diff --git "a/checkpoint-1000/trainer_state.json" "b/checkpoint-1000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1000/trainer_state.json"
@@ -0,0 +1,6016 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.6534879921581441,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4492753623188408e-07,
+      "loss": 2.5958,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8985507246376816e-07,
+      "loss": 2.4792,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.347826086956522e-07,
+      "loss": 1.9376,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.797101449275363e-07,
+      "loss": 2.0731,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.246376811594204e-07,
+      "loss": 1.7538,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.695652173913044e-07,
+      "loss": 1.8296,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.0144927536231885e-06,
+      "loss": 1.9283,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1594202898550726e-06,
+      "loss": 1.8517,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3043478260869566e-06,
+      "loss": 1.7672,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4492753623188408e-06,
+      "loss": 1.7437,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5942028985507246e-06,
+      "loss": 1.7038,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7391304347826088e-06,
+      "loss": 2.0155,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.884057971014493e-06,
+      "loss": 2.0078,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.028985507246377e-06,
+      "loss": 1.9901,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.173913043478261e-06,
+      "loss": 1.7954,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3188405797101453e-06,
+      "loss": 1.7183,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4637681159420295e-06,
+      "loss": 1.842,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6086956521739132e-06,
+      "loss": 1.7957,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7536231884057974e-06,
+      "loss": 1.7775,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8985507246376816e-06,
+      "loss": 1.8612,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.043478260869566e-06,
+      "loss": 1.68,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.188405797101449e-06,
+      "loss": 1.928,
+      "step": 22
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.7262,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4782608695652175e-06,
+      "loss": 1.9312,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6231884057971017e-06,
+      "loss": 1.9374,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.768115942028986e-06,
+      "loss": 1.5449,
+      "step": 26
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.91304347826087e-06,
+      "loss": 1.7894,
+      "step": 27
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.057971014492754e-06,
+      "loss": 2.0061,
+      "step": 28
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.202898550724638e-06,
+      "loss": 1.775,
+      "step": 29
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.347826086956522e-06,
+      "loss": 1.856,
+      "step": 30
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.492753623188406e-06,
+      "loss": 1.7222,
+      "step": 31
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637681159420291e-06,
+      "loss": 2.0042,
+      "step": 32
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.782608695652174e-06,
+      "loss": 1.7216,
+      "step": 33
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.927536231884059e-06,
+      "loss": 1.6443,
+      "step": 34
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.072463768115943e-06,
+      "loss": 1.6971,
+      "step": 35
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.2173913043478265e-06,
+      "loss": 1.7971,
+      "step": 36
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.362318840579711e-06,
+      "loss": 1.771,
+      "step": 37
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.507246376811595e-06,
+      "loss": 1.7365,
+      "step": 38
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.652173913043479e-06,
+      "loss": 1.9443,
+      "step": 39
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.797101449275363e-06,
+      "loss": 1.8931,
+      "step": 40
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.942028985507247e-06,
+      "loss": 1.7748,
+      "step": 41
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.086956521739132e-06,
+      "loss": 1.8616,
+      "step": 42
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.2318840579710145e-06,
+      "loss": 1.8224,
+      "step": 43
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.376811594202898e-06,
+      "loss": 1.5656,
+      "step": 44
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.521739130434783e-06,
+      "loss": 1.6967,
+      "step": 45
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.7999,
+      "step": 46
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.811594202898551e-06,
+      "loss": 1.7706,
+      "step": 47
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.956521739130435e-06,
+      "loss": 1.732,
+      "step": 48
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 7.10144927536232e-06,
+      "loss": 1.9167,
+      "step": 49
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 7.246376811594203e-06,
+      "loss": 1.7641,
+      "step": 50
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 7.391304347826087e-06,
+      "loss": 1.5833,
+      "step": 51
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 7.536231884057972e-06,
+      "loss": 1.7085,
+      "step": 52
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 7.681159420289856e-06,
+      "loss": 1.7761,
+      "step": 53
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.82608695652174e-06,
+      "loss": 1.7624,
+      "step": 54
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.971014492753623e-06,
+      "loss": 1.7813,
+      "step": 55
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.115942028985508e-06,
+      "loss": 1.6884,
+      "step": 56
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.260869565217392e-06,
+      "loss": 1.7478,
+      "step": 57
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.405797101449275e-06,
+      "loss": 1.6749,
+      "step": 58
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.55072463768116e-06,
+      "loss": 1.5301,
+      "step": 59
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.695652173913044e-06,
+      "loss": 1.6166,
+      "step": 60
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.840579710144929e-06,
+      "loss": 1.5818,
+      "step": 61
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.985507246376812e-06,
+      "loss": 1.7006,
+      "step": 62
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.130434782608697e-06,
+      "loss": 1.7545,
+      "step": 63
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.275362318840581e-06,
+      "loss": 1.7942,
+      "step": 64
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.420289855072464e-06,
+      "loss": 1.8001,
+      "step": 65
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.565217391304349e-06,
+      "loss": 1.8964,
+      "step": 66
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.710144927536233e-06,
+      "loss": 1.8787,
+      "step": 67
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.855072463768118e-06,
+      "loss": 1.6214,
+      "step": 68
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1e-05,
+      "loss": 1.7235,
+      "step": 69
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0144927536231885e-05,
+      "loss": 1.6727,
+      "step": 70
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0289855072463768e-05,
+      "loss": 1.5962,
+      "step": 71
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0434782608695653e-05,
+      "loss": 1.5443,
+      "step": 72
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0579710144927538e-05,
+      "loss": 1.8079,
+      "step": 73
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0724637681159422e-05,
+      "loss": 1.7635,
+      "step": 74
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0869565217391305e-05,
+      "loss": 1.6006,
+      "step": 75
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.101449275362319e-05,
+      "loss": 1.8047,
+      "step": 76
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.1159420289855074e-05,
+      "loss": 1.7687,
+      "step": 77
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.1304347826086957e-05,
+      "loss": 1.6998,
+      "step": 78
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.1449275362318842e-05,
+      "loss": 1.717,
+      "step": 79
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.1594202898550726e-05,
+      "loss": 1.6178,
+      "step": 80
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.1739130434782611e-05,
+      "loss": 1.9585,
+      "step": 81
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.1884057971014494e-05,
+      "loss": 1.6892,
+      "step": 82
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.2028985507246379e-05,
+      "loss": 1.6169,
+      "step": 83
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.2173913043478263e-05,
+      "loss": 2.0966,
+      "step": 84
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2318840579710144e-05,
+      "loss": 2.0971,
+      "step": 85
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2463768115942029e-05,
+      "loss": 1.6939,
+      "step": 86
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2608695652173915e-05,
+      "loss": 1.7684,
+      "step": 87
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2753623188405797e-05,
+      "loss": 1.5717,
+      "step": 88
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2898550724637681e-05,
+      "loss": 1.7117,
+      "step": 89
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.3043478260869566e-05,
+      "loss": 1.9092,
+      "step": 90
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.318840579710145e-05,
+      "loss": 1.9193,
+      "step": 91
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 1.8541,
+      "step": 92
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.3478260869565218e-05,
+      "loss": 2.1758,
+      "step": 93
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.3623188405797103e-05,
+      "loss": 1.897,
+      "step": 94
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.3768115942028985e-05,
+      "loss": 1.6832,
+      "step": 95
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.391304347826087e-05,
+      "loss": 1.7025,
+      "step": 96
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.4057971014492755e-05,
+      "loss": 1.6876,
+      "step": 97
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.420289855072464e-05,
+      "loss": 1.7254,
+      "step": 98
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.4347826086956522e-05,
+      "loss": 1.7035,
+      "step": 99
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4492753623188407e-05,
+      "loss": 1.6701,
+      "step": 100
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4637681159420291e-05,
+      "loss": 1.755,
+      "step": 101
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4782608695652174e-05,
+      "loss": 1.8233,
+      "step": 102
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4927536231884059e-05,
+      "loss": 1.5721,
+      "step": 103
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.5072463768115944e-05,
+      "loss": 1.92,
+      "step": 104
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.5217391304347828e-05,
+      "loss": 1.7907,
+      "step": 105
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.536231884057971e-05,
+      "loss": 1.7155,
+      "step": 106
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.5507246376811594e-05,
+      "loss": 1.5529,
+      "step": 107
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.565217391304348e-05,
+      "loss": 1.6133,
+      "step": 108
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.5797101449275363e-05,
+      "loss": 1.6704,
+      "step": 109
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.5942028985507246e-05,
+      "loss": 1.5658,
+      "step": 110
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.6086956521739132e-05,
+      "loss": 1.638,
+      "step": 111
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.6231884057971015e-05,
+      "loss": 1.9667,
+      "step": 112
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.6376811594202898e-05,
+      "loss": 1.5857,
+      "step": 113
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.6521739130434785e-05,
+      "loss": 1.8468,
+      "step": 114
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 1.8011,
+      "step": 115
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.681159420289855e-05,
+      "loss": 1.5584,
+      "step": 116
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6956521739130437e-05,
+      "loss": 1.5325,
+      "step": 117
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.710144927536232e-05,
+      "loss": 1.6412,
+      "step": 118
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.7246376811594206e-05,
+      "loss": 1.7962,
+      "step": 119
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.739130434782609e-05,
+      "loss": 1.771,
+      "step": 120
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.7536231884057972e-05,
+      "loss": 1.619,
+      "step": 121
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.7681159420289858e-05,
+      "loss": 1.7047,
+      "step": 122
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.782608695652174e-05,
+      "loss": 1.7217,
+      "step": 123
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.7971014492753624e-05,
+      "loss": 1.819,
+      "step": 124
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.811594202898551e-05,
+      "loss": 1.5605,
+      "step": 125
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.8260869565217393e-05,
+      "loss": 1.7051,
+      "step": 126
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.8405797101449276e-05,
+      "loss": 1.6241,
+      "step": 127
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.8550724637681162e-05,
+      "loss": 1.7919,
+      "step": 128
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.8695652173913045e-05,
+      "loss": 1.971,
+      "step": 129
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.8840579710144928e-05,
+      "loss": 1.8275,
+      "step": 130
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8985507246376814e-05,
+      "loss": 1.6933,
+      "step": 131
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9130434782608697e-05,
+      "loss": 1.7722,
+      "step": 132
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.927536231884058e-05,
+      "loss": 1.7086,
+      "step": 133
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9420289855072467e-05,
+      "loss": 1.6655,
+      "step": 134
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.956521739130435e-05,
+      "loss": 1.7234,
+      "step": 135
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9710144927536236e-05,
+      "loss": 1.729,
+      "step": 136
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9855072463768115e-05,
+      "loss": 1.6154,
+      "step": 137
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 1.8506,
+      "step": 138
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9999997510228907e-05,
+      "loss": 1.886,
+      "step": 139
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.999999004091687e-05,
+      "loss": 1.768,
+      "step": 140
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9999977592067603e-05,
+      "loss": 1.6536,
+      "step": 141
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9999960163687307e-05,
+      "loss": 1.8171,
+      "step": 142
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.999993775578466e-05,
+      "loss": 1.62,
+      "step": 143
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9999910368370826e-05,
+      "loss": 1.9765,
+      "step": 144
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9999878001459436e-05,
+      "loss": 1.7381,
+      "step": 145
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999840655066608e-05,
+      "loss": 1.8361,
+      "step": 146
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999798329210938e-05,
+      "loss": 1.7863,
+      "step": 147
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999751023913506e-05,
+      "loss": 1.8941,
+      "step": 148
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999969873919787e-05,
+      "loss": 1.8171,
+      "step": 149
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999964147509006e-05,
+      "loss": 1.5924,
+      "step": 150
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999579231618588e-05,
+      "loss": 1.6363,
+      "step": 151
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999951200881446e-05,
+      "loss": 1.6188,
+      "step": 152
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999439806711137e-05,
+      "loss": 1.6399,
+      "step": 153
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999362625344584e-05,
+      "loss": 1.7759,
+      "step": 154
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999928046475322e-05,
+      "loss": 1.7508,
+      "step": 155
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999193324977974e-05,
+      "loss": 1.7955,
+      "step": 156
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999101206062224e-05,
+      "loss": 1.9154,
+      "step": 157
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999004108051846e-05,
+      "loss": 1.7574,
+      "step": 158
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999890203099519e-05,
+      "loss": 1.6867,
+      "step": 159
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9998794974943087e-05,
+      "loss": 1.8837,
+      "step": 160
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9998682939948843e-05,
+      "loss": 1.7995,
+      "step": 161
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9998565926068253e-05,
+      "loss": 1.7261,
+      "step": 162
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9998443933359576e-05,
+      "loss": 1.7059,
+      "step": 163
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9998316961883563e-05,
+      "loss": 1.8152,
+      "step": 164
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999818501170344e-05,
+      "loss": 1.7966,
+      "step": 165
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999804808288491e-05,
+      "loss": 1.4982,
+      "step": 166
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999790617549616e-05,
+      "loss": 1.9083,
+      "step": 167
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9997759289607854e-05,
+      "loss": 1.6917,
+      "step": 168
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999760742529313e-05,
+      "loss": 1.7672,
+      "step": 169
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9997450582627614e-05,
+      "loss": 1.8419,
+      "step": 170
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9997288761689404e-05,
+      "loss": 1.7393,
+      "step": 171
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999712196255908e-05,
+      "loss": 1.9489,
+      "step": 172
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.99969501853197e-05,
+      "loss": 1.6927,
+      "step": 173
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9996773430056806e-05,
+      "loss": 1.729,
+      "step": 174
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9996591696858407e-05,
+      "loss": 1.8146,
+      "step": 175
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9996404985814998e-05,
+      "loss": 1.5406,
+      "step": 176
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9996213297019556e-05,
+      "loss": 1.834,
+      "step": 177
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9996016630567535e-05,
+      "loss": 1.8505,
+      "step": 178
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.999581498655686e-05,
+      "loss": 1.5933,
+      "step": 179
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9995608365087945e-05,
+      "loss": 1.7946,
+      "step": 180
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9995396766263677e-05,
+      "loss": 1.6388,
+      "step": 181
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9995180190189424e-05,
+      "loss": 1.9924,
+      "step": 182
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9994958636973024e-05,
+      "loss": 1.5948,
+      "step": 183
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.999473210672481e-05,
+      "loss": 1.5846,
+      "step": 184
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9994500599557577e-05,
+      "loss": 1.6471,
+      "step": 185
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.999426411558661e-05,
+      "loss": 1.9565,
+      "step": 186
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9994022654929665e-05,
+      "loss": 1.6896,
+      "step": 187
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9993776217706972e-05,
+      "loss": 1.9445,
+      "step": 188
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9993524804041255e-05,
+      "loss": 1.714,
+      "step": 189
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9993268414057704e-05,
+      "loss": 1.7137,
+      "step": 190
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9993007047883988e-05,
+      "loss": 1.6729,
+      "step": 191
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9992740705650252e-05,
+      "loss": 1.691,
+      "step": 192
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9992469387489127e-05,
+      "loss": 1.7324,
+      "step": 193
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.999219309353572e-05,
+      "loss": 1.57,
+      "step": 194
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9991911823927607e-05,
+      "loss": 2.0345,
+      "step": 195
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.999162557880485e-05,
+      "loss": 1.5059,
+      "step": 196
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.999133435830998e-05,
+      "loss": 1.8392,
+      "step": 197
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9991038162588018e-05,
+      "loss": 1.7923,
+      "step": 198
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9990736991786453e-05,
+      "loss": 1.78,
+      "step": 199
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.999043084605526e-05,
+      "loss": 1.8779,
+      "step": 200
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.999011972554688e-05,
+      "loss": 1.6544,
+      "step": 201
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.998980363041624e-05,
+      "loss": 1.8014,
+      "step": 202
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9989482560820734e-05,
+      "loss": 1.6421,
+      "step": 203
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9989156516920248e-05,
+      "loss": 1.8874,
+      "step": 204
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.998882549887713e-05,
+      "loss": 1.7511,
+      "step": 205
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9988489506856218e-05,
+      "loss": 1.7904,
+      "step": 206
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.998814854102482e-05,
+      "loss": 1.7859,
+      "step": 207
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9987802601552717e-05,
+      "loss": 1.7957,
+      "step": 208
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9987451688612176e-05,
+      "loss": 1.6068,
+      "step": 209
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9987095802377933e-05,
+      "loss": 1.7852,
+      "step": 210
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.99867349430272e-05,
+      "loss": 1.7431,
+      "step": 211
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9986369110739675e-05,
+      "loss": 1.8569,
+      "step": 212
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.998599830569752e-05,
+      "loss": 1.6872,
+      "step": 213
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9985622528085382e-05,
+      "loss": 1.7792,
+      "step": 214
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.998524177809038e-05,
+      "loss": 1.9195,
+      "step": 215
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.998485605590211e-05,
+      "loss": 1.6664,
+      "step": 216
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.998446536171265e-05,
+      "loss": 1.6921,
+      "step": 217
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9984069695716534e-05,
+      "loss": 1.7893,
+      "step": 218
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.99836690581108e-05,
+      "loss": 1.8192,
+      "step": 219
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.998326344909494e-05,
+      "loss": 1.885,
+      "step": 220
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9982852868870927e-05,
+      "loss": 1.556,
+      "step": 221
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9982437317643218e-05,
+      "loss": 1.7575,
+      "step": 222
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.998201679561873e-05,
+      "loss": 1.7023,
+      "step": 223
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.998159130300687e-05,
+      "loss": 1.6913,
+      "step": 224
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9981160840019513e-05,
+      "loss": 1.7819,
+      "step": 225
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9980725406871007e-05,
+      "loss": 1.5825,
+      "step": 226
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.998028500377818e-05,
+      "loss": 1.9634,
+      "step": 227
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9979839630960333e-05,
+      "loss": 1.8114,
+      "step": 228
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9979389288639237e-05,
+      "loss": 1.6686,
+      "step": 229
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.997893397703915e-05,
+      "loss": 1.8282,
+      "step": 230
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9978473696386787e-05,
+      "loss": 1.7462,
+      "step": 231
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9978008446911354e-05,
+      "loss": 1.7526,
+      "step": 232
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.997753822884452e-05,
+      "loss": 1.6126,
+      "step": 233
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9977063042420438e-05,
+      "loss": 1.8654,
+      "step": 234
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.997658288787572e-05,
+      "loss": 1.5399,
+      "step": 235
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9976097765449463e-05,
+      "loss": 1.8716,
+      "step": 236
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9975607675383243e-05,
+      "loss": 1.9406,
+      "step": 237
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9975112617921097e-05,
+      "loss": 1.7728,
+      "step": 238
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.997461259330954e-05,
+      "loss": 1.8756,
+      "step": 239
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.997410760179756e-05,
+      "loss": 1.8006,
+      "step": 240
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.997359764363663e-05,
+      "loss": 1.7244,
+      "step": 241
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9973082719080673e-05,
+      "loss": 1.8612,
+      "step": 242
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.99725628283861e-05,
+      "loss": 1.7037,
+      "step": 243
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9972037971811802e-05,
+      "loss": 1.939,
+      "step": 244
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9971508149619126e-05,
+      "loss": 1.5057,
+      "step": 245
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.99709733620719e-05,
+      "loss": 1.9402,
+      "step": 246
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9970433609436426e-05,
+      "loss": 1.6762,
+      "step": 247
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9969888891981473e-05,
+      "loss": 1.8236,
+      "step": 248
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9969339209978287e-05,
+      "loss": 1.7346,
+      "step": 249
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9968784563700586e-05,
+      "loss": 1.5757,
+      "step": 250
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9968224953424553e-05,
+      "loss": 1.7573,
+      "step": 251
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9967660379428855e-05,
+      "loss": 1.5817,
+      "step": 252
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.996709084199462e-05,
+      "loss": 1.5974,
+      "step": 253
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9966516341405452e-05,
+      "loss": 1.8537,
+      "step": 254
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.996593687794743e-05,
+      "loss": 1.71,
+      "step": 255
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.996535245190909e-05,
+      "loss": 1.7969,
+      "step": 256
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9964763063581465e-05,
+      "loss": 1.8761,
+      "step": 257
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.996416871325803e-05,
+      "loss": 1.8265,
+      "step": 258
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.996356940123475e-05,
+      "loss": 1.5745,
+      "step": 259
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.996296512781005e-05,
+      "loss": 1.575,
+      "step": 260
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.996235589328484e-05,
+      "loss": 1.537,
+      "step": 261
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.996174169796248e-05,
+      "loss": 1.6117,
+      "step": 262
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9961122542148823e-05,
+      "loss": 1.7155,
+      "step": 263
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.996049842615217e-05,
+      "loss": 1.6558,
+      "step": 264
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9959869350283306e-05,
+      "loss": 1.6215,
+      "step": 265
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9959235314855485e-05,
+      "loss": 1.877,
+      "step": 266
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.995859632018442e-05,
+      "loss": 1.7136,
+      "step": 267
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9957952366588307e-05,
+      "loss": 1.6772,
+      "step": 268
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9957303454387803e-05,
+      "loss": 1.6511,
+      "step": 269
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.995664958390604e-05,
+      "loss": 1.717,
+      "step": 270
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9955990755468614e-05,
+      "loss": 1.9313,
+      "step": 271
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9955326969403587e-05,
+      "loss": 1.8296,
+      "step": 272
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.99546582260415e-05,
+      "loss": 1.7657,
+      "step": 273
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9953984525715354e-05,
+      "loss": 1.6089,
+      "step": 274
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.995330586876062e-05,
+      "loss": 1.6183,
+      "step": 275
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.995262225551524e-05,
+      "loss": 1.8083,
+      "step": 276
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9951933686319624e-05,
+      "loss": 1.7245,
+      "step": 277
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9951240161516643e-05,
+      "loss": 1.9381,
+      "step": 278
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9950541681451644e-05,
+      "loss": 1.6028,
+      "step": 279
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9949838246472436e-05,
+      "loss": 1.6434,
+      "step": 280
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9949129856929295e-05,
+      "loss": 1.7555,
+      "step": 281
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9948416513174976e-05,
+      "loss": 1.7311,
+      "step": 282
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.994769821556468e-05,
+      "loss": 1.6387,
+      "step": 283
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9946974964456094e-05,
+      "loss": 1.8235,
+      "step": 284
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.994624676020936e-05,
+      "loss": 1.9122,
+      "step": 285
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9945513603187096e-05,
+      "loss": 1.838,
+      "step": 286
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9944775493754374e-05,
+      "loss": 1.6418,
+      "step": 287
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9944032432278743e-05,
+      "loss": 1.5999,
+      "step": 288
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.994328441913021e-05,
+      "loss": 1.8707,
+      "step": 289
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9942531454681254e-05,
+      "loss": 1.7181,
+      "step": 290
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9941773539306818e-05,
+      "loss": 1.999,
+      "step": 291
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9941010673384307e-05,
+      "loss": 1.7784,
+      "step": 292
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9940242857293594e-05,
+      "loss": 1.6533,
+      "step": 293
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9939470091417012e-05,
+      "loss": 1.7313,
+      "step": 294
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.993869237613937e-05,
+      "loss": 1.6005,
+      "step": 295
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.993790971184793e-05,
+      "loss": 1.8541,
+      "step": 296
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9937122098932428e-05,
+      "loss": 1.7487,
+      "step": 297
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9936329537785054e-05,
+      "loss": 1.9028,
+      "step": 298
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9935532028800465e-05,
+      "loss": 1.7109,
+      "step": 299
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9934729572375792e-05,
+      "loss": 1.5248,
+      "step": 300
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9933922168910617e-05,
+      "loss": 1.7893,
+      "step": 301
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.993310981880699e-05,
+      "loss": 1.6346,
+      "step": 302
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9932292522469424e-05,
+      "loss": 1.6256,
+      "step": 303
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9931470280304895e-05,
+      "loss": 1.8888,
+      "step": 304
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9930643092722843e-05,
+      "loss": 1.6717,
+      "step": 305
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.992981096013517e-05,
+      "loss": 1.9557,
+      "step": 306
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.992897388295624e-05,
+      "loss": 1.9722,
+      "step": 307
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9928131861602878e-05,
+      "loss": 1.7194,
+      "step": 308
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.992728489649437e-05,
+      "loss": 1.7614,
+      "step": 309
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.992643298805247e-05,
+      "loss": 1.7441,
+      "step": 310
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9925576136701386e-05,
+      "loss": 1.728,
+      "step": 311
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.992471434286779e-05,
+      "loss": 1.7892,
+      "step": 312
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9923847606980824e-05,
+      "loss": 1.6874,
+      "step": 313
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9922975929472076e-05,
+      "loss": 1.8102,
+      "step": 314
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.99220993107756e-05,
+      "loss": 1.682,
+      "step": 315
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9921217751327916e-05,
+      "loss": 1.9336,
+      "step": 316
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9920331251568e-05,
+      "loss": 1.6113,
+      "step": 317
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9919439811937283e-05,
+      "loss": 1.9143,
+      "step": 318
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9918543432879667e-05,
+      "loss": 1.7181,
+      "step": 319
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9917642114841505e-05,
+      "loss": 1.6744,
+      "step": 320
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9916735858271615e-05,
+      "loss": 1.7523,
+      "step": 321
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9915824663621267e-05,
+      "loss": 1.78,
+      "step": 322
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9914908531344198e-05,
+      "loss": 1.9025,
+      "step": 323
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9913987461896597e-05,
+      "loss": 1.7448,
+      "step": 324
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9913061455737116e-05,
+      "loss": 1.8273,
+      "step": 325
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9912130513326863e-05,
+      "loss": 1.6921,
+      "step": 326
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.991119463512941e-05,
+      "loss": 1.7164,
+      "step": 327
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.991025382161077e-05,
+      "loss": 1.6556,
+      "step": 328
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9909308073239433e-05,
+      "loss": 1.6554,
+      "step": 329
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9908357390486342e-05,
+      "loss": 1.7639,
+      "step": 330
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9907401773824887e-05,
+      "loss": 1.7496,
+      "step": 331
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.990644122373092e-05,
+      "loss": 1.8797,
+      "step": 332
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9905475740682756e-05,
+      "loss": 1.5464,
+      "step": 333
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.990450532516116e-05,
+      "loss": 1.7168,
+      "step": 334
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.990352997764935e-05,
+      "loss": 1.7238,
+      "step": 335
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.990254969863301e-05,
+      "loss": 1.983,
+      "step": 336
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9901564488600274e-05,
+      "loss": 1.6583,
+      "step": 337
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9900574348041728e-05,
+      "loss": 1.9771,
+      "step": 338
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9899579277450417e-05,
+      "loss": 1.7111,
+      "step": 339
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9898579277321845e-05,
+      "loss": 1.8824,
+      "step": 340
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9897574348153963e-05,
+      "loss": 1.7126,
+      "step": 341
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.989656449044718e-05,
+      "loss": 1.6109,
+      "step": 342
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9895549704704358e-05,
+      "loss": 1.6111,
+      "step": 343
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9894529991430814e-05,
+      "loss": 1.5684,
+      "step": 344
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9893505351134322e-05,
+      "loss": 1.7286,
+      "step": 345
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.98924757843251e-05,
+      "loss": 1.6518,
+      "step": 346
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.989144129151583e-05,
+      "loss": 1.6504,
+      "step": 347
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9890401873221642e-05,
+      "loss": 1.4915,
+      "step": 348
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9889357529960114e-05,
+      "loss": 1.7182,
+      "step": 349
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9888308262251286e-05,
+      "loss": 1.6,
+      "step": 350
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9887254070617644e-05,
+      "loss": 1.6773,
+      "step": 351
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9886194955584126e-05,
+      "loss": 1.7222,
+      "step": 352
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9885130917678123e-05,
+      "loss": 1.799,
+      "step": 353
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.988406195742948e-05,
+      "loss": 1.9064,
+      "step": 354
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.988298807537049e-05,
+      "loss": 1.652,
+      "step": 355
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9881909272035888e-05,
+      "loss": 1.8517,
+      "step": 356
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9880825547962882e-05,
+      "loss": 1.7441,
+      "step": 357
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9879736903691107e-05,
+      "loss": 1.6021,
+      "step": 358
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9878643339762663e-05,
+      "loss": 1.7011,
+      "step": 359
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9877544856722094e-05,
+      "loss": 1.6778,
+      "step": 360
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9876441455116393e-05,
+      "loss": 1.7144,
+      "step": 361
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9875333135495e-05,
+      "loss": 2.0308,
+      "step": 362
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.987421989840982e-05,
+      "loss": 1.7471,
+      "step": 363
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.987310174441518e-05,
+      "loss": 1.5851,
+      "step": 364
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.987197867406788e-05,
+      "loss": 1.873,
+      "step": 365
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.987085068792715e-05,
+      "loss": 1.7713,
+      "step": 366
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.986971778655468e-05,
+      "loss": 1.6572,
+      "step": 367
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.98685799705146e-05,
+      "loss": 1.7438,
+      "step": 368
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.986743724037349e-05,
+      "loss": 1.7512,
+      "step": 369
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9866289596700383e-05,
+      "loss": 1.7058,
+      "step": 370
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9865137040066744e-05,
+      "loss": 1.7243,
+      "step": 371
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.98639795710465e-05,
+      "loss": 1.5387,
+      "step": 372
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.986281719021602e-05,
+      "loss": 1.8603,
+      "step": 373
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9861649898154107e-05,
+      "loss": 1.7283,
+      "step": 374
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.986047769544203e-05,
+      "loss": 1.6644,
+      "step": 375
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.985930058266348e-05,
+      "loss": 1.455,
+      "step": 376
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.985811856040462e-05,
+      "loss": 1.6884,
+      "step": 377
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9856931629254032e-05,
+      "loss": 1.7179,
+      "step": 378
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9855739789802753e-05,
+      "loss": 1.9469,
+      "step": 379
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.985454304264427e-05,
+      "loss": 1.6094,
+      "step": 380
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9853341388374504e-05,
+      "loss": 1.6679,
+      "step": 381
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.985213482759183e-05,
+      "loss": 1.5478,
+      "step": 382
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9850923360897055e-05,
+      "loss": 1.6236,
+      "step": 383
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9849706988893433e-05,
+      "loss": 1.8009,
+      "step": 384
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9848485712186665e-05,
+      "loss": 1.4946,
+      "step": 385
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.984725953138489e-05,
+      "loss": 1.6641,
+      "step": 386
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9846028447098687e-05,
+      "loss": 1.6904,
+      "step": 387
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.984479245994108e-05,
+      "loss": 1.6803,
+      "step": 388
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9843551570527543e-05,
+      "loss": 1.8788,
+      "step": 389
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.984230577947597e-05,
+      "loss": 1.8476,
+      "step": 390
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9841055087406715e-05,
+      "loss": 1.5684,
+      "step": 391
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.983979949494256e-05,
+      "loss": 1.4454,
+      "step": 392
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.983853900270874e-05,
+      "loss": 1.6985,
+      "step": 393
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9837273611332918e-05,
+      "loss": 1.9378,
+      "step": 394
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9836003321445198e-05,
+      "loss": 1.5318,
+      "step": 395
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.983472813367813e-05,
+      "loss": 1.7142,
+      "step": 396
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.98334480486667e-05,
+      "loss": 1.7121,
+      "step": 397
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9832163067048335e-05,
+      "loss": 1.7841,
+      "step": 398
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9830873189462886e-05,
+      "loss": 1.6868,
+      "step": 399
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.982957841655266e-05,
+      "loss": 1.9112,
+      "step": 400
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9828278748962396e-05,
+      "loss": 1.5438,
+      "step": 401
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9826974187339267e-05,
+      "loss": 1.8109,
+      "step": 402
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9825664732332886e-05,
+      "loss": 1.784,
+      "step": 403
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9824350384595295e-05,
+      "loss": 1.5111,
+      "step": 404
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.982303114478099e-05,
+      "loss": 1.5256,
+      "step": 405
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9821707013546885e-05,
+      "loss": 1.6807,
+      "step": 406
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9820377991552338e-05,
+      "loss": 1.606,
+      "step": 407
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9819044079459142e-05,
+      "loss": 1.5803,
+      "step": 408
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9817705277931526e-05,
+      "loss": 1.7415,
+      "step": 409
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9816361587636143e-05,
+      "loss": 1.7781,
+      "step": 410
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9815013009242103e-05,
+      "loss": 1.679,
+      "step": 411
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9813659543420924e-05,
+      "loss": 1.7402,
+      "step": 412
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.981230119084658e-05,
+      "loss": 1.6561,
+      "step": 413
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.981093795219546e-05,
+      "loss": 1.4753,
+      "step": 414
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9809569828146396e-05,
+      "loss": 1.8676,
+      "step": 415
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9808196819380656e-05,
+      "loss": 1.5367,
+      "step": 416
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.980681892658193e-05,
+      "loss": 1.7324,
+      "step": 417
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9805436150436352e-05,
+      "loss": 1.6161,
+      "step": 418
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9804048491632475e-05,
+      "loss": 1.5056,
+      "step": 419
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.980265595086129e-05,
+      "loss": 1.7321,
+      "step": 420
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9801258528816223e-05,
+      "loss": 1.5752,
+      "step": 421
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9799856226193125e-05,
+      "loss": 1.8505,
+      "step": 422
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9798449043690272e-05,
+      "loss": 1.7899,
+      "step": 423
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9797036982008385e-05,
+      "loss": 1.8264,
+      "step": 424
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9795620041850602e-05,
+      "loss": 1.9919,
+      "step": 425
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9794198223922496e-05,
+      "loss": 1.6524,
+      "step": 426
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9792771528932064e-05,
+      "loss": 1.7086,
+      "step": 427
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.979133995758974e-05,
+      "loss": 1.8144,
+      "step": 428
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9789903510608374e-05,
+      "loss": 1.5938,
+      "step": 429
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.978846218870326e-05,
+      "loss": 1.8841,
+      "step": 430
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.97870159925921e-05,
+      "loss": 1.6414,
+      "step": 431
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9785564922995042e-05,
+      "loss": 1.6442,
+      "step": 432
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.978410898063465e-05,
+      "loss": 1.6781,
+      "step": 433
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.978264816623591e-05,
+      "loss": 1.6857,
+      "step": 434
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.978118248052625e-05,
+      "loss": 1.6302,
+      "step": 435
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.977971192423551e-05,
+      "loss": 1.7229,
+      "step": 436
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.977823649809596e-05,
+      "loss": 1.8492,
+      "step": 437
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9776756202842297e-05,
+      "loss": 1.7095,
+      "step": 438
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9775271039211633e-05,
+      "loss": 1.8962,
+      "step": 439
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.977378100794352e-05,
+      "loss": 1.809,
+      "step": 440
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.977228610977992e-05,
+      "loss": 1.5157,
+      "step": 441
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.977078634546523e-05,
+      "loss": 1.8447,
+      "step": 442
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9769281715746258e-05,
+      "loss": 1.5596,
+      "step": 443
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.976777222137224e-05,
+      "loss": 1.6913,
+      "step": 444
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9766257863094843e-05,
+      "loss": 1.5567,
+      "step": 445
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9764738641668137e-05,
+      "loss": 1.7594,
+      "step": 446
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9763214557848634e-05,
+      "loss": 1.8343,
+      "step": 447
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9761685612395253e-05,
+      "loss": 1.8573,
+      "step": 448
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.976015180606934e-05,
+      "loss": 1.6969,
+      "step": 449
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9758613139634662e-05,
+      "loss": 1.6318,
+      "step": 450
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9757069613857404e-05,
+      "loss": 1.7388,
+      "step": 451
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9755521229506164e-05,
+      "loss": 1.7768,
+      "step": 452
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.975396798735198e-05,
+      "loss": 1.684,
+      "step": 453
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9752409888168285e-05,
+      "loss": 1.6109,
+      "step": 454
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9750846932730947e-05,
+      "loss": 1.8364,
+      "step": 455
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9749279121818235e-05,
+      "loss": 1.8766,
+      "step": 456
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9747706456210865e-05,
+      "loss": 1.5277,
+      "step": 457
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.974612893669194e-05,
+      "loss": 1.8831,
+      "step": 458
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9744546564046995e-05,
+      "loss": 1.8072,
+      "step": 459
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9742959339063977e-05,
+      "loss": 1.7829,
+      "step": 460
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.974136726253326e-05,
+      "loss": 1.708,
+      "step": 461
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9739770335247616e-05,
+      "loss": 1.6817,
+      "step": 462
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9738168558002245e-05,
+      "loss": 1.8286,
+      "step": 463
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.973656193159476e-05,
+      "loss": 1.8612,
+      "step": 464
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9734950456825187e-05,
+      "loss": 1.7412,
+      "step": 465
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9733334134495963e-05,
+      "loss": 1.6253,
+      "step": 466
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9731712965411947e-05,
+      "loss": 1.8272,
+      "step": 467
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9730086950380404e-05,
+      "loss": 1.6659,
+      "step": 468
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9728456090211017e-05,
+      "loss": 1.7135,
+      "step": 469
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9726820385715877e-05,
+      "loss": 1.9411,
+      "step": 470
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9725179837709494e-05,
+      "loss": 1.7032,
+      "step": 471
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9723534447008785e-05,
+      "loss": 1.7129,
+      "step": 472
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9721884214433077e-05,
+      "loss": 1.8704,
+      "step": 473
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.972022914080411e-05,
+      "loss": 1.667,
+      "step": 474
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9718569226946035e-05,
+      "loss": 1.5918,
+      "step": 475
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9716904473685417e-05,
+      "loss": 1.5348,
+      "step": 476
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9715234881851226e-05,
+      "loss": 1.7724,
+      "step": 477
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.971356045227484e-05,
+      "loss": 1.927,
+      "step": 478
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.971188118579004e-05,
+      "loss": 1.8668,
+      "step": 479
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9710197083233044e-05,
+      "loss": 1.8976,
+      "step": 480
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9708508145442443e-05,
+      "loss": 1.7757,
+      "step": 481
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.970681437325925e-05,
+      "loss": 1.8585,
+      "step": 482
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9705115767526894e-05,
+      "loss": 1.7761,
+      "step": 483
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.97034123290912e-05,
+      "loss": 1.7709,
+      "step": 484
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.97017040588004e-05,
+      "loss": 1.6896,
+      "step": 485
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9699990957505136e-05,
+      "loss": 1.7458,
+      "step": 486
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9698273026058454e-05,
+      "loss": 1.7141,
+      "step": 487
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9696550265315805e-05,
+      "loss": 1.6869,
+      "step": 488
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9694822676135045e-05,
+      "loss": 1.7167,
+      "step": 489
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9693090259376436e-05,
+      "loss": 1.8046,
+      "step": 490
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9691353015902638e-05,
+      "loss": 1.9046,
+      "step": 491
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9689610946578725e-05,
+      "loss": 1.5962,
+      "step": 492
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.968786405227216e-05,
+      "loss": 1.7309,
+      "step": 493
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9686112333852826e-05,
+      "loss": 1.6692,
+      "step": 494
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.968435579219299e-05,
+      "loss": 1.6645,
+      "step": 495
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.968259442816733e-05,
+      "loss": 1.9673,
+      "step": 496
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.968082824265293e-05,
+      "loss": 1.8449,
+      "step": 497
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9679057236529266e-05,
+      "loss": 1.4464,
+      "step": 498
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9677281410678216e-05,
+      "loss": 1.6464,
+      "step": 499
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9675500765984066e-05,
+      "loss": 1.6744,
+      "step": 500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9673715303333495e-05,
+      "loss": 1.4672,
+      "step": 501
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9671925023615572e-05,
+      "loss": 1.6775,
+      "step": 502
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9670129927721783e-05,
+      "loss": 1.7813,
+      "step": 503
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9668330016546004e-05,
+      "loss": 1.6196,
+      "step": 504
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9666525290984505e-05,
+      "loss": 1.6561,
+      "step": 505
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9664715751935958e-05,
+      "loss": 1.7315,
+      "step": 506
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9662901400301432e-05,
+      "loss": 1.8063,
+      "step": 507
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9661082236984387e-05,
+      "loss": 1.4929,
+      "step": 508
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9659258262890683e-05,
+      "loss": 1.8631,
+      "step": 509
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.965742947892858e-05,
+      "loss": 1.6901,
+      "step": 510
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.965559588600873e-05,
+      "loss": 1.7438,
+      "step": 511
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.965375748504417e-05,
+      "loss": 1.6929,
+      "step": 512
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.965191427695035e-05,
+      "loss": 1.6071,
+      "step": 513
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9650066262645097e-05,
+      "loss": 1.5969,
+      "step": 514
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9648213443048635e-05,
+      "loss": 1.634,
+      "step": 515
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.964635581908359e-05,
+      "loss": 1.8549,
+      "step": 516
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9644493391674966e-05,
+      "loss": 1.6202,
+      "step": 517
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9642626161750176e-05,
+      "loss": 1.8871,
+      "step": 518
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9640754130239004e-05,
+      "loss": 1.6687,
+      "step": 519
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9638877298073645e-05,
+      "loss": 1.612,
+      "step": 520
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9636995666188677e-05,
+      "loss": 1.785,
+      "step": 521
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9635109235521057e-05,
+      "loss": 1.5052,
+      "step": 522
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9633218007010146e-05,
+      "loss": 1.7157,
+      "step": 523
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.963132198159769e-05,
+      "loss": 1.5638,
+      "step": 524
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9629421160227825e-05,
+      "loss": 1.5479,
+      "step": 525
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9627515543847068e-05,
+      "loss": 1.5164,
+      "step": 526
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.962560513340433e-05,
+      "loss": 1.648,
+      "step": 527
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.962368992985091e-05,
+      "loss": 1.7224,
+      "step": 528
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.962176993414049e-05,
+      "loss": 1.7266,
+      "step": 529
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.961984514722914e-05,
+      "loss": 1.5928,
+      "step": 530
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9617915570075317e-05,
+      "loss": 1.5717,
+      "step": 531
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.961598120363986e-05,
+      "loss": 1.6539,
+      "step": 532
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9614042048886e-05,
+      "loss": 1.7196,
+      "step": 533
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.961209810677934e-05,
+      "loss": 1.6336,
+      "step": 534
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9610149378287877e-05,
+      "loss": 1.6449,
+      "step": 535
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9608195864381994e-05,
+      "loss": 1.7273,
+      "step": 536
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9606237566034443e-05,
+      "loss": 1.9775,
+      "step": 537
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.960427448422037e-05,
+      "loss": 1.6844,
+      "step": 538
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9602306619917305e-05,
+      "loss": 1.4817,
+      "step": 539
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9600333974105145e-05,
+      "loss": 1.6722,
+      "step": 540
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9598356547766183e-05,
+      "loss": 1.7554,
+      "step": 541
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9596374341885093e-05,
+      "loss": 1.6304,
+      "step": 542
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9594387357448908e-05,
+      "loss": 1.922,
+      "step": 543
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9592395595447064e-05,
+      "loss": 1.7577,
+      "step": 544
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9590399056871366e-05,
+      "loss": 1.764,
+      "step": 545
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9588397742716004e-05,
+      "loss": 1.794,
+      "step": 546
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.958639165397753e-05,
+      "loss": 1.846,
+      "step": 547
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9584380791654896e-05,
+      "loss": 1.7184,
+      "step": 548
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.958236515674941e-05,
+      "loss": 1.7393,
+      "step": 549
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.958034475026477e-05,
+      "loss": 1.6775,
+      "step": 550
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9578319573207047e-05,
+      "loss": 1.8334,
+      "step": 551
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9576289626584685e-05,
+      "loss": 1.4963,
+      "step": 552
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9574254911408502e-05,
+      "loss": 1.5934,
+      "step": 553
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.95722154286917e-05,
+      "loss": 1.8733,
+      "step": 554
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.957017117944984e-05,
+      "loss": 1.6129,
+      "step": 555
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9568122164700867e-05,
+      "loss": 1.5116,
+      "step": 556
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.95660683854651e-05,
+      "loss": 1.7035,
+      "step": 557
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9564009842765225e-05,
+      "loss": 1.4937,
+      "step": 558
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.95619465376263e-05,
+      "loss": 1.7986,
+      "step": 559
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9559878471075763e-05,
+      "loss": 1.7467,
+      "step": 560
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9557805644143408e-05,
+      "loss": 1.4965,
+      "step": 561
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.955572805786141e-05,
+      "loss": 1.7962,
+      "step": 562
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9553645713264312e-05,
+      "loss": 1.6441,
+      "step": 563
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.955155861138903e-05,
+      "loss": 1.7992,
+      "step": 564
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9549466753274843e-05,
+      "loss": 1.7547,
+      "step": 565
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9547370139963406e-05,
+      "loss": 1.6163,
+      "step": 566
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9545268772498723e-05,
+      "loss": 1.6678,
+      "step": 567
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.954316265192719e-05,
+      "loss": 1.7166,
+      "step": 568
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9541051779297553e-05,
+      "loss": 1.8128,
+      "step": 569
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9538936155660934e-05,
+      "loss": 2.1209,
+      "step": 570
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9536815782070816e-05,
+      "loss": 1.5158,
+      "step": 571
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9534690659583045e-05,
+      "loss": 1.7709,
+      "step": 572
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9532560789255838e-05,
+      "loss": 1.5841,
+      "step": 573
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.953042617214977e-05,
+      "loss": 1.5374,
+      "step": 574
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9528286809327786e-05,
+      "loss": 1.5786,
+      "step": 575
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9526142701855187e-05,
+      "loss": 1.688,
+      "step": 576
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9523993850799644e-05,
+      "loss": 1.5085,
+      "step": 577
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9521840257231183e-05,
+      "loss": 1.6474,
+      "step": 578
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9519681922222195e-05,
+      "loss": 1.9484,
+      "step": 579
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9517518846847437e-05,
+      "loss": 1.7364,
+      "step": 580
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.951535103218402e-05,
+      "loss": 1.7895,
+      "step": 581
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.951317847931141e-05,
+      "loss": 1.8405,
+      "step": 582
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9511001189311446e-05,
+      "loss": 1.6786,
+      "step": 583
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9508819163268315e-05,
+      "loss": 1.8383,
+      "step": 584
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.950663240226857e-05,
+      "loss": 1.732,
+      "step": 585
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9504440907401113e-05,
+      "loss": 1.6186,
+      "step": 586
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.950224467975721e-05,
+      "loss": 1.6698,
+      "step": 587
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9500043720430484e-05,
+      "loss": 1.5525,
+      "step": 588
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9497838030516905e-05,
+      "loss": 1.5394,
+      "step": 589
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9495627611114817e-05,
+      "loss": 1.7961,
+      "step": 590
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9493412463324903e-05,
+      "loss": 1.7375,
+      "step": 591
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9491192588250198e-05,
+      "loss": 1.6577,
+      "step": 592
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9488967986996105e-05,
+      "loss": 1.6901,
+      "step": 593
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9486738660670373e-05,
+      "loss": 1.6328,
+      "step": 594
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9484504610383103e-05,
+      "loss": 1.7063,
+      "step": 595
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.948226583724675e-05,
+      "loss": 1.7627,
+      "step": 596
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9480022342376125e-05,
+      "loss": 1.8574,
+      "step": 597
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.947777412688838e-05,
+      "loss": 1.5624,
+      "step": 598
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9475521191903023e-05,
+      "loss": 1.5426,
+      "step": 599
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9473263538541916e-05,
+      "loss": 1.5822,
+      "step": 600
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9471001167929266e-05,
+      "loss": 1.7343,
+      "step": 601
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9468734081191627e-05,
+      "loss": 1.9196,
+      "step": 602
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.946646227945791e-05,
+      "loss": 1.6072,
+      "step": 603
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9464185763859365e-05,
+      "loss": 1.7154,
+      "step": 604
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9461904535529593e-05,
+      "loss": 1.6671,
+      "step": 605
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.945961859560454e-05,
+      "loss": 1.5347,
+      "step": 606
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9457327945222505e-05,
+      "loss": 1.6004,
+      "step": 607
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9455032585524115e-05,
+      "loss": 1.6925,
+      "step": 608
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9452732517652366e-05,
+      "loss": 1.801,
+      "step": 609
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9450427742752583e-05,
+      "loss": 1.534,
+      "step": 610
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9448118261972437e-05,
+      "loss": 1.6647,
+      "step": 611
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9445804076461946e-05,
+      "loss": 1.8082,
+      "step": 612
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9443485187373463e-05,
+      "loss": 1.6061,
+      "step": 613
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.944116159586169e-05,
+      "loss": 1.535,
+      "step": 614
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9438833303083677e-05,
+      "loss": 1.7386,
+      "step": 615
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.94365003101988e-05,
+      "loss": 1.7329,
+      "step": 616
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9434162618368784e-05,
+      "loss": 1.9582,
+      "step": 617
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.943182022875769e-05,
+      "loss": 1.6357,
+      "step": 618
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.942947314253193e-05,
+      "loss": 1.6327,
+      "step": 619
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9427121360860233e-05,
+      "loss": 1.9452,
+      "step": 620
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.942476488491369e-05,
+      "loss": 1.5513,
+      "step": 621
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9422403715865708e-05,
+      "loss": 1.5723,
+      "step": 622
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9420037854892053e-05,
+      "loss": 1.6096,
+      "step": 623
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9417667303170803e-05,
+      "loss": 1.693,
+      "step": 624
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9415292061882386e-05,
+      "loss": 1.7972,
+      "step": 625
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9412912132209573e-05,
+      "loss": 1.6269,
+      "step": 626
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.941052751533745e-05,
+      "loss": 1.6658,
+      "step": 627
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9408138212453456e-05,
+      "loss": 1.6587,
+      "step": 628
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9405744224747348e-05,
+      "loss": 1.6423,
+      "step": 629
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.940334555341122e-05,
+      "loss": 1.7249,
+      "step": 630
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.940094219963951e-05,
+      "loss": 1.5317,
+      "step": 631
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9398534164628968e-05,
+      "loss": 1.762,
+      "step": 632
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9396121449578694e-05,
+      "loss": 1.7897,
+      "step": 633
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.93937040556901e-05,
+      "loss": 1.7555,
+      "step": 634
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9391281984166944e-05,
+      "loss": 1.3809,
+      "step": 635
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.938885523621531e-05,
+      "loss": 1.7534,
+      "step": 636
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.93864238130436e-05,
+      "loss": 1.881,
+      "step": 637
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9383987715862554e-05,
+      "loss": 1.5816,
+      "step": 638
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9381546945885236e-05,
+      "loss": 1.8615,
+      "step": 639
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9379101504327044e-05,
+      "loss": 1.6472,
+      "step": 640
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9376651392405685e-05,
+      "loss": 1.8553,
+      "step": 641
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9374196611341212e-05,
+      "loss": 1.5499,
+      "step": 642
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9371737162355986e-05,
+      "loss": 1.5872,
+      "step": 643
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9369273046674708e-05,
+      "loss": 1.4952,
+      "step": 644
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9366804265524388e-05,
+      "loss": 1.7629,
+      "step": 645
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.936433082013437e-05,
+      "loss": 1.5493,
+      "step": 646
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9361852711736317e-05,
+      "loss": 1.5898,
+      "step": 647
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.935936994156421e-05,
+      "loss": 1.6559,
+      "step": 648
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.935688251085436e-05,
+      "loss": 1.6534,
+      "step": 649
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9354390420845387e-05,
+      "loss": 1.8454,
+      "step": 650
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.935189367277824e-05,
+      "loss": 1.661,
+      "step": 651
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.934939226789619e-05,
+      "loss": 1.6613,
+      "step": 652
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9346886207444817e-05,
+      "loss": 1.4516,
+      "step": 653
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9344375492672024e-05,
+      "loss": 1.6293,
+      "step": 654
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9341860124828037e-05,
+      "loss": 1.5812,
+      "step": 655
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.933934010516539e-05,
+      "loss": 1.6494,
+      "step": 656
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9336815434938935e-05,
+      "loss": 1.8236,
+      "step": 657
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.933428611540585e-05,
+      "loss": 1.4262,
+      "step": 658
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9331752147825615e-05,
+      "loss": 1.8037,
+      "step": 659
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.932921353346003e-05,
+      "loss": 1.5748,
+      "step": 660
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9326670273573208e-05,
+      "loss": 1.7351,
+      "step": 661
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.932412236943158e-05,
+      "loss": 1.6881,
+      "step": 662
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.932156982230388e-05,
+      "loss": 1.7435,
+      "step": 663
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9319012633461164e-05,
+      "loss": 1.8803,
+      "step": 664
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9316450804176794e-05,
+      "loss": 2.0769,
+      "step": 665
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9313884335726443e-05,
+      "loss": 1.6239,
+      "step": 666
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9311313229388094e-05,
+      "loss": 1.6221,
+      "step": 667
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9308737486442045e-05,
+      "loss": 1.6181,
+      "step": 668
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9306157108170892e-05,
+      "loss": 2.0024,
+      "step": 669
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9303572095859545e-05,
+      "loss": 1.6205,
+      "step": 670
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.930098245079523e-05,
+      "loss": 1.4555,
+      "step": 671
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9298388174267463e-05,
+      "loss": 1.7335,
+      "step": 672
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9295789267568078e-05,
+      "loss": 1.6883,
+      "step": 673
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9293185731991212e-05,
+      "loss": 1.6186,
+      "step": 674
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9290577568833305e-05,
+      "loss": 1.6508,
+      "step": 675
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9287964779393105e-05,
+      "loss": 1.4298,
+      "step": 676
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9285347364971662e-05,
+      "loss": 1.868,
+      "step": 677
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9282725326872324e-05,
+      "loss": 1.5846,
+      "step": 678
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9280098666400752e-05,
+      "loss": 1.733,
+      "step": 679
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.92774673848649e-05,
+      "loss": 1.5303,
+      "step": 680
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.927483148357502e-05,
+      "loss": 1.4126,
+      "step": 681
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.927219096384368e-05,
+      "loss": 1.5526,
+      "step": 682
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9269545826985733e-05,
+      "loss": 1.7014,
+      "step": 683
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9266896074318335e-05,
+      "loss": 1.7935,
+      "step": 684
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9264241707160946e-05,
+      "loss": 1.7879,
+      "step": 685
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9261582726835316e-05,
+      "loss": 1.4795,
+      "step": 686
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9258919134665494e-05,
+      "loss": 1.73,
+      "step": 687
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.925625093197783e-05,
+      "loss": 1.6534,
+      "step": 688
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9253578120100962e-05,
+      "loss": 1.6437,
+      "step": 689
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9250900700365837e-05,
+      "loss": 1.6342,
+      "step": 690
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9248218674105678e-05,
+      "loss": 1.3695,
+      "step": 691
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.924553204265602e-05,
+      "loss": 1.8255,
+      "step": 692
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9242840807354676e-05,
+      "loss": 1.9887,
+      "step": 693
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9240144969541754e-05,
+      "loss": 1.7133,
+      "step": 694
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.923744453055967e-05,
+      "loss": 1.5942,
+      "step": 695
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.923473949175311e-05,
+      "loss": 1.5758,
+      "step": 696
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9232029854469064e-05,
+      "loss": 1.8235,
+      "step": 697
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9229315620056805e-05,
+      "loss": 1.6101,
+      "step": 698
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9226596789867895e-05,
+      "loss": 1.6482,
+      "step": 699
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9223873365256192e-05,
+      "loss": 1.6308,
+      "step": 700
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9221145347577833e-05,
+      "loss": 1.9061,
+      "step": 701
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.921841273819125e-05,
+      "loss": 1.7377,
+      "step": 702
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9215675538457155e-05,
+      "loss": 1.7004,
+      "step": 703
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9212933749738546e-05,
+      "loss": 1.5572,
+      "step": 704
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.921018737340071e-05,
+      "loss": 1.5642,
+      "step": 705
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.920743641081122e-05,
+      "loss": 1.476,
+      "step": 706
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9204680863339922e-05,
+      "loss": 1.6446,
+      "step": 707
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.920192073235896e-05,
+      "loss": 1.7077,
+      "step": 708
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.919915601924275e-05,
+      "loss": 1.7782,
+      "step": 709
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.919638672536799e-05,
+      "loss": 1.8962,
+      "step": 710
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9193612852113666e-05,
+      "loss": 1.7171,
+      "step": 711
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9190834400861035e-05,
+      "loss": 1.6728,
+      "step": 712
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9188051372993645e-05,
+      "loss": 1.5485,
+      "step": 713
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.918526376989731e-05,
+      "loss": 1.7059,
+      "step": 714
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.918247159296013e-05,
+      "loss": 1.6561,
+      "step": 715
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9179674843572487e-05,
+      "loss": 1.6245,
+      "step": 716
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9176873523127025e-05,
+      "loss": 1.449,
+      "step": 717
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9174067633018682e-05,
+      "loss": 1.553,
+      "step": 718
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9171257174644658e-05,
+      "loss": 1.7433,
+      "step": 719
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9168442149404438e-05,
+      "loss": 1.4599,
+      "step": 720
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9165622558699763e-05,
+      "loss": 1.7968,
+      "step": 721
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.916279840393467e-05,
+      "loss": 1.761,
+      "step": 722
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9159969686515462e-05,
+      "loss": 1.764,
+      "step": 723
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9157136407850706e-05,
+      "loss": 1.8362,
+      "step": 724
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.915429856935124e-05,
+      "loss": 1.7414,
+      "step": 725
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9151456172430186e-05,
+      "loss": 1.6869,
+      "step": 726
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9148609218502923e-05,
+      "loss": 1.7816,
+      "step": 727
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9145757708987103e-05,
+      "loss": 1.7204,
+      "step": 728
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9142901645302654e-05,
+      "loss": 1.7609,
+      "step": 729
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.914004102887176e-05,
+      "loss": 1.7315,
+      "step": 730
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9137175861118873e-05,
+      "loss": 1.6165,
+      "step": 731
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9134306143470722e-05,
+      "loss": 1.3642,
+      "step": 732
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9131431877356292e-05,
+      "loss": 1.5392,
+      "step": 733
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9128553064206835e-05,
+      "loss": 1.854,
+      "step": 734
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.912566970545587e-05,
+      "loss": 1.8508,
+      "step": 735
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.912278180253918e-05,
+      "loss": 1.7225,
+      "step": 736
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9119889356894804e-05,
+      "loss": 1.841,
+      "step": 737
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.911699236996305e-05,
+      "loss": 1.622,
+      "step": 738
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9114090843186482e-05,
+      "loss": 1.8323,
+      "step": 739
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9111184778009934e-05,
+      "loss": 1.7356,
+      "step": 740
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9108274175880483e-05,
+      "loss": 1.6885,
+      "step": 741
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9105359038247484e-05,
+      "loss": 1.7559,
+      "step": 742
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.910243936656254e-05,
+      "loss": 1.6461,
+      "step": 743
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9099515162279515e-05,
+      "loss": 1.4499,
+      "step": 744
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9096586426854523e-05,
+      "loss": 1.6665,
+      "step": 745
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.909365316174595e-05,
+      "loss": 2.0526,
+      "step": 746
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.909071536841442e-05,
+      "loss": 1.4342,
+      "step": 747
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.908777304832282e-05,
+      "loss": 1.5707,
+      "step": 748
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9084826202936292e-05,
+      "loss": 1.5996,
+      "step": 749
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9081874833722234e-05,
+      "loss": 1.6187,
+      "step": 750
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9078918942150292e-05,
+      "loss": 1.6321,
+      "step": 751
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9075958529692355e-05,
+      "loss": 1.7505,
+      "step": 752
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9072993597822586e-05,
+      "loss": 1.7084,
+      "step": 753
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9070024148017375e-05,
+      "loss": 1.8183,
+      "step": 754
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9067050181755377e-05,
+      "loss": 1.4917,
+      "step": 755
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.906407170051749e-05,
+      "loss": 1.6107,
+      "step": 756
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9061088705786863e-05,
+      "loss": 1.7112,
+      "step": 757
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.905810119904889e-05,
+      "loss": 1.7943,
+      "step": 758
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.905510918179121e-05,
+      "loss": 1.6517,
+      "step": 759
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9052112655503713e-05,
+      "loss": 1.5343,
+      "step": 760
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9049111621678534e-05,
+      "loss": 1.6211,
+      "step": 761
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9046106081810047e-05,
+      "loss": 1.5274,
+      "step": 762
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.904309603739487e-05,
+      "loss": 1.6842,
+      "step": 763
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.904008148993188e-05,
+      "loss": 1.4002,
+      "step": 764
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.903706244092217e-05,
+      "loss": 1.5549,
+      "step": 765
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.903403889186909e-05,
+      "loss": 1.7794,
+      "step": 766
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.903101084427824e-05,
+      "loss": 1.6031,
+      "step": 767
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9027978299657436e-05,
+      "loss": 1.8203,
+      "step": 768
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9024941259516753e-05,
+      "loss": 1.6061,
+      "step": 769
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9021899725368498e-05,
+      "loss": 1.535,
+      "step": 770
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9018853698727216e-05,
+      "loss": 1.6099,
+      "step": 771
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9015803181109685e-05,
+      "loss": 1.5597,
+      "step": 772
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.9012748174034926e-05,
+      "loss": 1.54,
+      "step": 773
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.900968867902419e-05,
+      "loss": 1.6177,
+      "step": 774
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.900662469760097e-05,
+      "loss": 1.6051,
+      "step": 775
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.9003556231290988e-05,
+      "loss": 1.5715,
+      "step": 776
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.9000483281622198e-05,
+      "loss": 1.528,
+      "step": 777
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.8997405850124786e-05,
+      "loss": 1.4962,
+      "step": 778
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.8994323938331174e-05,
+      "loss": 1.5675,
+      "step": 779
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.8991237547776014e-05,
+      "loss": 1.7189,
+      "step": 780
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.8988146679996184e-05,
+      "loss": 1.5303,
+      "step": 781
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.89850513365308e-05,
+      "loss": 1.581,
+      "step": 782
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.8981951518921194e-05,
+      "loss": 1.4948,
+      "step": 783
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.897884722871094e-05,
+      "loss": 1.5518,
+      "step": 784
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.897573846744583e-05,
+      "loss": 1.637,
+      "step": 785
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.8972625236673887e-05,
+      "loss": 1.6985,
+      "step": 786
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.8969507537945354e-05,
+      "loss": 1.7558,
+      "step": 787
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.8966385372812703e-05,
+      "loss": 1.5349,
+      "step": 788
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.896325874283063e-05,
+      "loss": 1.6001,
+      "step": 789
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.896012764955605e-05,
+      "loss": 1.5696,
+      "step": 790
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.895699209454811e-05,
+      "loss": 1.5245,
+      "step": 791
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.895385207936817e-05,
+      "loss": 1.8194,
+      "step": 792
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8950707605579816e-05,
+      "loss": 1.6387,
+      "step": 793
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8947558674748844e-05,
+      "loss": 1.685,
+      "step": 794
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.894440528844329e-05,
+      "loss": 1.5585,
+      "step": 795
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8941247448233386e-05,
+      "loss": 1.5872,
+      "step": 796
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8938085155691598e-05,
+      "loss": 1.5518,
+      "step": 797
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8934918412392596e-05,
+      "loss": 1.601,
+      "step": 798
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.893174721991328e-05,
+      "loss": 1.4921,
+      "step": 799
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8928571579832756e-05,
+      "loss": 1.5792,
+      "step": 800
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8925391493732346e-05,
+      "loss": 1.6429,
+      "step": 801
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.892220696319559e-05,
+      "loss": 1.5469,
+      "step": 802
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8919017989808238e-05,
+      "loss": 1.8114,
+      "step": 803
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.891582457515825e-05,
+      "loss": 1.8066,
+      "step": 804
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.89126267208358e-05,
+      "loss": 1.4935,
+      "step": 805
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8909424428433278e-05,
+      "loss": 1.9584,
+      "step": 806
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.890621769954528e-05,
+      "loss": 1.49,
+      "step": 807
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.89030065357686e-05,
+      "loss": 1.6055,
+      "step": 808
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.889979093870226e-05,
+      "loss": 1.4847,
+      "step": 809
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8896570909947477e-05,
+      "loss": 1.8265,
+      "step": 810
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8893346451107674e-05,
+      "loss": 1.6387,
+      "step": 811
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.889011756378849e-05,
+      "loss": 1.9597,
+      "step": 812
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8886884249597763e-05,
+      "loss": 1.5448,
+      "step": 813
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.888364651014553e-05,
+      "loss": 1.5469,
+      "step": 814
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8880404347044046e-05,
+      "loss": 1.7137,
+      "step": 815
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.887715776190775e-05,
+      "loss": 1.5354,
+      "step": 816
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.88739067563533e-05,
+      "loss": 1.6245,
+      "step": 817
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8870651331999542e-05,
+      "loss": 1.7649,
+      "step": 818
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8867391490467534e-05,
+      "loss": 1.886,
+      "step": 819
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.886412723338052e-05,
+      "loss": 1.7526,
+      "step": 820
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8860858562363957e-05,
+      "loss": 1.7467,
+      "step": 821
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8857585479045493e-05,
+      "loss": 1.5234,
+      "step": 822
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8854307985054973e-05,
+      "loss": 1.5892,
+      "step": 823
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.885102608202444e-05,
+      "loss": 1.4823,
+      "step": 824
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.884773977158813e-05,
+      "loss": 1.8852,
+      "step": 825
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8844449055382473e-05,
+      "loss": 1.8052,
+      "step": 826
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8841153935046098e-05,
+      "loss": 1.6245,
+      "step": 827
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8837854412219828e-05,
+      "loss": 1.5381,
+      "step": 828
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8834550488546663e-05,
+      "loss": 1.6746,
+      "step": 829
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8831242165671816e-05,
+      "loss": 1.6597,
+      "step": 830
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8827929445242678e-05,
+      "loss": 1.658,
+      "step": 831
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8824612328908828e-05,
+      "loss": 1.6956,
+      "step": 832
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.882129081832204e-05,
+      "loss": 1.6015,
+      "step": 833
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8817964915136277e-05,
+      "loss": 1.7438,
+      "step": 834
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8814634621007685e-05,
+      "loss": 2.0059,
+      "step": 835
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8811299937594598e-05,
+      "loss": 1.7808,
+      "step": 836
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8807960866557535e-05,
+      "loss": 1.7095,
+      "step": 837
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.88046174095592e-05,
+      "loss": 1.4648,
+      "step": 838
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.880126956826448e-05,
+      "loss": 1.5142,
+      "step": 839
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8797917344340453e-05,
+      "loss": 1.7125,
+      "step": 840
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.879456073945637e-05,
+      "loss": 1.5195,
+      "step": 841
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8791199755283664e-05,
+      "loss": 1.777,
+      "step": 842
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8787834393495952e-05,
+      "loss": 1.6173,
+      "step": 843
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8784464655769033e-05,
+      "loss": 1.6622,
+      "step": 844
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8781090543780876e-05,
+      "loss": 1.5369,
+      "step": 845
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8777712059211643e-05,
+      "loss": 1.5513,
+      "step": 846
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.877432920374366e-05,
+      "loss": 1.5864,
+      "step": 847
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8770941979061432e-05,
+      "loss": 1.5691,
+      "step": 848
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8767550386851646e-05,
+      "loss": 1.5007,
+      "step": 849
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8764154428803155e-05,
+      "loss": 1.5827,
+      "step": 850
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8760754106606997e-05,
+      "loss": 1.5528,
+      "step": 851
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.875734942195637e-05,
+      "loss": 1.7157,
+      "step": 852
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8753940376546656e-05,
+      "loss": 1.5487,
+      "step": 853
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.87505269720754e-05,
+      "loss": 1.6451,
+      "step": 854
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8747109210242326e-05,
+      "loss": 1.8242,
+      "step": 855
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8743687092749318e-05,
+      "loss": 1.7555,
+      "step": 856
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8740260621300433e-05,
+      "loss": 1.6325,
+      "step": 857
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8736829797601903e-05,
+      "loss": 1.6707,
+      "step": 858
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8733394623362116e-05,
+      "loss": 1.7867,
+      "step": 859
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8729955100291634e-05,
+      "loss": 1.5048,
+      "step": 860
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8726511230103183e-05,
+      "loss": 1.6509,
+      "step": 861
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.872306301451165e-05,
+      "loss": 1.6407,
+      "step": 862
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.871961045523409e-05,
+      "loss": 1.6166,
+      "step": 863
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8716153553989716e-05,
+      "loss": 1.508,
+      "step": 864
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.871269231249991e-05,
+      "loss": 1.6792,
+      "step": 865
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8709226732488216e-05,
+      "loss": 1.6021,
+      "step": 866
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8705756815680324e-05,
+      "loss": 1.6887,
+      "step": 867
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.87022825638041e-05,
+      "loss": 1.6528,
+      "step": 868
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8698803978589564e-05,
+      "loss": 1.6668,
+      "step": 869
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8695321061768886e-05,
+      "loss": 1.4839,
+      "step": 870
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8691833815076406e-05,
+      "loss": 1.6164,
+      "step": 871
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.868834224024861e-05,
+      "loss": 1.8147,
+      "step": 872
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8684846339024145e-05,
+      "loss": 1.8088,
+      "step": 873
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.86813461131438e-05,
+      "loss": 1.5114,
+      "step": 874
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.867784156435054e-05,
+      "loss": 1.4502,
+      "step": 875
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8674332694389464e-05,
+      "loss": 1.5312,
+      "step": 876
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8670819505007826e-05,
+      "loss": 1.6685,
+      "step": 877
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8667301997955038e-05,
+      "loss": 1.9059,
+      "step": 878
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8663780174982656e-05,
+      "loss": 1.5351,
+      "step": 879
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.866025403784439e-05,
+      "loss": 1.673,
+      "step": 880
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8656723588296085e-05,
+      "loss": 1.7394,
+      "step": 881
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8653188828095754e-05,
+      "loss": 1.8356,
+      "step": 882
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.864964975900354e-05,
+      "loss": 1.5036,
+      "step": 883
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8646106382781738e-05,
+      "loss": 1.6354,
+      "step": 884
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8642558701194787e-05,
+      "loss": 1.6454,
+      "step": 885
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8639006716009275e-05,
+      "loss": 1.7203,
+      "step": 886
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.863545042899392e-05,
+      "loss": 1.7076,
+      "step": 887
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8631889841919596e-05,
+      "loss": 1.4176,
+      "step": 888
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8628324956559313e-05,
+      "loss": 1.5047,
+      "step": 889
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.862475577468821e-05,
+      "loss": 1.9395,
+      "step": 890
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.862118229808359e-05,
+      "loss": 1.4564,
+      "step": 891
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8617604528524876e-05,
+      "loss": 1.5646,
+      "step": 892
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.861402246779363e-05,
+      "loss": 1.7551,
+      "step": 893
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8610436117673557e-05,
+      "loss": 1.5105,
+      "step": 894
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8606845479950494e-05,
+      "loss": 1.6701,
+      "step": 895
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8603250556412418e-05,
+      "loss": 1.6202,
+      "step": 896
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.859965134884943e-05,
+      "loss": 1.5629,
+      "step": 897
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8596047859053776e-05,
+      "loss": 1.6441,
+      "step": 898
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.859244008881983e-05,
+      "loss": 1.6198,
+      "step": 899
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8588828039944086e-05,
+      "loss": 1.5501,
+      "step": 900
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.858521171422519e-05,
+      "loss": 1.5689,
+      "step": 901
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8581591113463903e-05,
+      "loss": 1.9153,
+      "step": 902
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.857796623946312e-05,
+      "loss": 1.6043,
+      "step": 903
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8574337094027858e-05,
+      "loss": 1.6687,
+      "step": 904
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.857070367896527e-05,
+      "loss": 1.7897,
+      "step": 905
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8567065996084628e-05,
+      "loss": 1.5528,
+      "step": 906
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.856342404719733e-05,
+      "loss": 1.6756,
+      "step": 907
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8559777834116906e-05,
+      "loss": 1.6897,
+      "step": 908
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8556127358658995e-05,
+      "loss": 1.6393,
+      "step": 909
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8552472622641372e-05,
+      "loss": 1.8571,
+      "step": 910
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8548813627883924e-05,
+      "loss": 1.922,
+      "step": 911
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.854515037620867e-05,
+      "loss": 1.3198,
+      "step": 912
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8541482869439736e-05,
+      "loss": 1.6532,
+      "step": 913
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8537811109403372e-05,
+      "loss": 1.6219,
+      "step": 914
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.853413509792795e-05,
+      "loss": 1.6325,
+      "step": 915
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8530454836843953e-05,
+      "loss": 1.5912,
+      "step": 916
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.852677032798398e-05,
+      "loss": 1.7899,
+      "step": 917
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8523081573182754e-05,
+      "loss": 1.6063,
+      "step": 918
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8519388574277105e-05,
+      "loss": 1.763,
+      "step": 919
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.851569133310597e-05,
+      "loss": 1.4422,
+      "step": 920
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8511989851510416e-05,
+      "loss": 1.6333,
+      "step": 921
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8508284131333604e-05,
+      "loss": 1.6701,
+      "step": 922
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8504574174420812e-05,
+      "loss": 1.4986,
+      "step": 923
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8500859982619438e-05,
+      "loss": 1.809,
+      "step": 924
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.849714155777897e-05,
+      "loss": 1.5901,
+      "step": 925
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8493418901751016e-05,
+      "loss": 1.7828,
+      "step": 926
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.848969201638929e-05,
+      "loss": 1.5578,
+      "step": 927
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8485960903549614e-05,
+      "loss": 1.593,
+      "step": 928
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8482225565089905e-05,
+      "loss": 1.7663,
+      "step": 929
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.847848600287019e-05,
+      "loss": 1.7714,
+      "step": 930
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.84747422187526e-05,
+      "loss": 1.5569,
+      "step": 931
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8470994214601378e-05,
+      "loss": 1.6464,
+      "step": 932
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8467241992282842e-05,
+      "loss": 1.5306,
+      "step": 933
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.846348555366544e-05,
+      "loss": 1.5146,
+      "step": 934
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8459724900619705e-05,
+      "loss": 1.6297,
+      "step": 935
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.845596003501826e-05,
+      "loss": 1.4991,
+      "step": 936
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.845219095873585e-05,
+      "loss": 1.6013,
+      "step": 937
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8448417673649292e-05,
+      "loss": 1.6493,
+      "step": 938
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8444640181637514e-05,
+      "loss": 1.6812,
+      "step": 939
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.844085848458153e-05,
+      "loss": 1.4978,
+      "step": 940
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8437072584364456e-05,
+      "loss": 1.3872,
+      "step": 941
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8433282482871497e-05,
+      "loss": 1.6746,
+      "step": 942
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8429488181989946e-05,
+      "loss": 1.6504,
+      "step": 943
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8425689683609198e-05,
+      "loss": 1.6379,
+      "step": 944
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8421886989620724e-05,
+      "loss": 1.5317,
+      "step": 945
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8418080101918095e-05,
+      "loss": 1.9125,
+      "step": 946
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8414269022396967e-05,
+      "loss": 1.5284,
+      "step": 947
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.841045375295508e-05,
+      "loss": 1.7975,
+      "step": 948
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8406634295492266e-05,
+      "loss": 1.6434,
+      "step": 949
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8402810651910444e-05,
+      "loss": 1.457,
+      "step": 950
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.839898282411361e-05,
+      "loss": 1.6691,
+      "step": 951
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.839515081400784e-05,
+      "loss": 1.8307,
+      "step": 952
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8391314623501315e-05,
+      "loss": 1.5053,
+      "step": 953
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8387474254504265e-05,
+      "loss": 1.5939,
+      "step": 954
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.838362970892903e-05,
+      "loss": 1.6342,
+      "step": 955
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8379780988690014e-05,
+      "loss": 1.4749,
+      "step": 956
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8375928095703703e-05,
+      "loss": 1.7633,
+      "step": 957
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.837207103188866e-05,
+      "loss": 1.7029,
+      "step": 958
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.836820979916553e-05,
+      "loss": 1.4715,
+      "step": 959
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8364344399457028e-05,
+      "loss": 1.4499,
+      "step": 960
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8360474834687944e-05,
+      "loss": 1.5667,
+      "step": 961
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8356601106785148e-05,
+      "loss": 1.6502,
+      "step": 962
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.835272321767758e-05,
+      "loss": 1.6614,
+      "step": 963
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8348841169296247e-05,
+      "loss": 1.7366,
+      "step": 964
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8344954963574227e-05,
+      "loss": 1.5804,
+      "step": 965
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8341064602446686e-05,
+      "loss": 1.5135,
+      "step": 966
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8337170087850834e-05,
+      "loss": 1.5833,
+      "step": 967
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8333271421725968e-05,
+      "loss": 1.6302,
+      "step": 968
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.832936860601344e-05,
+      "loss": 1.4175,
+      "step": 969
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8325461642656676e-05,
+      "loss": 1.6856,
+      "step": 970
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8321550533601166e-05,
+      "loss": 1.4986,
+      "step": 971
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8317635280794466e-05,
+      "loss": 1.7103,
+      "step": 972
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8313715886186183e-05,
+      "loss": 1.6611,
+      "step": 973
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8309792351728006e-05,
+      "loss": 1.7768,
+      "step": 974
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8305864679373667e-05,
+      "loss": 1.4449,
+      "step": 975
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8301932871078975e-05,
+      "loss": 1.6403,
+      "step": 976
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8297996928801787e-05,
+      "loss": 1.8645,
+      "step": 977
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.829405685450202e-05,
+      "loss": 1.9234,
+      "step": 978
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8290112650141652e-05,
+      "loss": 1.6654,
+      "step": 979
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.828616431768472e-05,
+      "loss": 1.4603,
+      "step": 980
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8282211859097308e-05,
+      "loss": 1.2843,
+      "step": 981
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8278255276347563e-05,
+      "loss": 1.7198,
+      "step": 982
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8274294571405678e-05,
+      "loss": 1.6672,
+      "step": 983
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8270329746243903e-05,
+      "loss": 1.5261,
+      "step": 984
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8266360802836542e-05,
+      "loss": 1.6983,
+      "step": 985
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.826238774315995e-05,
+      "loss": 1.4945,
+      "step": 986
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8258410569192526e-05,
+      "loss": 1.4757,
+      "step": 987
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8254429282914715e-05,
+      "loss": 1.61,
+      "step": 988
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8250443886309023e-05,
+      "loss": 1.6111,
+      "step": 989
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.824645438135999e-05,
+      "loss": 1.7097,
+      "step": 990
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.824246077005421e-05,
+      "loss": 1.5042,
+      "step": 991
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.823846305438032e-05,
+      "loss": 1.4909,
+      "step": 992
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8234461236328992e-05,
+      "loss": 1.5205,
+      "step": 993
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8230455317892957e-05,
+      "loss": 1.6378,
+      "step": 994
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8226445301066974e-05,
+      "loss": 1.6407,
+      "step": 995
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.822243118784785e-05,
+      "loss": 1.7994,
+      "step": 996
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8218412980234426e-05,
+      "loss": 1.6106,
+      "step": 997
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8214390680227588e-05,
+      "loss": 1.8341,
+      "step": 998
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.821036428983026e-05,
+      "loss": 1.5347,
+      "step": 999
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.82063338110474e-05,
+      "loss": 1.5895,
+      "step": 1000
+    }
+  ],
+  "max_steps": 4590,
+  "num_train_epochs": 3,
+  "total_flos": 1.1673189840689562e+17,
+  "trial_name": null,
+  "trial_params": null
+}