diff --git "a/checkpoint-1188/trainer_state.json" "b/checkpoint-1188/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1188/trainer_state.json"
@@ -0,0 +1,7171 @@
+{
+  "best_metric": 0.3537317216396332,
+  "best_model_checkpoint": "accent_classification/checkpoint-792",
+  "epoch": 3.0,
+  "global_step": 1188,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 0.6956,
+      "step": 1
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 0.6979,
+      "step": 2
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8e-06,
+      "loss": 0.7014,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.6984,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3e-06,
+      "loss": 0.6978,
+      "step": 5
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6e-06,
+      "loss": 0.6783,
+      "step": 6
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 0.6956,
+      "step": 7
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 0.6994,
+      "step": 8
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.4e-06,
+      "loss": 0.6902,
+      "step": 9
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6e-06,
+      "loss": 0.6937,
+      "step": 10
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.6e-06,
+      "loss": 0.6497,
+      "step": 11
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 7.2e-06,
+      "loss": 0.6911,
+      "step": 12
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 7.8e-06,
+      "loss": 0.6904,
+      "step": 13
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 0.6854,
+      "step": 14
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9e-06,
+      "loss": 0.6843,
+      "step": 15
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 0.6851,
+      "step": 16
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.02e-05,
+      "loss": 0.6733,
+      "step": 17
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.08e-05,
+      "loss": 0.6819,
+      "step": 18
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.1400000000000001e-05,
+      "loss": 0.681,
+      "step": 19
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.2e-05,
+      "loss": 0.6998,
+      "step": 20
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.26e-05,
+      "loss": 0.6743,
+      "step": 21
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.32e-05,
+      "loss": 0.6818,
+      "step": 22
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.3800000000000002e-05,
+      "loss": 0.5331,
+      "step": 23
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.44e-05,
+      "loss": 0.6681,
+      "step": 24
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.5e-05,
+      "loss": 0.6955,
+      "step": 25
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.56e-05,
+      "loss": 0.7,
+      "step": 26
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.62e-05,
+      "loss": 0.6975,
+      "step": 27
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 0.6249,
+      "step": 28
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.74e-05,
+      "loss": 0.6979,
+      "step": 29
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.8e-05,
+      "loss": 0.6076,
+      "step": 30
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.86e-05,
+      "loss": 0.6091,
+      "step": 31
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 0.6438,
+      "step": 32
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.98e-05,
+      "loss": 0.6624,
+      "step": 33
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.04e-05,
+      "loss": 0.6452,
+      "step": 34
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.1e-05,
+      "loss": 0.6963,
+      "step": 35
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.16e-05,
+      "loss": 0.758,
+      "step": 36
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.22e-05,
+      "loss": 0.6471,
+      "step": 37
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.2800000000000002e-05,
+      "loss": 0.6313,
+      "step": 38
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.3400000000000003e-05,
+      "loss": 0.7019,
+      "step": 39
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.4e-05,
+      "loss": 0.6336,
+      "step": 40
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.4599999999999998e-05,
+      "loss": 0.7091,
+      "step": 41
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.52e-05,
+      "loss": 0.701,
+      "step": 42
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.58e-05,
+      "loss": 0.7036,
+      "step": 43
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.64e-05,
+      "loss": 0.7073,
+      "step": 44
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 0.614,
+      "step": 45
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.7600000000000003e-05,
+      "loss": 0.7543,
+      "step": 46
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.8199999999999998e-05,
+      "loss": 0.538,
+      "step": 47
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.88e-05,
+      "loss": 0.5179,
+      "step": 48
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.94e-05,
+      "loss": 0.6164,
+      "step": 49
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 3e-05,
+      "loss": 0.6093,
+      "step": 50
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.99923273657289e-05,
+      "loss": 0.6983,
+      "step": 51
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.9984654731457804e-05,
+      "loss": 0.6069,
+      "step": 52
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.9976982097186702e-05,
+      "loss": 0.5933,
+      "step": 53
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.99693094629156e-05,
+      "loss": 0.552,
+      "step": 54
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.9961636828644502e-05,
+      "loss": 0.5865,
+      "step": 55
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.9953964194373404e-05,
+      "loss": 0.7146,
+      "step": 56
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.9946291560102302e-05,
+      "loss": 0.5591,
+      "step": 57
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.9938618925831203e-05,
+      "loss": 0.7366,
+      "step": 58
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.9930946291560102e-05,
+      "loss": 0.7006,
+      "step": 59
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.9923273657289003e-05,
+      "loss": 0.4148,
+      "step": 60
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.9915601023017905e-05,
+      "loss": 0.9938,
+      "step": 61
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.9907928388746803e-05,
+      "loss": 0.911,
+      "step": 62
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.99002557544757e-05,
+      "loss": 0.5682,
+      "step": 63
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.9892583120204606e-05,
+      "loss": 0.5833,
+      "step": 64
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.9884910485933505e-05,
+      "loss": 0.7237,
+      "step": 65
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.9877237851662403e-05,
+      "loss": 0.4013,
+      "step": 66
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.9869565217391304e-05,
+      "loss": 0.9695,
+      "step": 67
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.9861892583120206e-05,
+      "loss": 0.568,
+      "step": 68
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.9854219948849104e-05,
+      "loss": 0.5697,
+      "step": 69
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.9846547314578006e-05,
+      "loss": 0.7675,
+      "step": 70
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.9838874680306904e-05,
+      "loss": 0.8831,
+      "step": 71
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.9831202046035806e-05,
+      "loss": 0.7008,
+      "step": 72
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.9823529411764707e-05,
+      "loss": 0.7085,
+      "step": 73
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.9815856777493606e-05,
+      "loss": 0.5374,
+      "step": 74
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.9808184143222507e-05,
+      "loss": 0.4131,
+      "step": 75
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.980051150895141e-05,
+      "loss": 0.7309,
+      "step": 76
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.9792838874680307e-05,
+      "loss": 0.4279,
+      "step": 77
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2.978516624040921e-05,
+      "loss": 0.5813,
+      "step": 78
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2.977749360613811e-05,
+      "loss": 0.6276,
+      "step": 79
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2.976982097186701e-05,
+      "loss": 0.7892,
+      "step": 80
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2.976214833759591e-05,
+      "loss": 0.7134,
+      "step": 81
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2.9754475703324808e-05,
+      "loss": 0.3682,
+      "step": 82
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2.974680306905371e-05,
+      "loss": 0.5562,
+      "step": 83
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2.973913043478261e-05,
+      "loss": 0.7158,
+      "step": 84
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2.973145780051151e-05,
+      "loss": 0.524,
+      "step": 85
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2.9723785166240408e-05,
+      "loss": 0.3372,
+      "step": 86
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2.9716112531969313e-05,
+      "loss": 0.4979,
+      "step": 87
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2.970843989769821e-05,
+      "loss": 0.3177,
+      "step": 88
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2.970076726342711e-05,
+      "loss": 0.3488,
+      "step": 89
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2.969309462915601e-05,
+      "loss": 0.5158,
+      "step": 90
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2.9685421994884913e-05,
+      "loss": 0.3208,
+      "step": 91
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2.967774936061381e-05,
+      "loss": 0.6654,
+      "step": 92
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2.9670076726342712e-05,
+      "loss": 0.5581,
+      "step": 93
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.966240409207161e-05,
+      "loss": 0.7456,
+      "step": 94
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.9654731457800512e-05,
+      "loss": 0.4437,
+      "step": 95
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.9647058823529414e-05,
+      "loss": 0.5525,
+      "step": 96
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.9639386189258312e-05,
+      "loss": 0.6673,
+      "step": 97
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2.963171355498721e-05,
+      "loss": 0.435,
+      "step": 98
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2.9624040920716115e-05,
+      "loss": 0.7415,
+      "step": 99
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2.9616368286445014e-05,
+      "loss": 0.2981,
+      "step": 100
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.9608695652173912e-05,
+      "loss": 0.5179,
+      "step": 101
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.9601023017902817e-05,
+      "loss": 0.4005,
+      "step": 102
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.9593350383631715e-05,
+      "loss": 0.4598,
+      "step": 103
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.9585677749360613e-05,
+      "loss": 0.2553,
+      "step": 104
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.9578005115089515e-05,
+      "loss": 0.4302,
+      "step": 105
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.9570332480818416e-05,
+      "loss": 0.7894,
+      "step": 106
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.9562659846547315e-05,
+      "loss": 0.213,
+      "step": 107
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.9554987212276216e-05,
+      "loss": 0.5507,
+      "step": 108
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2.9547314578005115e-05,
+      "loss": 0.7676,
+      "step": 109
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2.9539641943734016e-05,
+      "loss": 0.3418,
+      "step": 110
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2.9531969309462918e-05,
+      "loss": 0.8533,
+      "step": 111
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2.9524296675191816e-05,
+      "loss": 0.3659,
+      "step": 112
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.9516624040920714e-05,
+      "loss": 0.6482,
+      "step": 113
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.950895140664962e-05,
+      "loss": 0.5484,
+      "step": 114
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.9501278772378517e-05,
+      "loss": 0.4508,
+      "step": 115
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.9493606138107416e-05,
+      "loss": 0.4401,
+      "step": 116
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.9485933503836317e-05,
+      "loss": 0.5455,
+      "step": 117
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.947826086956522e-05,
+      "loss": 0.722,
+      "step": 118
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.9470588235294117e-05,
+      "loss": 0.2874,
+      "step": 119
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.946291560102302e-05,
+      "loss": 0.5973,
+      "step": 120
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2.9455242966751917e-05,
+      "loss": 0.5898,
+      "step": 121
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2.944757033248082e-05,
+      "loss": 1.0125,
+      "step": 122
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2.943989769820972e-05,
+      "loss": 0.2511,
+      "step": 123
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2.943222506393862e-05,
+      "loss": 0.545,
+      "step": 124
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.942455242966752e-05,
+      "loss": 0.593,
+      "step": 125
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.941687979539642e-05,
+      "loss": 0.2319,
+      "step": 126
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.940920716112532e-05,
+      "loss": 0.2215,
+      "step": 127
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.9401534526854218e-05,
+      "loss": 0.1964,
+      "step": 128
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2.9393861892583123e-05,
+      "loss": 0.1471,
+      "step": 129
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2.938618925831202e-05,
+      "loss": 1.5965,
+      "step": 130
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2.9378516624040923e-05,
+      "loss": 1.0772,
+      "step": 131
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2.937084398976982e-05,
+      "loss": 0.2005,
+      "step": 132
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2.9363171355498723e-05,
+      "loss": 0.9628,
+      "step": 133
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2.9355498721227624e-05,
+      "loss": 0.1503,
+      "step": 134
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2.9347826086956523e-05,
+      "loss": 0.6201,
+      "step": 135
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2.934015345268542e-05,
+      "loss": 0.6094,
+      "step": 136
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2.9332480818414326e-05,
+      "loss": 1.385,
+      "step": 137
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2.9324808184143224e-05,
+      "loss": 0.1552,
+      "step": 138
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2.9317135549872122e-05,
+      "loss": 0.5828,
+      "step": 139
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2.9309462915601024e-05,
+      "loss": 0.6018,
+      "step": 140
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2.9301790281329925e-05,
+      "loss": 1.0614,
+      "step": 141
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2.9294117647058824e-05,
+      "loss": 0.566,
+      "step": 142
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2.9286445012787725e-05,
+      "loss": 0.5958,
+      "step": 143
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2.9278772378516624e-05,
+      "loss": 0.5838,
+      "step": 144
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2.9271099744245525e-05,
+      "loss": 0.5909,
+      "step": 145
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2.9263427109974427e-05,
+      "loss": 0.5862,
+      "step": 146
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2.9255754475703325e-05,
+      "loss": 0.2031,
+      "step": 147
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2.9248081841432223e-05,
+      "loss": 0.1995,
+      "step": 148
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2.9240409207161128e-05,
+      "loss": 0.24,
+      "step": 149
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2.9232736572890026e-05,
+      "loss": 0.5845,
+      "step": 150
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2.9225063938618925e-05,
+      "loss": 0.964,
+      "step": 151
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2.921739130434783e-05,
+      "loss": 0.9595,
+      "step": 152
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.9209718670076728e-05,
+      "loss": 0.2033,
+      "step": 153
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.9202046035805626e-05,
+      "loss": 0.213,
+      "step": 154
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.9194373401534528e-05,
+      "loss": 0.2042,
+      "step": 155
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.918670076726343e-05,
+      "loss": 0.5783,
+      "step": 156
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2.9179028132992328e-05,
+      "loss": 1.0822,
+      "step": 157
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2.917135549872123e-05,
+      "loss": 1.2643,
+      "step": 158
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2.9163682864450127e-05,
+      "loss": 0.2593,
+      "step": 159
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2.915601023017903e-05,
+      "loss": 0.5783,
+      "step": 160
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2.914833759590793e-05,
+      "loss": 0.5721,
+      "step": 161
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2.914066496163683e-05,
+      "loss": 0.5808,
+      "step": 162
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2.9132992327365727e-05,
+      "loss": 1.2578,
+      "step": 163
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2.9125319693094632e-05,
+      "loss": 0.5709,
+      "step": 164
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.911764705882353e-05,
+      "loss": 0.2159,
+      "step": 165
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.910997442455243e-05,
+      "loss": 0.2329,
+      "step": 166
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.910230179028133e-05,
+      "loss": 0.926,
+      "step": 167
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.9094629156010232e-05,
+      "loss": 0.9095,
+      "step": 168
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.908695652173913e-05,
+      "loss": 0.5622,
+      "step": 169
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.907928388746803e-05,
+      "loss": 0.5703,
+      "step": 170
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.907161125319693e-05,
+      "loss": 0.5669,
+      "step": 171
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.906393861892583e-05,
+      "loss": 0.2637,
+      "step": 172
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.9056265984654733e-05,
+      "loss": 1.1077,
+      "step": 173
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.904859335038363e-05,
+      "loss": 0.5653,
+      "step": 174
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.9040920716112533e-05,
+      "loss": 0.5584,
+      "step": 175
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.9033248081841434e-05,
+      "loss": 0.5544,
+      "step": 176
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.9025575447570333e-05,
+      "loss": 0.5663,
+      "step": 177
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.901790281329923e-05,
+      "loss": 1.0808,
+      "step": 178
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.9010230179028136e-05,
+      "loss": 0.8736,
+      "step": 179
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.9002557544757034e-05,
+      "loss": 0.6435,
+      "step": 180
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.8994884910485932e-05,
+      "loss": 0.786,
+      "step": 181
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.8987212276214834e-05,
+      "loss": 0.5717,
+      "step": 182
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.8979539641943736e-05,
+      "loss": 0.5618,
+      "step": 183
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.8971867007672637e-05,
+      "loss": 0.5785,
+      "step": 184
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.8964194373401535e-05,
+      "loss": 0.7601,
+      "step": 185
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.8956521739130434e-05,
+      "loss": 0.8112,
+      "step": 186
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.894884910485934e-05,
+      "loss": 0.5663,
+      "step": 187
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.8941176470588237e-05,
+      "loss": 0.5631,
+      "step": 188
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.8933503836317135e-05,
+      "loss": 0.7363,
+      "step": 189
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.8925831202046037e-05,
+      "loss": 0.4552,
+      "step": 190
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.8918158567774938e-05,
+      "loss": 0.4599,
+      "step": 191
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.8910485933503836e-05,
+      "loss": 0.7848,
+      "step": 192
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.8902813299232738e-05,
+      "loss": 0.5907,
+      "step": 193
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.8895140664961636e-05,
+      "loss": 0.4264,
+      "step": 194
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.8887468030690538e-05,
+      "loss": 0.5697,
+      "step": 195
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.887979539641944e-05,
+      "loss": 0.7643,
+      "step": 196
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.8872122762148338e-05,
+      "loss": 0.3635,
+      "step": 197
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.8864450127877236e-05,
+      "loss": 0.3715,
+      "step": 198
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.885677749360614e-05,
+      "loss": 0.554,
+      "step": 199
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.884910485933504e-05,
+      "loss": 0.8044,
+      "step": 200
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.8841432225063937e-05,
+      "loss": 0.5673,
+      "step": 201
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.8833759590792842e-05,
+      "loss": 0.8042,
+      "step": 202
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.882608695652174e-05,
+      "loss": 0.5638,
+      "step": 203
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.881841432225064e-05,
+      "loss": 0.3584,
+      "step": 204
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.881074168797954e-05,
+      "loss": 0.8211,
+      "step": 205
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.8803069053708442e-05,
+      "loss": 0.5708,
+      "step": 206
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.879539641943734e-05,
+      "loss": 0.8301,
+      "step": 207
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.8787723785166242e-05,
+      "loss": 0.8229,
+      "step": 208
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.878005115089514e-05,
+      "loss": 0.5645,
+      "step": 209
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.8772378516624042e-05,
+      "loss": 0.3529,
+      "step": 210
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.8764705882352943e-05,
+      "loss": 0.5578,
+      "step": 211
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.875703324808184e-05,
+      "loss": 0.5659,
+      "step": 212
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.874936061381074e-05,
+      "loss": 0.7495,
+      "step": 213
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.8741687979539645e-05,
+      "loss": 0.3162,
+      "step": 214
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.8734015345268543e-05,
+      "loss": 0.5616,
+      "step": 215
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.872634271099744e-05,
+      "loss": 0.4034,
+      "step": 216
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.8718670076726343e-05,
+      "loss": 0.3163,
+      "step": 217
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.8710997442455245e-05,
+      "loss": 0.5514,
+      "step": 218
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.8703324808184143e-05,
+      "loss": 0.826,
+      "step": 219
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.8695652173913044e-05,
+      "loss": 0.8363,
+      "step": 220
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.8687979539641943e-05,
+      "loss": 0.8288,
+      "step": 221
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.8680306905370844e-05,
+      "loss": 0.5929,
+      "step": 222
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.8672634271099746e-05,
+      "loss": 0.565,
+      "step": 223
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.8664961636828644e-05,
+      "loss": 0.5642,
+      "step": 224
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.8657289002557546e-05,
+      "loss": 0.5609,
+      "step": 225
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.8649616368286447e-05,
+      "loss": 0.5641,
+      "step": 226
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.8641943734015345e-05,
+      "loss": 0.8535,
+      "step": 227
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.8634271099744244e-05,
+      "loss": 0.5652,
+      "step": 228
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.862659846547315e-05,
+      "loss": 0.4397,
+      "step": 229
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.8618925831202047e-05,
+      "loss": 0.567,
+      "step": 230
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.8611253196930945e-05,
+      "loss": 0.7517,
+      "step": 231
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.8603580562659847e-05,
+      "loss": 0.5839,
+      "step": 232
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.859590792838875e-05,
+      "loss": 0.5677,
+      "step": 233
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.8588235294117647e-05,
+      "loss": 0.5678,
+      "step": 234
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.8580562659846548e-05,
+      "loss": 1.3045,
+      "step": 235
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.8572890025575446e-05,
+      "loss": 0.804,
+      "step": 236
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.856521739130435e-05,
+      "loss": 1.0303,
+      "step": 237
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.855754475703325e-05,
+      "loss": 0.3303,
+      "step": 238
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.8549872122762148e-05,
+      "loss": 0.5677,
+      "step": 239
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.854219948849105e-05,
+      "loss": 0.7972,
+      "step": 240
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.853452685421995e-05,
+      "loss": 0.5644,
+      "step": 241
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.852685421994885e-05,
+      "loss": 0.9294,
+      "step": 242
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.851918158567775e-05,
+      "loss": 0.5665,
+      "step": 243
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.851150895140665e-05,
+      "loss": 0.7963,
+      "step": 244
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.850383631713555e-05,
+      "loss": 0.5659,
+      "step": 245
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.8496163682864452e-05,
+      "loss": 0.5718,
+      "step": 246
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.848849104859335e-05,
+      "loss": 0.3526,
+      "step": 247
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.8480818414322252e-05,
+      "loss": 0.574,
+      "step": 248
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.8473145780051154e-05,
+      "loss": 0.9996,
+      "step": 249
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.8465473145780052e-05,
+      "loss": 0.7634,
+      "step": 250
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.845780051150895e-05,
+      "loss": 0.5639,
+      "step": 251
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.8450127877237855e-05,
+      "loss": 0.5737,
+      "step": 252
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.8442455242966753e-05,
+      "loss": 0.7793,
+      "step": 253
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.8434782608695652e-05,
+      "loss": 0.6769,
+      "step": 254
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.8427109974424553e-05,
+      "loss": 0.5748,
+      "step": 255
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.8419437340153455e-05,
+      "loss": 0.3602,
+      "step": 256
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.8411764705882353e-05,
+      "loss": 0.5714,
+      "step": 257
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.8404092071611255e-05,
+      "loss": 0.5738,
+      "step": 258
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.8396419437340153e-05,
+      "loss": 0.5719,
+      "step": 259
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.8388746803069055e-05,
+      "loss": 0.5753,
+      "step": 260
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.8381074168797956e-05,
+      "loss": 0.5717,
+      "step": 261
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.8373401534526854e-05,
+      "loss": 0.9794,
+      "step": 262
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.8365728900255753e-05,
+      "loss": 0.5737,
+      "step": 263
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.8358056265984658e-05,
+      "loss": 0.5723,
+      "step": 264
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.8350383631713556e-05,
+      "loss": 0.5656,
+      "step": 265
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.8342710997442454e-05,
+      "loss": 0.4259,
+      "step": 266
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.8335038363171356e-05,
+      "loss": 0.3076,
+      "step": 267
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.8327365728900257e-05,
+      "loss": 0.338,
+      "step": 268
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.8319693094629156e-05,
+      "loss": 0.5612,
+      "step": 269
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.8312020460358057e-05,
+      "loss": 0.5249,
+      "step": 270
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.8304347826086955e-05,
+      "loss": 0.5671,
+      "step": 271
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.8296675191815857e-05,
+      "loss": 0.5574,
+      "step": 272
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.828900255754476e-05,
+      "loss": 0.8152,
+      "step": 273
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.8281329923273657e-05,
+      "loss": 0.5647,
+      "step": 274
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.827365728900256e-05,
+      "loss": 0.5632,
+      "step": 275
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.826598465473146e-05,
+      "loss": 0.3391,
+      "step": 276
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.8258312020460358e-05,
+      "loss": 0.55,
+      "step": 277
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.8250639386189257e-05,
+      "loss": 0.5672,
+      "step": 278
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.824296675191816e-05,
+      "loss": 0.285,
+      "step": 279
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.823529411764706e-05,
+      "loss": 0.2809,
+      "step": 280
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.8227621483375958e-05,
+      "loss": 0.5552,
+      "step": 281
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.821994884910486e-05,
+      "loss": 0.563,
+      "step": 282
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.821227621483376e-05,
+      "loss": 0.5628,
+      "step": 283
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2.820460358056266e-05,
+      "loss": 0.2548,
+      "step": 284
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2.819693094629156e-05,
+      "loss": 0.871,
+      "step": 285
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2.818925831202046e-05,
+      "loss": 0.8797,
+      "step": 286
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2.818158567774936e-05,
+      "loss": 0.2453,
+      "step": 287
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.8173913043478262e-05,
+      "loss": 0.2464,
+      "step": 288
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.816624040920716e-05,
+      "loss": 0.3134,
+      "step": 289
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.8158567774936062e-05,
+      "loss": 0.2431,
+      "step": 290
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.8150895140664964e-05,
+      "loss": 0.2308,
+      "step": 291
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2.8143222506393862e-05,
+      "loss": 0.2173,
+      "step": 292
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2.8135549872122764e-05,
+      "loss": 0.5665,
+      "step": 293
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2.8127877237851662e-05,
+      "loss": 0.5592,
+      "step": 294
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2.8120204603580564e-05,
+      "loss": 0.473,
+      "step": 295
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.8112531969309465e-05,
+      "loss": 0.9458,
+      "step": 296
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.8104859335038363e-05,
+      "loss": 0.1879,
+      "step": 297
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.8097186700767265e-05,
+      "loss": 0.6797,
+      "step": 298
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.8089514066496167e-05,
+      "loss": 0.7592,
+      "step": 299
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.8081841432225065e-05,
+      "loss": 0.4109,
+      "step": 300
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.8074168797953963e-05,
+      "loss": 0.5821,
+      "step": 301
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.8066496163682868e-05,
+      "loss": 0.2057,
+      "step": 302
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.8058823529411766e-05,
+      "loss": 0.5322,
+      "step": 303
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.8051150895140665e-05,
+      "loss": 0.5208,
+      "step": 304
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.8043478260869566e-05,
+      "loss": 0.6799,
+      "step": 305
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.8035805626598468e-05,
+      "loss": 0.5894,
+      "step": 306
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.8028132992327366e-05,
+      "loss": 0.4051,
+      "step": 307
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.8020460358056268e-05,
+      "loss": 0.3695,
+      "step": 308
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.8012787723785166e-05,
+      "loss": 0.4869,
+      "step": 309
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.8005115089514067e-05,
+      "loss": 0.6075,
+      "step": 310
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.799744245524297e-05,
+      "loss": 0.4245,
+      "step": 311
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.7989769820971867e-05,
+      "loss": 0.5122,
+      "step": 312
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.7982097186700765e-05,
+      "loss": 0.452,
+      "step": 313
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.797442455242967e-05,
+      "loss": 0.9568,
+      "step": 314
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.796675191815857e-05,
+      "loss": 0.5846,
+      "step": 315
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.7959079283887467e-05,
+      "loss": 0.3743,
+      "step": 316
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.795140664961637e-05,
+      "loss": 0.1922,
+      "step": 317
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.794373401534527e-05,
+      "loss": 0.5303,
+      "step": 318
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.793606138107417e-05,
+      "loss": 0.7253,
+      "step": 319
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.792838874680307e-05,
+      "loss": 0.2794,
+      "step": 320
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.7920716112531968e-05,
+      "loss": 0.577,
+      "step": 321
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.791304347826087e-05,
+      "loss": 0.614,
+      "step": 322
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.790537084398977e-05,
+      "loss": 0.5882,
+      "step": 323
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.789769820971867e-05,
+      "loss": 0.2503,
+      "step": 324
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.789002557544757e-05,
+      "loss": 0.4195,
+      "step": 325
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.7882352941176473e-05,
+      "loss": 0.2959,
+      "step": 326
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.787468030690537e-05,
+      "loss": 0.3732,
+      "step": 327
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.786700767263427e-05,
+      "loss": 0.5729,
+      "step": 328
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.7859335038363174e-05,
+      "loss": 0.3179,
+      "step": 329
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.7851662404092073e-05,
+      "loss": 0.6823,
+      "step": 330
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.784398976982097e-05,
+      "loss": 0.2663,
+      "step": 331
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.7836317135549872e-05,
+      "loss": 0.4436,
+      "step": 332
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.7828644501278774e-05,
+      "loss": 0.3089,
+      "step": 333
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.7820971867007672e-05,
+      "loss": 0.5888,
+      "step": 334
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.7813299232736574e-05,
+      "loss": 0.2115,
+      "step": 335
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.7805626598465472e-05,
+      "loss": 0.4005,
+      "step": 336
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.7797953964194374e-05,
+      "loss": 0.3097,
+      "step": 337
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.7790281329923275e-05,
+      "loss": 0.5847,
+      "step": 338
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.7782608695652174e-05,
+      "loss": 1.0164,
+      "step": 339
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.7774936061381072e-05,
+      "loss": 1.1321,
+      "step": 340
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.7767263427109977e-05,
+      "loss": 0.592,
+      "step": 341
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.7759590792838875e-05,
+      "loss": 1.0191,
+      "step": 342
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.7751918158567777e-05,
+      "loss": 0.9199,
+      "step": 343
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.7744245524296675e-05,
+      "loss": 0.2874,
+      "step": 344
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.7736572890025576e-05,
+      "loss": 0.2469,
+      "step": 345
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.7728900255754478e-05,
+      "loss": 0.8364,
+      "step": 346
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.7721227621483376e-05,
+      "loss": 0.7365,
+      "step": 347
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.7713554987212278e-05,
+      "loss": 0.5595,
+      "step": 348
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.770588235294118e-05,
+      "loss": 0.2546,
+      "step": 349
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.7698209718670078e-05,
+      "loss": 0.8183,
+      "step": 350
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.7690537084398976e-05,
+      "loss": 0.5385,
+      "step": 351
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.768286445012788e-05,
+      "loss": 0.2763,
+      "step": 352
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.767519181585678e-05,
+      "loss": 0.3045,
+      "step": 353
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.7667519181585677e-05,
+      "loss": 0.3354,
+      "step": 354
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.765984654731458e-05,
+      "loss": 0.9366,
+      "step": 355
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.765217391304348e-05,
+      "loss": 0.6212,
+      "step": 356
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.764450127877238e-05,
+      "loss": 0.1161,
+      "step": 357
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.763682864450128e-05,
+      "loss": 0.4929,
+      "step": 358
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.762915601023018e-05,
+      "loss": 0.8167,
+      "step": 359
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.762148337595908e-05,
+      "loss": 0.2521,
+      "step": 360
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.7613810741687982e-05,
+      "loss": 0.4174,
+      "step": 361
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.760613810741688e-05,
+      "loss": 0.3822,
+      "step": 362
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.759846547314578e-05,
+      "loss": 0.5355,
+      "step": 363
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.7590792838874683e-05,
+      "loss": 0.561,
+      "step": 364
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.758312020460358e-05,
+      "loss": 0.5126,
+      "step": 365
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.757544757033248e-05,
+      "loss": 0.2301,
+      "step": 366
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.756777493606138e-05,
+      "loss": 0.1698,
+      "step": 367
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.7560102301790283e-05,
+      "loss": 0.1634,
+      "step": 368
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.755242966751918e-05,
+      "loss": 0.5955,
+      "step": 369
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.7544757033248083e-05,
+      "loss": 0.1568,
+      "step": 370
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.753708439897698e-05,
+      "loss": 0.5167,
+      "step": 371
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.7529411764705883e-05,
+      "loss": 1.0028,
+      "step": 372
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.7521739130434784e-05,
+      "loss": 1.2567,
+      "step": 373
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.7514066496163682e-05,
+      "loss": 0.6094,
+      "step": 374
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.7506393861892584e-05,
+      "loss": 0.6059,
+      "step": 375
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.7498721227621486e-05,
+      "loss": 1.0593,
+      "step": 376
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.7491048593350384e-05,
+      "loss": 0.1506,
+      "step": 377
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.7483375959079282e-05,
+      "loss": 1.5106,
+      "step": 378
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.7475703324808187e-05,
+      "loss": 0.1827,
+      "step": 379
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.7468030690537085e-05,
+      "loss": 1.4961,
+      "step": 380
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.7460358056265984e-05,
+      "loss": 1.0456,
+      "step": 381
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.7452685421994885e-05,
+      "loss": 1.2063,
+      "step": 382
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.7445012787723787e-05,
+      "loss": 0.5837,
+      "step": 383
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.7437340153452685e-05,
+      "loss": 1.4536,
+      "step": 384
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.7429667519181587e-05,
+      "loss": 0.5803,
+      "step": 385
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.7421994884910485e-05,
+      "loss": 0.5896,
+      "step": 386
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.7414322250639386e-05,
+      "loss": 0.1822,
+      "step": 387
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.7406649616368288e-05,
+      "loss": 1.7893,
+      "step": 388
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.7398976982097186e-05,
+      "loss": 0.9723,
+      "step": 389
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.7391304347826085e-05,
+      "loss": 1.2815,
+      "step": 390
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.738363171355499e-05,
+      "loss": 0.5765,
+      "step": 391
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.7375959079283888e-05,
+      "loss": 1.2429,
+      "step": 392
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.7368286445012786e-05,
+      "loss": 0.1343,
+      "step": 393
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.7360613810741688e-05,
+      "loss": 0.2306,
+      "step": 394
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.735294117647059e-05,
+      "loss": 0.624,
+      "step": 395
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.734526854219949e-05,
+      "loss": 0.2371,
+      "step": 396
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.7348484848484849,
+      "eval_loss": 0.5804409980773926,
+      "eval_runtime": 57.3336,
+      "eval_samples_per_second": 6.907,
+      "eval_steps_per_second": 1.727,
+      "step": 396
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.733759590792839e-05,
+      "loss": 0.8989,
+      "step": 397
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.732992327365729e-05,
+      "loss": 0.5664,
+      "step": 398
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7322250639386192e-05,
+      "loss": 0.8852,
+      "step": 399
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.731457800511509e-05,
+      "loss": 0.5661,
+      "step": 400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.730690537084399e-05,
+      "loss": 0.5648,
+      "step": 401
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7299232736572894e-05,
+      "loss": 0.7333,
+      "step": 402
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7291560102301792e-05,
+      "loss": 0.3037,
+      "step": 403
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.728388746803069e-05,
+      "loss": 0.1463,
+      "step": 404
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7276214833759592e-05,
+      "loss": 0.2702,
+      "step": 405
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.7268542199488493e-05,
+      "loss": 0.2693,
+      "step": 406
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.726086956521739e-05,
+      "loss": 0.5647,
+      "step": 407
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.7253196930946293e-05,
+      "loss": 0.8507,
+      "step": 408
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.724552429667519e-05,
+      "loss": 0.8645,
+      "step": 409
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.7237851662404093e-05,
+      "loss": 0.2649,
+      "step": 410
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.7230179028132995e-05,
+      "loss": 0.6942,
+      "step": 411
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.7222506393861893e-05,
+      "loss": 0.5656,
+      "step": 412
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.721483375959079e-05,
+      "loss": 0.5489,
+      "step": 413
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.7207161125319696e-05,
+      "loss": 0.2652,
+      "step": 414
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.7199488491048594e-05,
+      "loss": 0.8676,
+      "step": 415
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.7191815856777493e-05,
+      "loss": 0.8621,
+      "step": 416
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.7184143222506394e-05,
+      "loss": 0.2641,
+      "step": 417
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.7176470588235296e-05,
+      "loss": 0.859,
+      "step": 418
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.7168797953964194e-05,
+      "loss": 0.6492,
+      "step": 419
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.7161125319693096e-05,
+      "loss": 1.1494,
+      "step": 420
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.7153452685421994e-05,
+      "loss": 0.8278,
+      "step": 421
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.7145780051150895e-05,
+      "loss": 0.764,
+      "step": 422
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.7138107416879797e-05,
+      "loss": 0.564,
+      "step": 423
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.7130434782608695e-05,
+      "loss": 0.8051,
+      "step": 424
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.7122762148337597e-05,
+      "loss": 0.3052,
+      "step": 425
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.71150895140665e-05,
+      "loss": 0.2945,
+      "step": 426
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.7107416879795397e-05,
+      "loss": 0.4118,
+      "step": 427
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.7099744245524295e-05,
+      "loss": 0.559,
+      "step": 428
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.70920716112532e-05,
+      "loss": 0.5593,
+      "step": 429
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.7084398976982098e-05,
+      "loss": 0.564,
+      "step": 430
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.7076726342710996e-05,
+      "loss": 0.4454,
+      "step": 431
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.7069053708439898e-05,
+      "loss": 0.4381,
+      "step": 432
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.70613810741688e-05,
+      "loss": 0.3718,
+      "step": 433
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.7053708439897698e-05,
+      "loss": 0.6627,
+      "step": 434
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.70460358056266e-05,
+      "loss": 0.4004,
+      "step": 435
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.7038363171355498e-05,
+      "loss": 0.3855,
+      "step": 436
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.70306905370844e-05,
+      "loss": 0.4799,
+      "step": 437
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.70230179028133e-05,
+      "loss": 0.6714,
+      "step": 438
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.70153452685422e-05,
+      "loss": 0.622,
+      "step": 439
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.7007672634271097e-05,
+      "loss": 0.3916,
+      "step": 440
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 0.5487,
+      "step": 441
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.69923273657289e-05,
+      "loss": 0.376,
+      "step": 442
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.69846547314578e-05,
+      "loss": 0.3616,
+      "step": 443
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.69769820971867e-05,
+      "loss": 0.8747,
+      "step": 444
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.6969309462915602e-05,
+      "loss": 0.283,
+      "step": 445
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.69616368286445e-05,
+      "loss": 0.3988,
+      "step": 446
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.6953964194373402e-05,
+      "loss": 0.4569,
+      "step": 447
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.6946291560102303e-05,
+      "loss": 0.2005,
+      "step": 448
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.6938618925831205e-05,
+      "loss": 0.3087,
+      "step": 449
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.6930946291560103e-05,
+      "loss": 0.5814,
+      "step": 450
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.6923273657289e-05,
+      "loss": 0.2995,
+      "step": 451
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.6915601023017907e-05,
+      "loss": 0.7806,
+      "step": 452
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.6907928388746805e-05,
+      "loss": 0.2666,
+      "step": 453
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.6900255754475703e-05,
+      "loss": 0.2571,
+      "step": 454
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.6892583120204605e-05,
+      "loss": 0.5931,
+      "step": 455
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.6884910485933506e-05,
+      "loss": 0.475,
+      "step": 456
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.6877237851662404e-05,
+      "loss": 0.2372,
+      "step": 457
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.6869565217391306e-05,
+      "loss": 0.1577,
+      "step": 458
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.6861892583120204e-05,
+      "loss": 0.1953,
+      "step": 459
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.6854219948849106e-05,
+      "loss": 0.1931,
+      "step": 460
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.6846547314578008e-05,
+      "loss": 0.1829,
+      "step": 461
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.6838874680306906e-05,
+      "loss": 0.1601,
+      "step": 462
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.6831202046035804e-05,
+      "loss": 0.2505,
+      "step": 463
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.682352941176471e-05,
+      "loss": 0.181,
+      "step": 464
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.6815856777493607e-05,
+      "loss": 0.5978,
+      "step": 465
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.6808184143222505e-05,
+      "loss": 0.5458,
+      "step": 466
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.6800511508951407e-05,
+      "loss": 0.1289,
+      "step": 467
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.679283887468031e-05,
+      "loss": 0.4383,
+      "step": 468
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.6785166240409207e-05,
+      "loss": 0.5892,
+      "step": 469
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.677749360613811e-05,
+      "loss": 1.3565,
+      "step": 470
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.676982097186701e-05,
+      "loss": 0.7267,
+      "step": 471
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.6762148337595908e-05,
+      "loss": 0.1354,
+      "step": 472
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.675447570332481e-05,
+      "loss": 0.9291,
+      "step": 473
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.6746803069053708e-05,
+      "loss": 0.5265,
+      "step": 474
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.673913043478261e-05,
+      "loss": 0.4841,
+      "step": 475
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.673145780051151e-05,
+      "loss": 1.2391,
+      "step": 476
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.672378516624041e-05,
+      "loss": 0.4924,
+      "step": 477
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.6716112531969308e-05,
+      "loss": 1.0076,
+      "step": 478
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.6708439897698213e-05,
+      "loss": 0.6314,
+      "step": 479
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.670076726342711e-05,
+      "loss": 0.4237,
+      "step": 480
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.669309462915601e-05,
+      "loss": 0.1077,
+      "step": 481
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.668542199488491e-05,
+      "loss": 0.3038,
+      "step": 482
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.6677749360613812e-05,
+      "loss": 0.6081,
+      "step": 483
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.667007672634271e-05,
+      "loss": 0.1078,
+      "step": 484
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.6662404092071612e-05,
+      "loss": 0.4296,
+      "step": 485
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.665473145780051e-05,
+      "loss": 0.3981,
+      "step": 486
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.6647058823529412e-05,
+      "loss": 0.5853,
+      "step": 487
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.6639386189258314e-05,
+      "loss": 0.802,
+      "step": 488
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.6631713554987212e-05,
+      "loss": 0.4576,
+      "step": 489
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.662404092071611e-05,
+      "loss": 0.2682,
+      "step": 490
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.6616368286445015e-05,
+      "loss": 0.1991,
+      "step": 491
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.6608695652173913e-05,
+      "loss": 0.1409,
+      "step": 492
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.660102301790281e-05,
+      "loss": 0.3869,
+      "step": 493
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.6593350383631713e-05,
+      "loss": 0.5301,
+      "step": 494
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.6585677749360615e-05,
+      "loss": 0.8116,
+      "step": 495
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.6578005115089513e-05,
+      "loss": 0.0984,
+      "step": 496
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.6570332480818415e-05,
+      "loss": 0.7773,
+      "step": 497
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.6562659846547316e-05,
+      "loss": 1.3522,
+      "step": 498
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.6554987212276215e-05,
+      "loss": 0.0862,
+      "step": 499
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.6547314578005116e-05,
+      "loss": 0.0894,
+      "step": 500
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.6539641943734014e-05,
+      "loss": 0.0885,
+      "step": 501
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.653196930946292e-05,
+      "loss": 0.3031,
+      "step": 502
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.6524296675191818e-05,
+      "loss": 0.187,
+      "step": 503
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.6516624040920716e-05,
+      "loss": 1.3063,
+      "step": 504
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.6508951406649617e-05,
+      "loss": 0.6847,
+      "step": 505
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.650127877237852e-05,
+      "loss": 0.078,
+      "step": 506
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.6493606138107417e-05,
+      "loss": 0.1265,
+      "step": 507
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.648593350383632e-05,
+      "loss": 0.0804,
+      "step": 508
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.6478260869565217e-05,
+      "loss": 0.6976,
+      "step": 509
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.647058823529412e-05,
+      "loss": 0.7977,
+      "step": 510
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.646291560102302e-05,
+      "loss": 0.413,
+      "step": 511
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.645524296675192e-05,
+      "loss": 0.6944,
+      "step": 512
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.6447570332480817e-05,
+      "loss": 0.7177,
+      "step": 513
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.6439897698209722e-05,
+      "loss": 0.0581,
+      "step": 514
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.643222506393862e-05,
+      "loss": 0.9308,
+      "step": 515
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.6424552429667518e-05,
+      "loss": 0.2518,
+      "step": 516
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.641687979539642e-05,
+      "loss": 0.1014,
+      "step": 517
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.640920716112532e-05,
+      "loss": 0.1641,
+      "step": 518
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.640153452685422e-05,
+      "loss": 0.101,
+      "step": 519
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.639386189258312e-05,
+      "loss": 0.1097,
+      "step": 520
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.6386189258312023e-05,
+      "loss": 0.987,
+      "step": 521
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.637851662404092e-05,
+      "loss": 0.6465,
+      "step": 522
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.6370843989769823e-05,
+      "loss": 0.0914,
+      "step": 523
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.636317135549872e-05,
+      "loss": 0.139,
+      "step": 524
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.6355498721227623e-05,
+      "loss": 0.6455,
+      "step": 525
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.6347826086956524e-05,
+      "loss": 0.204,
+      "step": 526
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.6340153452685422e-05,
+      "loss": 0.1269,
+      "step": 527
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.633248081841432e-05,
+      "loss": 0.2214,
+      "step": 528
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.6324808184143226e-05,
+      "loss": 0.0736,
+      "step": 529
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.6317135549872124e-05,
+      "loss": 0.0714,
+      "step": 530
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.6309462915601022e-05,
+      "loss": 0.1388,
+      "step": 531
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.6301790281329924e-05,
+      "loss": 0.7758,
+      "step": 532
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.6294117647058825e-05,
+      "loss": 0.3388,
+      "step": 533
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.6286445012787724e-05,
+      "loss": 1.1831,
+      "step": 534
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.6278772378516625e-05,
+      "loss": 0.1166,
+      "step": 535
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.6271099744245523e-05,
+      "loss": 0.2389,
+      "step": 536
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.6263427109974425e-05,
+      "loss": 0.1622,
+      "step": 537
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.6255754475703327e-05,
+      "loss": 0.1015,
+      "step": 538
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.6248081841432225e-05,
+      "loss": 0.0658,
+      "step": 539
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.6240409207161123e-05,
+      "loss": 0.0954,
+      "step": 540
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.6232736572890028e-05,
+      "loss": 0.1182,
+      "step": 541
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.6225063938618926e-05,
+      "loss": 0.0605,
+      "step": 542
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.6217391304347824e-05,
+      "loss": 0.9265,
+      "step": 543
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.6209718670076726e-05,
+      "loss": 1.2994,
+      "step": 544
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.6202046035805628e-05,
+      "loss": 0.081,
+      "step": 545
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.6194373401534526e-05,
+      "loss": 0.5962,
+      "step": 546
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.6186700767263428e-05,
+      "loss": 0.0665,
+      "step": 547
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.617902813299233e-05,
+      "loss": 0.0655,
+      "step": 548
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.6171355498721227e-05,
+      "loss": 0.0642,
+      "step": 549
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.616368286445013e-05,
+      "loss": 0.0834,
+      "step": 550
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.6156010230179027e-05,
+      "loss": 0.047,
+      "step": 551
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.6148337595907932e-05,
+      "loss": 0.0642,
+      "step": 552
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.614066496163683e-05,
+      "loss": 0.0594,
+      "step": 553
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.613299232736573e-05,
+      "loss": 0.0573,
+      "step": 554
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.612531969309463e-05,
+      "loss": 0.1101,
+      "step": 555
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.6117647058823532e-05,
+      "loss": 0.5121,
+      "step": 556
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.610997442455243e-05,
+      "loss": 0.0374,
+      "step": 557
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.6102301790281332e-05,
+      "loss": 0.0957,
+      "step": 558
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.609462915601023e-05,
+      "loss": 0.0488,
+      "step": 559
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.608695652173913e-05,
+      "loss": 0.0482,
+      "step": 560
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.6079283887468033e-05,
+      "loss": 1.2041,
+      "step": 561
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.607161125319693e-05,
+      "loss": 0.8833,
+      "step": 562
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.606393861892583e-05,
+      "loss": 0.1544,
+      "step": 563
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.6056265984654735e-05,
+      "loss": 0.046,
+      "step": 564
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.6048593350383633e-05,
+      "loss": 0.9073,
+      "step": 565
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.604092071611253e-05,
+      "loss": 0.052,
+      "step": 566
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.6033248081841433e-05,
+      "loss": 0.0308,
+      "step": 567
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.6025575447570334e-05,
+      "loss": 0.7032,
+      "step": 568
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.6017902813299232e-05,
+      "loss": 0.6947,
+      "step": 569
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.6010230179028134e-05,
+      "loss": 1.5603,
+      "step": 570
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.6002557544757036e-05,
+      "loss": 0.8836,
+      "step": 571
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.5994884910485934e-05,
+      "loss": 1.7439,
+      "step": 572
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.5987212276214836e-05,
+      "loss": 0.0313,
+      "step": 573
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.5979539641943734e-05,
+      "loss": 0.0317,
+      "step": 574
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.5971867007672635e-05,
+      "loss": 0.0325,
+      "step": 575
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.5964194373401537e-05,
+      "loss": 0.5,
+      "step": 576
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.5956521739130435e-05,
+      "loss": 0.0675,
+      "step": 577
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.5948849104859333e-05,
+      "loss": 0.6364,
+      "step": 578
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.594117647058824e-05,
+      "loss": 0.0696,
+      "step": 579
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.5933503836317137e-05,
+      "loss": 0.6195,
+      "step": 580
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.5925831202046035e-05,
+      "loss": 0.5796,
+      "step": 581
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.5918158567774937e-05,
+      "loss": 0.0832,
+      "step": 582
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.5910485933503838e-05,
+      "loss": 0.0794,
+      "step": 583
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.5902813299232736e-05,
+      "loss": 0.5851,
+      "step": 584
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.5895140664961638e-05,
+      "loss": 0.104,
+      "step": 585
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.5887468030690536e-05,
+      "loss": 1.0492,
+      "step": 586
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.5879795396419438e-05,
+      "loss": 0.0645,
+      "step": 587
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.587212276214834e-05,
+      "loss": 0.514,
+      "step": 588
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.5864450127877238e-05,
+      "loss": 0.0701,
+      "step": 589
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.5856777493606136e-05,
+      "loss": 0.5409,
+      "step": 590
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.584910485933504e-05,
+      "loss": 0.0536,
+      "step": 591
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.584143222506394e-05,
+      "loss": 0.1728,
+      "step": 592
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.5833759590792837e-05,
+      "loss": 0.0845,
+      "step": 593
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.582608695652174e-05,
+      "loss": 0.2217,
+      "step": 594
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.581841432225064e-05,
+      "loss": 0.1397,
+      "step": 595
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.581074168797954e-05,
+      "loss": 0.4119,
+      "step": 596
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.580306905370844e-05,
+      "loss": 0.3638,
+      "step": 597
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.5795396419437342e-05,
+      "loss": 0.0781,
+      "step": 598
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.578772378516624e-05,
+      "loss": 0.5926,
+      "step": 599
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.5780051150895142e-05,
+      "loss": 0.1378,
+      "step": 600
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.577237851662404e-05,
+      "loss": 0.142,
+      "step": 601
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.576470588235294e-05,
+      "loss": 0.1722,
+      "step": 602
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.5757033248081843e-05,
+      "loss": 0.1529,
+      "step": 603
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.574936061381074e-05,
+      "loss": 1.6083,
+      "step": 604
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.5741687979539643e-05,
+      "loss": 1.8267,
+      "step": 605
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.5734015345268545e-05,
+      "loss": 0.0535,
+      "step": 606
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.5726342710997443e-05,
+      "loss": 0.9945,
+      "step": 607
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.5718670076726345e-05,
+      "loss": 0.9133,
+      "step": 608
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.5710997442455243e-05,
+      "loss": 0.5295,
+      "step": 609
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.5703324808184144e-05,
+      "loss": 0.058,
+      "step": 610
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.5695652173913046e-05,
+      "loss": 0.4789,
+      "step": 611
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.5687979539641944e-05,
+      "loss": 0.0572,
+      "step": 612
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.5680306905370842e-05,
+      "loss": 0.0897,
+      "step": 613
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.5672634271099747e-05,
+      "loss": 0.056,
+      "step": 614
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.5664961636828646e-05,
+      "loss": 0.0278,
+      "step": 615
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.5657289002557544e-05,
+      "loss": 0.9947,
+      "step": 616
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.5649616368286445e-05,
+      "loss": 0.0627,
+      "step": 617
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.5641943734015347e-05,
+      "loss": 0.9335,
+      "step": 618
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.5634271099744245e-05,
+      "loss": 0.9793,
+      "step": 619
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.5626598465473147e-05,
+      "loss": 1.8128,
+      "step": 620
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.561892583120205e-05,
+      "loss": 1.8015,
+      "step": 621
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.5611253196930947e-05,
+      "loss": 0.9259,
+      "step": 622
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.560358056265985e-05,
+      "loss": 0.9618,
+      "step": 623
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.5595907928388747e-05,
+      "loss": 0.0808,
+      "step": 624
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.5588235294117648e-05,
+      "loss": 1.7161,
+      "step": 625
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.558056265984655e-05,
+      "loss": 0.8879,
+      "step": 626
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.5572890025575448e-05,
+      "loss": 0.043,
+      "step": 627
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.5565217391304346e-05,
+      "loss": 0.9449,
+      "step": 628
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.555754475703325e-05,
+      "loss": 0.0624,
+      "step": 629
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.554987212276215e-05,
+      "loss": 0.0924,
+      "step": 630
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.5542199488491048e-05,
+      "loss": 0.0715,
+      "step": 631
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.553452685421995e-05,
+      "loss": 0.4995,
+      "step": 632
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.552685421994885e-05,
+      "loss": 0.5576,
+      "step": 633
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.551918158567775e-05,
+      "loss": 0.06,
+      "step": 634
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.551150895140665e-05,
+      "loss": 0.0873,
+      "step": 635
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.550383631713555e-05,
+      "loss": 0.6868,
+      "step": 636
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.549616368286445e-05,
+      "loss": 0.8487,
+      "step": 637
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.5488491048593352e-05,
+      "loss": 0.5641,
+      "step": 638
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.548081841432225e-05,
+      "loss": 1.1343,
+      "step": 639
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.547314578005115e-05,
+      "loss": 0.0614,
+      "step": 640
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.5465473145780054e-05,
+      "loss": 0.9414,
+      "step": 641
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.5457800511508952e-05,
+      "loss": 0.9226,
+      "step": 642
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.545012787723785e-05,
+      "loss": 0.3323,
+      "step": 643
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.5442455242966752e-05,
+      "loss": 0.9315,
+      "step": 644
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.5434782608695653e-05,
+      "loss": 0.5539,
+      "step": 645
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.542710997442455e-05,
+      "loss": 0.3376,
+      "step": 646
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.5419437340153453e-05,
+      "loss": 0.0694,
+      "step": 647
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.5411764705882355e-05,
+      "loss": 0.0706,
+      "step": 648
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.5404092071611253e-05,
+      "loss": 0.105,
+      "step": 649
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.5396419437340155e-05,
+      "loss": 0.912,
+      "step": 650
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.5388746803069053e-05,
+      "loss": 0.0998,
+      "step": 651
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.5381074168797954e-05,
+      "loss": 0.2247,
+      "step": 652
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.5373401534526856e-05,
+      "loss": 0.2869,
+      "step": 653
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.5365728900255754e-05,
+      "loss": 0.532,
+      "step": 654
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.5358056265984653e-05,
+      "loss": 0.4723,
+      "step": 655
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.5350383631713558e-05,
+      "loss": 0.1178,
+      "step": 656
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.5342710997442456e-05,
+      "loss": 0.0566,
+      "step": 657
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.5335038363171357e-05,
+      "loss": 0.059,
+      "step": 658
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.5327365728900256e-05,
+      "loss": 0.0881,
+      "step": 659
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.5319693094629157e-05,
+      "loss": 0.5762,
+      "step": 660
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.531202046035806e-05,
+      "loss": 0.0499,
+      "step": 661
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.5304347826086957e-05,
+      "loss": 0.087,
+      "step": 662
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.5296675191815855e-05,
+      "loss": 1.0719,
+      "step": 663
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.528900255754476e-05,
+      "loss": 0.6158,
+      "step": 664
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.528132992327366e-05,
+      "loss": 0.6117,
+      "step": 665
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.5273657289002557e-05,
+      "loss": 0.0515,
+      "step": 666
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.5265984654731458e-05,
+      "loss": 0.0608,
+      "step": 667
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.525831202046036e-05,
+      "loss": 0.0313,
+      "step": 668
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.5250639386189258e-05,
+      "loss": 0.1449,
+      "step": 669
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.524296675191816e-05,
+      "loss": 0.9883,
+      "step": 670
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.523529411764706e-05,
+      "loss": 0.0859,
+      "step": 671
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.522762148337596e-05,
+      "loss": 0.0561,
+      "step": 672
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.521994884910486e-05,
+      "loss": 0.0926,
+      "step": 673
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.521227621483376e-05,
+      "loss": 0.1101,
+      "step": 674
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.520460358056266e-05,
+      "loss": 0.068,
+      "step": 675
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.5196930946291563e-05,
+      "loss": 0.0773,
+      "step": 676
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.518925831202046e-05,
+      "loss": 0.0508,
+      "step": 677
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.518158567774936e-05,
+      "loss": 0.0375,
+      "step": 678
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.5173913043478264e-05,
+      "loss": 0.93,
+      "step": 679
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.5166240409207162e-05,
+      "loss": 0.1507,
+      "step": 680
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.515856777493606e-05,
+      "loss": 0.0917,
+      "step": 681
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.5150895140664962e-05,
+      "loss": 0.97,
+      "step": 682
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.5143222506393864e-05,
+      "loss": 0.5019,
+      "step": 683
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.5135549872122762e-05,
+      "loss": 0.0849,
+      "step": 684
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.5127877237851664e-05,
+      "loss": 1.2316,
+      "step": 685
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.5120204603580562e-05,
+      "loss": 0.0252,
+      "step": 686
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.5112531969309463e-05,
+      "loss": 1.0543,
+      "step": 687
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.5104859335038365e-05,
+      "loss": 0.0506,
+      "step": 688
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.5097186700767263e-05,
+      "loss": 0.0252,
+      "step": 689
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.508951406649616e-05,
+      "loss": 0.6297,
+      "step": 690
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.5081841432225066e-05,
+      "loss": 0.4497,
+      "step": 691
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.5074168797953965e-05,
+      "loss": 0.0516,
+      "step": 692
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.5066496163682863e-05,
+      "loss": 0.5501,
+      "step": 693
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.5058823529411768e-05,
+      "loss": 0.5621,
+      "step": 694
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.5051150895140666e-05,
+      "loss": 0.0577,
+      "step": 695
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.5043478260869564e-05,
+      "loss": 0.0271,
+      "step": 696
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.5035805626598466e-05,
+      "loss": 0.9496,
+      "step": 697
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.5028132992327368e-05,
+      "loss": 0.0877,
+      "step": 698
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.5020460358056266e-05,
+      "loss": 0.0523,
+      "step": 699
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.5012787723785167e-05,
+      "loss": 0.0577,
+      "step": 700
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.5005115089514066e-05,
+      "loss": 0.0522,
+      "step": 701
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.4997442455242967e-05,
+      "loss": 0.0517,
+      "step": 702
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.498976982097187e-05,
+      "loss": 0.6335,
+      "step": 703
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.4982097186700767e-05,
+      "loss": 0.05,
+      "step": 704
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.4974424552429665e-05,
+      "loss": 0.0753,
+      "step": 705
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.496675191815857e-05,
+      "loss": 0.0227,
+      "step": 706
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.495907928388747e-05,
+      "loss": 0.0755,
+      "step": 707
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.4951406649616367e-05,
+      "loss": 0.0215,
+      "step": 708
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.494373401534527e-05,
+      "loss": 0.0451,
+      "step": 709
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.493606138107417e-05,
+      "loss": 0.0343,
+      "step": 710
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.492838874680307e-05,
+      "loss": 0.0836,
+      "step": 711
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.492071611253197e-05,
+      "loss": 0.5287,
+      "step": 712
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.4913043478260868e-05,
+      "loss": 0.0686,
+      "step": 713
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.4905370843989773e-05,
+      "loss": 0.0211,
+      "step": 714
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.489769820971867e-05,
+      "loss": 0.5932,
+      "step": 715
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.489002557544757e-05,
+      "loss": 0.0371,
+      "step": 716
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.488235294117647e-05,
+      "loss": 1.0137,
+      "step": 717
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.4874680306905373e-05,
+      "loss": 0.9928,
+      "step": 718
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.486700767263427e-05,
+      "loss": 0.7431,
+      "step": 719
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.4859335038363173e-05,
+      "loss": 0.0207,
+      "step": 720
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.4851662404092074e-05,
+      "loss": 0.0681,
+      "step": 721
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.4843989769820972e-05,
+      "loss": 0.0306,
+      "step": 722
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.4836317135549874e-05,
+      "loss": 0.0204,
+      "step": 723
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.4828644501278772e-05,
+      "loss": 0.0344,
+      "step": 724
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.4820971867007674e-05,
+      "loss": 0.0345,
+      "step": 725
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.4813299232736575e-05,
+      "loss": 0.9275,
+      "step": 726
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.4805626598465474e-05,
+      "loss": 0.0196,
+      "step": 727
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.4797953964194372e-05,
+      "loss": 1.0137,
+      "step": 728
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.4790281329923277e-05,
+      "loss": 0.0318,
+      "step": 729
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.4782608695652175e-05,
+      "loss": 0.0688,
+      "step": 730
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 2.4774936061381073e-05,
+      "loss": 0.0314,
+      "step": 731
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 2.4767263427109975e-05,
+      "loss": 0.7135,
+      "step": 732
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 2.4759590792838877e-05,
+      "loss": 0.0299,
+      "step": 733
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 2.4751918158567775e-05,
+      "loss": 1.0127,
+      "step": 734
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 2.4744245524296676e-05,
+      "loss": 0.0304,
+      "step": 735
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 2.4736572890025575e-05,
+      "loss": 1.4618,
+      "step": 736
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 2.4728900255754476e-05,
+      "loss": 0.7167,
+      "step": 737
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 2.4721227621483378e-05,
+      "loss": 0.0307,
+      "step": 738
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 2.4713554987212276e-05,
+      "loss": 0.052,
+      "step": 739
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 2.4705882352941174e-05,
+      "loss": 0.778,
+      "step": 740
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 2.469820971867008e-05,
+      "loss": 0.5482,
+      "step": 741
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 2.4690537084398978e-05,
+      "loss": 1.4314,
+      "step": 742
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 2.4682864450127876e-05,
+      "loss": 0.0422,
+      "step": 743
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 2.467519181585678e-05,
+      "loss": 0.6614,
+      "step": 744
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 2.466751918158568e-05,
+      "loss": 0.0312,
+      "step": 745
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 2.4659846547314577e-05,
+      "loss": 0.043,
+      "step": 746
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 2.465217391304348e-05,
+      "loss": 0.0199,
+      "step": 747
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 2.464450127877238e-05,
+      "loss": 0.0832,
+      "step": 748
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 2.463682864450128e-05,
+      "loss": 0.0312,
+      "step": 749
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 2.462915601023018e-05,
+      "loss": 0.0309,
+      "step": 750
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 2.462148337595908e-05,
+      "loss": 0.0193,
+      "step": 751
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 2.461381074168798e-05,
+      "loss": 0.0446,
+      "step": 752
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 2.4606138107416882e-05,
+      "loss": 0.9993,
+      "step": 753
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 2.459846547314578e-05,
+      "loss": 0.016,
+      "step": 754
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 2.4590792838874678e-05,
+      "loss": 0.0433,
+      "step": 755
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 2.4583120204603583e-05,
+      "loss": 0.2614,
+      "step": 756
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 2.457544757033248e-05,
+      "loss": 0.018,
+      "step": 757
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 2.456777493606138e-05,
+      "loss": 0.6676,
+      "step": 758
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 2.456010230179028e-05,
+      "loss": 0.0599,
+      "step": 759
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 2.4552429667519183e-05,
+      "loss": 0.0338,
+      "step": 760
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 2.454475703324808e-05,
+      "loss": 0.0268,
+      "step": 761
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 2.4537084398976983e-05,
+      "loss": 0.0262,
+      "step": 762
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 2.452941176470588e-05,
+      "loss": 0.0252,
+      "step": 763
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 2.4521739130434786e-05,
+      "loss": 0.0377,
+      "step": 764
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 2.4514066496163684e-05,
+      "loss": 1.4866,
+      "step": 765
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 2.4506393861892582e-05,
+      "loss": 1.0524,
+      "step": 766
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 2.4498721227621484e-05,
+      "loss": 0.0328,
+      "step": 767
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 2.4491048593350386e-05,
+      "loss": 0.024,
+      "step": 768
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 2.4483375959079284e-05,
+      "loss": 0.022,
+      "step": 769
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 2.4475703324808185e-05,
+      "loss": 0.0245,
+      "step": 770
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 2.4468030690537087e-05,
+      "loss": 1.0107,
+      "step": 771
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 2.4460358056265985e-05,
+      "loss": 0.0336,
+      "step": 772
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 2.4452685421994887e-05,
+      "loss": 0.0391,
+      "step": 773
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 2.4445012787723785e-05,
+      "loss": 0.0303,
+      "step": 774
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2.4437340153452687e-05,
+      "loss": 0.0443,
+      "step": 775
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2.4429667519181588e-05,
+      "loss": 0.0292,
+      "step": 776
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2.4421994884910487e-05,
+      "loss": 0.0284,
+      "step": 777
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2.4414322250639385e-05,
+      "loss": 1.5985,
+      "step": 778
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 2.440664961636829e-05,
+      "loss": 0.0114,
+      "step": 779
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 2.4398976982097188e-05,
+      "loss": 0.0342,
+      "step": 780
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 2.4391304347826086e-05,
+      "loss": 1.8501,
+      "step": 781
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 2.4383631713554988e-05,
+      "loss": 0.5375,
+      "step": 782
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2.437595907928389e-05,
+      "loss": 1.589,
+      "step": 783
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2.4368286445012788e-05,
+      "loss": 0.0221,
+      "step": 784
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2.436061381074169e-05,
+      "loss": 0.0453,
+      "step": 785
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2.4352941176470587e-05,
+      "loss": 0.796,
+      "step": 786
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 2.434526854219949e-05,
+      "loss": 0.1249,
+      "step": 787
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 2.433759590792839e-05,
+      "loss": 0.7963,
+      "step": 788
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 2.432992327365729e-05,
+      "loss": 0.4231,
+      "step": 789
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 2.4322250639386187e-05,
+      "loss": 0.0247,
+      "step": 790
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4314578005115092e-05,
+      "loss": 0.0295,
+      "step": 791
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.430690537084399e-05,
+      "loss": 0.0249,
+      "step": 792
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.9065656565656566,
+      "eval_loss": 0.3537317216396332,
+      "eval_runtime": 58.1226,
+      "eval_samples_per_second": 6.813,
+      "eval_steps_per_second": 1.703,
+      "step": 792
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.429923273657289e-05,
+      "loss": 0.0157,
+      "step": 793
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4291560102301794e-05,
+      "loss": 0.5222,
+      "step": 794
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4283887468030692e-05,
+      "loss": 0.0226,
+      "step": 795
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.427621483375959e-05,
+      "loss": 0.0293,
+      "step": 796
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.426854219948849e-05,
+      "loss": 0.015,
+      "step": 797
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4260869565217393e-05,
+      "loss": 1.0658,
+      "step": 798
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.425319693094629e-05,
+      "loss": 0.0412,
+      "step": 799
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4245524296675193e-05,
+      "loss": 0.0219,
+      "step": 800
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.423785166240409e-05,
+      "loss": 0.0149,
+      "step": 801
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.4230179028132993e-05,
+      "loss": 0.0323,
+      "step": 802
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.4222506393861895e-05,
+      "loss": 1.0764,
+      "step": 803
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.4214833759590793e-05,
+      "loss": 0.7518,
+      "step": 804
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.420716112531969e-05,
+      "loss": 0.8166,
+      "step": 805
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2.4199488491048596e-05,
+      "loss": 0.0219,
+      "step": 806
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2.4191815856777494e-05,
+      "loss": 0.0154,
+      "step": 807
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2.4184143222506392e-05,
+      "loss": 0.0163,
+      "step": 808
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2.4176470588235294e-05,
+      "loss": 1.0562,
+      "step": 809
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 2.4168797953964196e-05,
+      "loss": 0.0218,
+      "step": 810
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 2.4161125319693094e-05,
+      "loss": 0.0232,
+      "step": 811
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 2.4153452685421995e-05,
+      "loss": 0.2384,
+      "step": 812
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 2.4145780051150894e-05,
+      "loss": 1.5286,
+      "step": 813
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.4138107416879795e-05,
+      "loss": 0.0262,
+      "step": 814
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.4130434782608697e-05,
+      "loss": 0.0334,
+      "step": 815
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.4122762148337595e-05,
+      "loss": 0.6509,
+      "step": 816
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.4115089514066497e-05,
+      "loss": 0.0159,
+      "step": 817
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.41074168797954e-05,
+      "loss": 0.6225,
+      "step": 818
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.4099744245524297e-05,
+      "loss": 0.752,
+      "step": 819
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.4092071611253198e-05,
+      "loss": 1.0594,
+      "step": 820
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.40843989769821e-05,
+      "loss": 0.0161,
+      "step": 821
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.4076726342710998e-05,
+      "loss": 0.7492,
+      "step": 822
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.40690537084399e-05,
+      "loss": 0.0302,
+      "step": 823
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.4061381074168798e-05,
+      "loss": 0.038,
+      "step": 824
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.40537084398977e-05,
+      "loss": 0.0243,
+      "step": 825
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.40460358056266e-05,
+      "loss": 0.0331,
+      "step": 826
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.40383631713555e-05,
+      "loss": 1.0252,
+      "step": 827
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.4030690537084398e-05,
+      "loss": 0.0172,
+      "step": 828
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.4023017902813303e-05,
+      "loss": 0.6168,
+      "step": 829
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.40153452685422e-05,
+      "loss": 0.3135,
+      "step": 830
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.40076726342711e-05,
+      "loss": 0.0168,
+      "step": 831
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.4e-05,
+      "loss": 0.0172,
+      "step": 832
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.3992327365728902e-05,
+      "loss": 1.0349,
+      "step": 833
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.39846547314578e-05,
+      "loss": 0.0978,
+      "step": 834
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.3976982097186702e-05,
+      "loss": 0.0167,
+      "step": 835
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.39693094629156e-05,
+      "loss": 1.0381,
+      "step": 836
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.3961636828644502e-05,
+      "loss": 0.6857,
+      "step": 837
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.3953964194373404e-05,
+      "loss": 1.0406,
+      "step": 838
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.3946291560102302e-05,
+      "loss": 0.071,
+      "step": 839
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.39386189258312e-05,
+      "loss": 0.0193,
+      "step": 840
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.3930946291560105e-05,
+      "loss": 0.0193,
+      "step": 841
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.3923273657289003e-05,
+      "loss": 0.9916,
+      "step": 842
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.39156010230179e-05,
+      "loss": 0.0723,
+      "step": 843
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.3907928388746806e-05,
+      "loss": 0.0626,
+      "step": 844
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.3900255754475705e-05,
+      "loss": 0.7264,
+      "step": 845
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.3892583120204603e-05,
+      "loss": 0.827,
+      "step": 846
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.3884910485933504e-05,
+      "loss": 1.2193,
+      "step": 847
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.3877237851662406e-05,
+      "loss": 0.6754,
+      "step": 848
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.3869565217391304e-05,
+      "loss": 0.6856,
+      "step": 849
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.3861892583120206e-05,
+      "loss": 0.7441,
+      "step": 850
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.3854219948849104e-05,
+      "loss": 0.023,
+      "step": 851
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.3846547314578006e-05,
+      "loss": 0.4742,
+      "step": 852
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.3838874680306907e-05,
+      "loss": 0.0363,
+      "step": 853
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.3831202046035806e-05,
+      "loss": 0.0409,
+      "step": 854
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.3823529411764704e-05,
+      "loss": 0.8353,
+      "step": 855
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.381585677749361e-05,
+      "loss": 0.0333,
+      "step": 856
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.3808184143222507e-05,
+      "loss": 0.0216,
+      "step": 857
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.3800511508951405e-05,
+      "loss": 0.9688,
+      "step": 858
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.3792838874680307e-05,
+      "loss": 1.0078,
+      "step": 859
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.378516624040921e-05,
+      "loss": 0.0183,
+      "step": 860
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.3777493606138107e-05,
+      "loss": 0.0183,
+      "step": 861
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.3769820971867008e-05,
+      "loss": 0.9786,
+      "step": 862
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.3762148337595907e-05,
+      "loss": 0.9166,
+      "step": 863
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.3754475703324808e-05,
+      "loss": 1.1408,
+      "step": 864
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.374680306905371e-05,
+      "loss": 1.1394,
+      "step": 865
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.3739130434782608e-05,
+      "loss": 0.0185,
+      "step": 866
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.373145780051151e-05,
+      "loss": 0.9979,
+      "step": 867
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.372378516624041e-05,
+      "loss": 1.0092,
+      "step": 868
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.371611253196931e-05,
+      "loss": 0.7276,
+      "step": 869
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.370843989769821e-05,
+      "loss": 0.0359,
+      "step": 870
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.3700767263427113e-05,
+      "loss": 0.0864,
+      "step": 871
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.369309462915601e-05,
+      "loss": 0.0246,
+      "step": 872
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.3685421994884912e-05,
+      "loss": 0.8518,
+      "step": 873
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.367774936061381e-05,
+      "loss": 1.433,
+      "step": 874
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.3670076726342712e-05,
+      "loss": 0.035,
+      "step": 875
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.3662404092071614e-05,
+      "loss": 0.5922,
+      "step": 876
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.3654731457800512e-05,
+      "loss": 0.529,
+      "step": 877
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.364705882352941e-05,
+      "loss": 0.6423,
+      "step": 878
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.3639386189258315e-05,
+      "loss": 0.0337,
+      "step": 879
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.3631713554987214e-05,
+      "loss": 0.0402,
+      "step": 880
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.3624040920716112e-05,
+      "loss": 0.0782,
+      "step": 881
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.3616368286445013e-05,
+      "loss": 0.4141,
+      "step": 882
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.3608695652173915e-05,
+      "loss": 0.0359,
+      "step": 883
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.3601023017902813e-05,
+      "loss": 0.4509,
+      "step": 884
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.3593350383631715e-05,
+      "loss": 0.6414,
+      "step": 885
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.3585677749360613e-05,
+      "loss": 0.6873,
+      "step": 886
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.3578005115089515e-05,
+      "loss": 0.0304,
+      "step": 887
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.3570332480818416e-05,
+      "loss": 0.0315,
+      "step": 888
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.3562659846547315e-05,
+      "loss": 0.0896,
+      "step": 889
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.3554987212276213e-05,
+      "loss": 0.0201,
+      "step": 890
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.3547314578005118e-05,
+      "loss": 0.027,
+      "step": 891
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.3539641943734016e-05,
+      "loss": 1.9853,
+      "step": 892
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.3531969309462914e-05,
+      "loss": 2.9647,
+      "step": 893
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.352429667519182e-05,
+      "loss": 1.9403,
+      "step": 894
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.3516624040920717e-05,
+      "loss": 0.0193,
+      "step": 895
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.3508951406649616e-05,
+      "loss": 0.9912,
+      "step": 896
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.3501278772378517e-05,
+      "loss": 0.02,
+      "step": 897
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.349360613810742e-05,
+      "loss": 0.9015,
+      "step": 898
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.3485933503836317e-05,
+      "loss": 0.1294,
+      "step": 899
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.347826086956522e-05,
+      "loss": 0.0559,
+      "step": 900
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.3470588235294117e-05,
+      "loss": 1.8039,
+      "step": 901
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.346291560102302e-05,
+      "loss": 0.0336,
+      "step": 902
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.345524296675192e-05,
+      "loss": 0.0383,
+      "step": 903
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.344757033248082e-05,
+      "loss": 0.0722,
+      "step": 904
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.3439897698209717e-05,
+      "loss": 0.0278,
+      "step": 905
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.343222506393862e-05,
+      "loss": 0.0684,
+      "step": 906
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.342455242966752e-05,
+      "loss": 0.0236,
+      "step": 907
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.3416879795396418e-05,
+      "loss": 0.0348,
+      "step": 908
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.340920716112532e-05,
+      "loss": 0.0204,
+      "step": 909
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.340153452685422e-05,
+      "loss": 0.8109,
+      "step": 910
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.339386189258312e-05,
+      "loss": 0.0279,
+      "step": 911
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.338618925831202e-05,
+      "loss": 0.0361,
+      "step": 912
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.337851662404092e-05,
+      "loss": 0.0355,
+      "step": 913
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.337084398976982e-05,
+      "loss": 0.7958,
+      "step": 914
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.3363171355498723e-05,
+      "loss": 0.0362,
+      "step": 915
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.335549872122762e-05,
+      "loss": 0.0252,
+      "step": 916
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.3347826086956522e-05,
+      "loss": 0.8165,
+      "step": 917
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.3340153452685424e-05,
+      "loss": 0.0237,
+      "step": 918
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.3332480818414322e-05,
+      "loss": 0.8162,
+      "step": 919
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.332480818414322e-05,
+      "loss": 0.7786,
+      "step": 920
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.3317135549872125e-05,
+      "loss": 0.0787,
+      "step": 921
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.3309462915601024e-05,
+      "loss": 0.1764,
+      "step": 922
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.3301790281329925e-05,
+      "loss": 0.7449,
+      "step": 923
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.3294117647058824e-05,
+      "loss": 0.0452,
+      "step": 924
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.3286445012787725e-05,
+      "loss": 0.0181,
+      "step": 925
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.3278772378516627e-05,
+      "loss": 0.1783,
+      "step": 926
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.3271099744245525e-05,
+      "loss": 0.5158,
+      "step": 927
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.3263427109974423e-05,
+      "loss": 0.7574,
+      "step": 928
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.3255754475703328e-05,
+      "loss": 2.6099,
+      "step": 929
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.3248081841432226e-05,
+      "loss": 0.0304,
+      "step": 930
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.3240409207161125e-05,
+      "loss": 0.0795,
+      "step": 931
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.3232736572890026e-05,
+      "loss": 0.0342,
+      "step": 932
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.3225063938618928e-05,
+      "loss": 0.0473,
+      "step": 933
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.3217391304347826e-05,
+      "loss": 0.0831,
+      "step": 934
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.3209718670076728e-05,
+      "loss": 0.0465,
+      "step": 935
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.3202046035805626e-05,
+      "loss": 0.6769,
+      "step": 936
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.3194373401534528e-05,
+      "loss": 0.0173,
+      "step": 937
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.318670076726343e-05,
+      "loss": 0.0169,
+      "step": 938
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.3179028132992327e-05,
+      "loss": 0.0469,
+      "step": 939
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.3171355498721226e-05,
+      "loss": 0.048,
+      "step": 940
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.316368286445013e-05,
+      "loss": 1.0704,
+      "step": 941
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.315601023017903e-05,
+      "loss": 1.0738,
+      "step": 942
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.3148337595907927e-05,
+      "loss": 0.0163,
+      "step": 943
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.3140664961636832e-05,
+      "loss": 2.0702,
+      "step": 944
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.313299232736573e-05,
+      "loss": 1.0433,
+      "step": 945
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.312531969309463e-05,
+      "loss": 1.0351,
+      "step": 946
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.311764705882353e-05,
+      "loss": 1.053,
+      "step": 947
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.3109974424552432e-05,
+      "loss": 0.1797,
+      "step": 948
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.310230179028133e-05,
+      "loss": 1.0259,
+      "step": 949
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.309462915601023e-05,
+      "loss": 0.03,
+      "step": 950
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.308695652173913e-05,
+      "loss": 0.905,
+      "step": 951
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.307928388746803e-05,
+      "loss": 0.0699,
+      "step": 952
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.3071611253196933e-05,
+      "loss": 1.0013,
+      "step": 953
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.306393861892583e-05,
+      "loss": 0.0319,
+      "step": 954
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.305626598465473e-05,
+      "loss": 0.0361,
+      "step": 955
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.3048593350383634e-05,
+      "loss": 0.0201,
+      "step": 956
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.3040920716112533e-05,
+      "loss": 0.033,
+      "step": 957
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.303324808184143e-05,
+      "loss": 0.0411,
+      "step": 958
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.3025575447570332e-05,
+      "loss": 0.7138,
+      "step": 959
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.3017902813299234e-05,
+      "loss": 0.0434,
+      "step": 960
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.3010230179028132e-05,
+      "loss": 0.8445,
+      "step": 961
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.3002557544757034e-05,
+      "loss": 0.0323,
+      "step": 962
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.2994884910485932e-05,
+      "loss": 0.0203,
+      "step": 963
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.2987212276214834e-05,
+      "loss": 0.2302,
+      "step": 964
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 2.2979539641943735e-05,
+      "loss": 0.0293,
+      "step": 965
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 2.2971867007672634e-05,
+      "loss": 0.0593,
+      "step": 966
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 2.2964194373401535e-05,
+      "loss": 0.6505,
+      "step": 967
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 2.2956521739130437e-05,
+      "loss": 0.5891,
+      "step": 968
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 2.2948849104859335e-05,
+      "loss": 0.0191,
+      "step": 969
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 2.2941176470588233e-05,
+      "loss": 0.0188,
+      "step": 970
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 2.2933503836317138e-05,
+      "loss": 0.9537,
+      "step": 971
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 2.2925831202046037e-05,
+      "loss": 0.0186,
+      "step": 972
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 2.2918158567774935e-05,
+      "loss": 1.0237,
+      "step": 973
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 2.2910485933503836e-05,
+      "loss": 0.0417,
+      "step": 974
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 2.2902813299232738e-05,
+      "loss": 0.0369,
+      "step": 975
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 2.289514066496164e-05,
+      "loss": 0.0184,
+      "step": 976
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 2.2887468030690538e-05,
+      "loss": 0.811,
+      "step": 977
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 2.2879795396419436e-05,
+      "loss": 0.0183,
+      "step": 978
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 2.287212276214834e-05,
+      "loss": 1.0442,
+      "step": 979
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 2.286445012787724e-05,
+      "loss": 0.027,
+      "step": 980
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 2.2856777493606137e-05,
+      "loss": 0.062,
+      "step": 981
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 2.284910485933504e-05,
+      "loss": 0.0186,
+      "step": 982
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 2.284143222506394e-05,
+      "loss": 0.6057,
+      "step": 983
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 2.283375959079284e-05,
+      "loss": 1.6594,
+      "step": 984
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 2.282608695652174e-05,
+      "loss": 0.0192,
+      "step": 985
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 2.281841432225064e-05,
+      "loss": 1.5361,
+      "step": 986
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 2.281074168797954e-05,
+      "loss": 0.0384,
+      "step": 987
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 2.2803069053708442e-05,
+      "loss": 2.0238,
+      "step": 988
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 2.279539641943734e-05,
+      "loss": 0.0334,
+      "step": 989
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 2.2787723785166242e-05,
+      "loss": 0.0571,
+      "step": 990
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 2.2780051150895143e-05,
+      "loss": 0.0211,
+      "step": 991
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 2.277237851662404e-05,
+      "loss": 0.0409,
+      "step": 992
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 2.276470588235294e-05,
+      "loss": 0.1059,
+      "step": 993
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 2.2757033248081845e-05,
+      "loss": 0.0277,
+      "step": 994
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 2.2749360613810743e-05,
+      "loss": 1.0023,
+      "step": 995
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 2.274168797953964e-05,
+      "loss": 0.0186,
+      "step": 996
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 2.2734015345268543e-05,
+      "loss": 0.0345,
+      "step": 997
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 2.2726342710997445e-05,
+      "loss": 0.0169,
+      "step": 998
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 2.2718670076726343e-05,
+      "loss": 0.7292,
+      "step": 999
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 2.2710997442455244e-05,
+      "loss": 1.2655,
+      "step": 1000
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 2.2703324808184143e-05,
+      "loss": 0.03,
+      "step": 1001
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 2.2695652173913044e-05,
+      "loss": 1.0266,
+      "step": 1002
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 2.2687979539641946e-05,
+      "loss": 0.7645,
+      "step": 1003
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 2.2680306905370844e-05,
+      "loss": 0.7985,
+      "step": 1004
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 2.2672634271099742e-05,
+      "loss": 0.76,
+      "step": 1005
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 2.2664961636828647e-05,
+      "loss": 0.7577,
+      "step": 1006
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 2.2657289002557545e-05,
+      "loss": 0.0248,
+      "step": 1007
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 2.2649616368286444e-05,
+      "loss": 0.0262,
+      "step": 1008
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 2.2641943734015345e-05,
+      "loss": 1.9984,
+      "step": 1009
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 2.2634271099744247e-05,
+      "loss": 0.1489,
+      "step": 1010
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 2.2626598465473145e-05,
+      "loss": 0.0683,
+      "step": 1011
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2.2618925831202047e-05,
+      "loss": 0.054,
+      "step": 1012
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2.2611253196930945e-05,
+      "loss": 0.0269,
+      "step": 1013
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2.2603580562659847e-05,
+      "loss": 1.0378,
+      "step": 1014
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2.2595907928388748e-05,
+      "loss": 0.0171,
+      "step": 1015
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2.2588235294117646e-05,
+      "loss": 0.8433,
+      "step": 1016
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2.2580562659846548e-05,
+      "loss": 0.0468,
+      "step": 1017
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2.257289002557545e-05,
+      "loss": 0.0331,
+      "step": 1018
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2.2565217391304348e-05,
+      "loss": 0.0308,
+      "step": 1019
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.2557544757033246e-05,
+      "loss": 1.0007,
+      "step": 1020
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.254987212276215e-05,
+      "loss": 2.0242,
+      "step": 1021
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.254219948849105e-05,
+      "loss": 0.7713,
+      "step": 1022
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.2534526854219948e-05,
+      "loss": 1.1112,
+      "step": 1023
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2.252685421994885e-05,
+      "loss": 0.0184,
+      "step": 1024
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2.251918158567775e-05,
+      "loss": 0.0243,
+      "step": 1025
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2.251150895140665e-05,
+      "loss": 0.046,
+      "step": 1026
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2.250383631713555e-05,
+      "loss": 0.0378,
+      "step": 1027
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.249616368286445e-05,
+      "loss": 0.0329,
+      "step": 1028
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.2488491048593354e-05,
+      "loss": 0.0245,
+      "step": 1029
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.2480818414322252e-05,
+      "loss": 0.0238,
+      "step": 1030
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.247314578005115e-05,
+      "loss": 1.0628,
+      "step": 1031
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.2465473145780052e-05,
+      "loss": 0.0703,
+      "step": 1032
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.2457800511508954e-05,
+      "loss": 0.0251,
+      "step": 1033
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.2450127877237852e-05,
+      "loss": 1.0561,
+      "step": 1034
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.2442455242966753e-05,
+      "loss": 0.0274,
+      "step": 1035
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.243478260869565e-05,
+      "loss": 0.023,
+      "step": 1036
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.2427109974424553e-05,
+      "loss": 0.0223,
+      "step": 1037
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.2419437340153455e-05,
+      "loss": 0.8113,
+      "step": 1038
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.2411764705882353e-05,
+      "loss": 0.0346,
+      "step": 1039
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.2404092071611255e-05,
+      "loss": 0.0164,
+      "step": 1040
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.2396419437340156e-05,
+      "loss": 0.0237,
+      "step": 1041
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.2388746803069054e-05,
+      "loss": 0.0305,
+      "step": 1042
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.2381074168797953e-05,
+      "loss": 0.0158,
+      "step": 1043
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.2373401534526858e-05,
+      "loss": 0.0225,
+      "step": 1044
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.2365728900255756e-05,
+      "loss": 0.0226,
+      "step": 1045
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.2358056265984654e-05,
+      "loss": 0.015,
+      "step": 1046
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.2350383631713556e-05,
+      "loss": 0.0219,
+      "step": 1047
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.2342710997442457e-05,
+      "loss": 1.0694,
+      "step": 1048
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.2335038363171356e-05,
+      "loss": 0.0348,
+      "step": 1049
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.2327365728900257e-05,
+      "loss": 0.0213,
+      "step": 1050
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.2319693094629155e-05,
+      "loss": 0.9405,
+      "step": 1051
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.2312020460358057e-05,
+      "loss": 0.0252,
+      "step": 1052
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.230434782608696e-05,
+      "loss": 0.0254,
+      "step": 1053
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.2296675191815857e-05,
+      "loss": 0.7353,
+      "step": 1054
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.2289002557544755e-05,
+      "loss": 0.0253,
+      "step": 1055
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.228132992327366e-05,
+      "loss": 0.8361,
+      "step": 1056
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.2273657289002558e-05,
+      "loss": 0.0196,
+      "step": 1057
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.2265984654731457e-05,
+      "loss": 0.014,
+      "step": 1058
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.2258312020460358e-05,
+      "loss": 0.8297,
+      "step": 1059
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.225063938618926e-05,
+      "loss": 1.0654,
+      "step": 1060
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.2242966751918158e-05,
+      "loss": 0.0248,
+      "step": 1061
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.223529411764706e-05,
+      "loss": 0.0226,
+      "step": 1062
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.2227621483375958e-05,
+      "loss": 0.0238,
+      "step": 1063
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.221994884910486e-05,
+      "loss": 0.0212,
+      "step": 1064
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.221227621483376e-05,
+      "loss": 0.0274,
+      "step": 1065
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.220460358056266e-05,
+      "loss": 1.0217,
+      "step": 1066
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.219693094629156e-05,
+      "loss": 0.0252,
+      "step": 1067
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2.2189258312020462e-05,
+      "loss": 0.0219,
+      "step": 1068
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2.218158567774936e-05,
+      "loss": 0.0349,
+      "step": 1069
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2.217391304347826e-05,
+      "loss": 0.0152,
+      "step": 1070
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2.2166240409207164e-05,
+      "loss": 0.021,
+      "step": 1071
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 2.2158567774936062e-05,
+      "loss": 0.015,
+      "step": 1072
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 2.215089514066496e-05,
+      "loss": 0.0277,
+      "step": 1073
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 2.2143222506393862e-05,
+      "loss": 0.1858,
+      "step": 1074
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 2.2135549872122764e-05,
+      "loss": 0.0136,
+      "step": 1075
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 2.2127877237851662e-05,
+      "loss": 0.0132,
+      "step": 1076
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 2.2120204603580563e-05,
+      "loss": 0.0126,
+      "step": 1077
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 2.211253196930946e-05,
+      "loss": 0.5171,
+      "step": 1078
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 2.2104859335038363e-05,
+      "loss": 0.0422,
+      "step": 1079
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 2.2097186700767265e-05,
+      "loss": 0.0186,
+      "step": 1080
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 2.2089514066496163e-05,
+      "loss": 0.8412,
+      "step": 1081
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 2.2081841432225065e-05,
+      "loss": 0.0183,
+      "step": 1082
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 2.2074168797953966e-05,
+      "loss": 1.0243,
+      "step": 1083
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 2.2066496163682865e-05,
+      "loss": 0.0212,
+      "step": 1084
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 2.2058823529411766e-05,
+      "loss": 0.2841,
+      "step": 1085
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 2.2051150895140664e-05,
+      "loss": 0.0118,
+      "step": 1086
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 2.2043478260869566e-05,
+      "loss": 0.0247,
+      "step": 1087
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 2.2035805626598468e-05,
+      "loss": 1.5121,
+      "step": 1088
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 2.2028132992327366e-05,
+      "loss": 0.0304,
+      "step": 1089
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 2.2020460358056267e-05,
+      "loss": 0.0329,
+      "step": 1090
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 2.201278772378517e-05,
+      "loss": 1.1288,
+      "step": 1091
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 2.2005115089514067e-05,
+      "loss": 0.0145,
+      "step": 1092
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 2.1997442455242966e-05,
+      "loss": 0.0498,
+      "step": 1093
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 2.198976982097187e-05,
+      "loss": 0.0244,
+      "step": 1094
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.198209718670077e-05,
+      "loss": 0.0112,
+      "step": 1095
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.1974424552429667e-05,
+      "loss": 0.8642,
+      "step": 1096
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.196675191815857e-05,
+      "loss": 0.8822,
+      "step": 1097
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.195907928388747e-05,
+      "loss": 0.0278,
+      "step": 1098
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.195140664961637e-05,
+      "loss": 0.6954,
+      "step": 1099
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.194373401534527e-05,
+      "loss": 0.0179,
+      "step": 1100
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.1936061381074168e-05,
+      "loss": 0.5607,
+      "step": 1101
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.192838874680307e-05,
+      "loss": 0.0191,
+      "step": 1102
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.192071611253197e-05,
+      "loss": 1.9519,
+      "step": 1103
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.191304347826087e-05,
+      "loss": 0.8347,
+      "step": 1104
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.1905370843989768e-05,
+      "loss": 0.0262,
+      "step": 1105
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.1897698209718673e-05,
+      "loss": 0.0477,
+      "step": 1106
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.189002557544757e-05,
+      "loss": 0.0227,
+      "step": 1107
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.188235294117647e-05,
+      "loss": 0.0205,
+      "step": 1108
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.187468030690537e-05,
+      "loss": 0.0208,
+      "step": 1109
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.1867007672634273e-05,
+      "loss": 0.0116,
+      "step": 1110
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.185933503836317e-05,
+      "loss": 1.1546,
+      "step": 1111
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.1851662404092072e-05,
+      "loss": 0.0114,
+      "step": 1112
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.184398976982097e-05,
+      "loss": 0.0113,
+      "step": 1113
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.1836317135549872e-05,
+      "loss": 0.6881,
+      "step": 1114
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.1828644501278774e-05,
+      "loss": 0.0294,
+      "step": 1115
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.1820971867007672e-05,
+      "loss": 0.1569,
+      "step": 1116
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.1813299232736574e-05,
+      "loss": 1.1362,
+      "step": 1117
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.1805626598465475e-05,
+      "loss": 0.0196,
+      "step": 1118
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.1797953964194374e-05,
+      "loss": 0.0205,
+      "step": 1119
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.1790281329923272e-05,
+      "loss": 0.0198,
+      "step": 1120
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.1782608695652177e-05,
+      "loss": 0.0113,
+      "step": 1121
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.1774936061381075e-05,
+      "loss": 0.02,
+      "step": 1122
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.1767263427109973e-05,
+      "loss": 0.0195,
+      "step": 1123
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.1759590792838875e-05,
+      "loss": 0.8016,
+      "step": 1124
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.1751918158567776e-05,
+      "loss": 0.0122,
+      "step": 1125
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.1744245524296675e-05,
+      "loss": 1.1132,
+      "step": 1126
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.1736572890025576e-05,
+      "loss": 0.011,
+      "step": 1127
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.1728900255754474e-05,
+      "loss": 0.7705,
+      "step": 1128
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.1721227621483376e-05,
+      "loss": 1.1467,
+      "step": 1129
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.1713554987212278e-05,
+      "loss": 0.0176,
+      "step": 1130
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.1705882352941176e-05,
+      "loss": 0.0293,
+      "step": 1131
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.1698209718670074e-05,
+      "loss": 0.0208,
+      "step": 1132
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.169053708439898e-05,
+      "loss": 0.0199,
+      "step": 1133
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.1682864450127877e-05,
+      "loss": 0.0431,
+      "step": 1134
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2.167519181585678e-05,
+      "loss": 1.6066,
+      "step": 1135
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2.1667519181585677e-05,
+      "loss": 0.029,
+      "step": 1136
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2.165984654731458e-05,
+      "loss": 0.029,
+      "step": 1137
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2.165217391304348e-05,
+      "loss": 1.1749,
+      "step": 1138
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 2.164450127877238e-05,
+      "loss": 0.0189,
+      "step": 1139
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 2.163682864450128e-05,
+      "loss": 0.0309,
+      "step": 1140
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 2.1629156010230182e-05,
+      "loss": 0.0287,
+      "step": 1141
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 2.162148337595908e-05,
+      "loss": 0.0202,
+      "step": 1142
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 2.161381074168798e-05,
+      "loss": 0.7889,
+      "step": 1143
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 2.1606138107416883e-05,
+      "loss": 0.0285,
+      "step": 1144
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 2.159846547314578e-05,
+      "loss": 0.0535,
+      "step": 1145
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 2.159079283887468e-05,
+      "loss": 1.0829,
+      "step": 1146
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 2.158312020460358e-05,
+      "loss": 0.0989,
+      "step": 1147
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 2.1575447570332483e-05,
+      "loss": 0.0115,
+      "step": 1148
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 2.156777493606138e-05,
+      "loss": 0.0227,
+      "step": 1149
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 2.1560102301790283e-05,
+      "loss": 0.5506,
+      "step": 1150
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 2.155242966751918e-05,
+      "loss": 1.1234,
+      "step": 1151
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 2.1544757033248083e-05,
+      "loss": 0.0564,
+      "step": 1152
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 2.1537084398976984e-05,
+      "loss": 0.0127,
+      "step": 1153
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 2.1529411764705882e-05,
+      "loss": 0.0218,
+      "step": 1154
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 2.152173913043478e-05,
+      "loss": 0.0124,
+      "step": 1155
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 2.1514066496163686e-05,
+      "loss": 0.0223,
+      "step": 1156
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 2.1506393861892584e-05,
+      "loss": 0.0505,
+      "step": 1157
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 2.1498721227621482e-05,
+      "loss": 0.0225,
+      "step": 1158
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.1491048593350384e-05,
+      "loss": 0.0308,
+      "step": 1159
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.1483375959079285e-05,
+      "loss": 0.036,
+      "step": 1160
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.1475703324808184e-05,
+      "loss": 0.0121,
+      "step": 1161
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.1468030690537085e-05,
+      "loss": 0.0274,
+      "step": 1162
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.1460358056265983e-05,
+      "loss": 1.0397,
+      "step": 1163
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.1452685421994885e-05,
+      "loss": 0.0332,
+      "step": 1164
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.1445012787723787e-05,
+      "loss": 0.0182,
+      "step": 1165
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.1437340153452685e-05,
+      "loss": 0.018,
+      "step": 1166
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 2.1429667519181587e-05,
+      "loss": 0.859,
+      "step": 1167
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 2.1421994884910488e-05,
+      "loss": 0.0321,
+      "step": 1168
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 2.1414322250639386e-05,
+      "loss": 0.0225,
+      "step": 1169
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 2.1406649616368285e-05,
+      "loss": 0.0224,
+      "step": 1170
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 2.139897698209719e-05,
+      "loss": 0.03,
+      "step": 1171
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 2.1391304347826088e-05,
+      "loss": 0.7473,
+      "step": 1172
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 2.1383631713554986e-05,
+      "loss": 0.8609,
+      "step": 1173
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 2.1375959079283888e-05,
+      "loss": 0.0105,
+      "step": 1174
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.136828644501279e-05,
+      "loss": 0.0217,
+      "step": 1175
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.1360613810741687e-05,
+      "loss": 0.0218,
+      "step": 1176
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.135294117647059e-05,
+      "loss": 0.016,
+      "step": 1177
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.1345268542199487e-05,
+      "loss": 0.7844,
+      "step": 1178
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.133759590792839e-05,
+      "loss": 0.01,
+      "step": 1179
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.132992327365729e-05,
+      "loss": 0.8774,
+      "step": 1180
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.132225063938619e-05,
+      "loss": 0.8375,
+      "step": 1181
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.1314578005115087e-05,
+      "loss": 0.0188,
+      "step": 1182
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 2.1306905370843992e-05,
+      "loss": 1.1026,
+      "step": 1183
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 2.129923273657289e-05,
+      "loss": 1.0447,
+      "step": 1184
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 2.129156010230179e-05,
+      "loss": 0.0101,
+      "step": 1185
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 2.128388746803069e-05,
+      "loss": 0.0247,
+      "step": 1186
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 2.127621483375959e-05,
+      "loss": 0.0295,
+      "step": 1187
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 2.1268542199488493e-05,
+      "loss": 0.0191,
+      "step": 1188
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.9065656565656566,
+      "eval_loss": 0.37390685081481934,
+      "eval_runtime": 57.4087,
+      "eval_samples_per_second": 6.898,
+      "eval_steps_per_second": 1.724,
+      "step": 1188
+    }
+  ],
+  "max_steps": 3960,
+  "num_train_epochs": 10,
+  "total_flos": 4.311442640736e+17,
+  "trial_name": null,
+  "trial_params": null
+}