diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5248 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 870,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.407407407407407e-07,
+      "loss": 0.8875,
+      "step": 1
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4814814814814815e-06,
+      "loss": 0.9001,
+      "step": 2
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 0.8933,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.962962962962963e-06,
+      "loss": 0.8382,
+      "step": 4
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7037037037037037e-06,
+      "loss": 0.8645,
+      "step": 5
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 0.8418,
+      "step": 6
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.185185185185185e-06,
+      "loss": 0.8209,
+      "step": 7
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.925925925925926e-06,
+      "loss": 0.8482,
+      "step": 8
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.7795,
+      "step": 9
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 0.7949,
+      "step": 10
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.148148148148148e-06,
+      "loss": 0.7894,
+      "step": 11
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 0.7611,
+      "step": 12
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.62962962962963e-06,
+      "loss": 0.7972,
+      "step": 13
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.037037037037037e-05,
+      "loss": 0.7677,
+      "step": 14
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.1111111111111113e-05,
+      "loss": 0.7112,
+      "step": 15
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1851851851851852e-05,
+      "loss": 0.7343,
+      "step": 16
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2592592592592593e-05,
+      "loss": 0.7336,
+      "step": 17
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.75,
+      "step": 18
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4074074074074075e-05,
+      "loss": 0.7225,
+      "step": 19
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4814814814814815e-05,
+      "loss": 0.7388,
+      "step": 20
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.555555555555556e-05,
+      "loss": 0.7575,
+      "step": 21
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6296296296296297e-05,
+      "loss": 0.708,
+      "step": 22
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.7037037037037038e-05,
+      "loss": 0.7119,
+      "step": 23
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.7777777777777777e-05,
+      "loss": 0.7272,
+      "step": 24
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.851851851851852e-05,
+      "loss": 0.704,
+      "step": 25
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.925925925925926e-05,
+      "loss": 0.7147,
+      "step": 26
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 0.7044,
+      "step": 27
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999930559298766e-05,
+      "loss": 0.7077,
+      "step": 28
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999972223815946e-05,
+      "loss": 0.6966,
+      "step": 29
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999375039475278e-05,
+      "loss": 0.7226,
+      "step": 30
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9998888968068158e-05,
+      "loss": 0.7271,
+      "step": 31
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9998264030688742e-05,
+      "loss": 0.6909,
+      "step": 32
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9997500236016233e-05,
+      "loss": 0.6988,
+      "step": 33
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9996597594658328e-05,
+      "loss": 0.7229,
+      "step": 34
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9995556119151032e-05,
+      "loss": 0.7141,
+      "step": 35
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9994375823958504e-05,
+      "loss": 0.6924,
+      "step": 36
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9993056725472846e-05,
+      "loss": 0.7206,
+      "step": 37
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9991598842013892e-05,
+      "loss": 0.7039,
+      "step": 38
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9990002193828923e-05,
+      "loss": 0.7018,
+      "step": 39
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.998826680309242e-05,
+      "loss": 0.702,
+      "step": 40
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9986392693905725e-05,
+      "loss": 0.7102,
+      "step": 41
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.998437989229673e-05,
+      "loss": 0.7018,
+      "step": 42
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9982228426219518e-05,
+      "loss": 0.7025,
+      "step": 43
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9979938325553937e-05,
+      "loss": 0.6791,
+      "step": 44
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9977509622105233e-05,
+      "loss": 0.7075,
+      "step": 45
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.997494234960358e-05,
+      "loss": 0.6959,
+      "step": 46
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.997223654370362e-05,
+      "loss": 0.6924,
+      "step": 47
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9969392241983957e-05,
+      "loss": 0.7094,
+      "step": 48
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9966409483946664e-05,
+      "loss": 0.7062,
+      "step": 49
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9963288311016695e-05,
+      "loss": 0.6836,
+      "step": 50
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9960028766541336e-05,
+      "loss": 0.7024,
+      "step": 51
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9956630895789604e-05,
+      "loss": 0.727,
+      "step": 52
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.99530947459516e-05,
+      "loss": 0.7025,
+      "step": 53
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.994942036613787e-05,
+      "loss": 0.6771,
+      "step": 54
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9945607807378717e-05,
+      "loss": 0.6974,
+      "step": 55
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9941657122623494e-05,
+      "loss": 0.6874,
+      "step": 56
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9937568366739858e-05,
+      "loss": 0.6956,
+      "step": 57
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9933341596513033e-05,
+      "loss": 0.6884,
+      "step": 58
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.992897687064499e-05,
+      "loss": 0.6885,
+      "step": 59
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9924474249753656e-05,
+      "loss": 0.6902,
+      "step": 60
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.991983379637206e-05,
+      "loss": 0.6952,
+      "step": 61
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.991505557494747e-05,
+      "loss": 0.7051,
+      "step": 62
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9910139651840497e-05,
+      "loss": 0.6926,
+      "step": 63
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.990508609532417e-05,
+      "loss": 0.7124,
+      "step": 64
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9899894975582997e-05,
+      "loss": 0.6998,
+      "step": 65
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9894566364711965e-05,
+      "loss": 0.6704,
+      "step": 66
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9889100336715575e-05,
+      "loss": 0.7083,
+      "step": 67
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9883496967506788e-05,
+      "loss": 0.678,
+      "step": 68
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9877756334905983e-05,
+      "loss": 0.6869,
+      "step": 69
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9871878518639875e-05,
+      "loss": 0.6569,
+      "step": 70
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9865863600340393e-05,
+      "loss": 0.6751,
+      "step": 71
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9859711663543573e-05,
+      "loss": 0.6961,
+      "step": 72
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.985342279368837e-05,
+      "loss": 0.6552,
+      "step": 73
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9846997078115493e-05,
+      "loss": 0.7056,
+      "step": 74
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9840434606066182e-05,
+      "loss": 0.6577,
+      "step": 75
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9833735468680967e-05,
+      "loss": 0.7007,
+      "step": 76
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.982689975899841e-05,
+      "loss": 0.6941,
+      "step": 77
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9819927571953804e-05,
+      "loss": 0.672,
+      "step": 78
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9812819004377862e-05,
+      "loss": 0.7093,
+      "step": 79
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.980557415499537e-05,
+      "loss": 0.6611,
+      "step": 80
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9798193124423804e-05,
+      "loss": 0.7045,
+      "step": 81
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.979067601517196e-05,
+      "loss": 0.6948,
+      "step": 82
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.97830229316385e-05,
+      "loss": 0.6792,
+      "step": 83
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9775233980110524e-05,
+      "loss": 0.6994,
+      "step": 84
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.976730926876208e-05,
+      "loss": 0.6815,
+      "step": 85
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.975924890765268e-05,
+      "loss": 0.6575,
+      "step": 86
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9751053008725736e-05,
+      "loss": 0.6996,
+      "step": 87
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9742721685807057e-05,
+      "loss": 0.6621,
+      "step": 88
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9734255054603212e-05,
+      "loss": 0.6772,
+      "step": 89
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9725653232699962e-05,
+      "loss": 0.6907,
+      "step": 90
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9716916339560626e-05,
+      "loss": 0.684,
+      "step": 91
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9708044496524394e-05,
+      "loss": 0.7,
+      "step": 92
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.969903782680467e-05,
+      "loss": 0.6888,
+      "step": 93
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9689896455487338e-05,
+      "loss": 0.6371,
+      "step": 94
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.968062050952906e-05,
+      "loss": 0.667,
+      "step": 95
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.967121011775546e-05,
+      "loss": 0.6658,
+      "step": 96
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.966166541085938e-05,
+      "loss": 0.6622,
+      "step": 97
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.965198652139906e-05,
+      "loss": 0.6877,
+      "step": 98
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9642173583796265e-05,
+      "loss": 0.6859,
+      "step": 99
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.963222673433445e-05,
+      "loss": 0.7151,
+      "step": 100
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9622146111156858e-05,
+      "loss": 0.6844,
+      "step": 101
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.961193185426459e-05,
+      "loss": 0.667,
+      "step": 102
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9601584105514683e-05,
+      "loss": 0.6579,
+      "step": 103
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9591103008618128e-05,
+      "loss": 0.6746,
+      "step": 104
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.958048870913786e-05,
+      "loss": 0.6807,
+      "step": 105
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9569741354486767e-05,
+      "loss": 0.6897,
+      "step": 106
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.955886109392561e-05,
+      "loss": 0.6934,
+      "step": 107
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9547848078560975e-05,
+      "loss": 0.6762,
+      "step": 108
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.953670246134317e-05,
+      "loss": 0.6695,
+      "step": 109
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9525424397064082e-05,
+      "loss": 0.6905,
+      "step": 110
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9514014042355057e-05,
+      "loss": 0.6596,
+      "step": 111
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9502471555684695e-05,
+      "loss": 0.6745,
+      "step": 112
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9490797097356674e-05,
+      "loss": 0.6531,
+      "step": 113
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9478990829507507e-05,
+      "loss": 0.6717,
+      "step": 114
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9467052916104297e-05,
+      "loss": 0.6315,
+      "step": 115
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.945498352294246e-05,
+      "loss": 0.6553,
+      "step": 116
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9442782817643425e-05,
+      "loss": 0.6512,
+      "step": 117
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.943045096965229e-05,
+      "loss": 0.6626,
+      "step": 118
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9417988150235496e-05,
+      "loss": 0.6495,
+      "step": 119
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9405394532478422e-05,
+      "loss": 0.6785,
+      "step": 120
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9392670291283005e-05,
+      "loss": 0.6899,
+      "step": 121
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9379815603365283e-05,
+      "loss": 0.6352,
+      "step": 122
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9366830647252974e-05,
+      "loss": 0.6567,
+      "step": 123
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.935371560328296e-05,
+      "loss": 0.6833,
+      "step": 124
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.934047065359881e-05,
+      "loss": 0.6589,
+      "step": 125
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9327095982148258e-05,
+      "loss": 0.6507,
+      "step": 126
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.93135917746806e-05,
+      "loss": 0.6725,
+      "step": 127
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.929995821874417e-05,
+      "loss": 0.66,
+      "step": 128
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.928619550368371e-05,
+      "loss": 0.6696,
+      "step": 129
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9272303820637727e-05,
+      "loss": 0.6732,
+      "step": 130
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9258283362535866e-05,
+      "loss": 0.6721,
+      "step": 131
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9244134324096223e-05,
+      "loss": 0.6602,
+      "step": 132
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.922985690182262e-05,
+      "loss": 0.6849,
+      "step": 133
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9215451294001902e-05,
+      "loss": 0.6797,
+      "step": 134
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9200917700701176e-05,
+      "loss": 0.6586,
+      "step": 135
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9186256323765014e-05,
+      "loss": 0.6822,
+      "step": 136
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9171467366812687e-05,
+      "loss": 0.6609,
+      "step": 137
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.915655103523529e-05,
+      "loss": 0.6821,
+      "step": 138
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.914150753619294e-05,
+      "loss": 0.6807,
+      "step": 139
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9126337078611855e-05,
+      "loss": 0.6516,
+      "step": 140
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9111039873181478e-05,
+      "loss": 0.6682,
+      "step": 141
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9095616132351545e-05,
+      "loss": 0.6375,
+      "step": 142
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9080066070329133e-05,
+      "loss": 0.6614,
+      "step": 143
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9064389903075676e-05,
+      "loss": 0.6633,
+      "step": 144
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9048587848303995e-05,
+      "loss": 0.6454,
+      "step": 145
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9032660125475238e-05,
+      "loss": 0.6768,
+      "step": 146
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.901660695579585e-05,
+      "loss": 0.6791,
+      "step": 147
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.900042856221451e-05,
+      "loss": 0.6799,
+      "step": 148
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.898412516941901e-05,
+      "loss": 0.658,
+      "step": 149
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8967697003833156e-05,
+      "loss": 0.6668,
+      "step": 150
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8951144293613623e-05,
+      "loss": 0.6641,
+      "step": 151
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.893446726864676e-05,
+      "loss": 0.6369,
+      "step": 152
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8917666160545446e-05,
+      "loss": 0.6798,
+      "step": 153
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8900741202645808e-05,
+      "loss": 0.6559,
+      "step": 154
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8883692630004046e-05,
+      "loss": 0.6695,
+      "step": 155
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8866520679393127e-05,
+      "loss": 0.6391,
+      "step": 156
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8849225589299508e-05,
+      "loss": 0.6635,
+      "step": 157
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8831807599919824e-05,
+      "loss": 0.6555,
+      "step": 158
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8814266953157557e-05,
+      "loss": 0.6683,
+      "step": 159
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8796603892619668e-05,
+      "loss": 0.6518,
+      "step": 160
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8778818663613217e-05,
+      "loss": 0.6165,
+      "step": 161
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.876091151314196e-05,
+      "loss": 0.6873,
+      "step": 162
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.874288268990292e-05,
+      "loss": 0.677,
+      "step": 163
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8724732444282917e-05,
+      "loss": 0.6483,
+      "step": 164
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8706461028355107e-05,
+      "loss": 0.6495,
+      "step": 165
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8688068695875477e-05,
+      "loss": 0.682,
+      "step": 166
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.866955570227932e-05,
+      "loss": 0.6506,
+      "step": 167
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.865092230467769e-05,
+      "loss": 0.6617,
+      "step": 168
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8632168761853827e-05,
+      "loss": 0.6407,
+      "step": 169
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.861329533425956e-05,
+      "loss": 0.6546,
+      "step": 170
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8594302284011704e-05,
+      "loss": 0.6515,
+      "step": 171
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8575189874888396e-05,
+      "loss": 0.6759,
+      "step": 172
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8555958372325456e-05,
+      "loss": 0.6676,
+      "step": 173
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.85366080434127e-05,
+      "loss": 0.6695,
+      "step": 174
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8517139156890196e-05,
+      "loss": 0.6647,
+      "step": 175
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8497551983144576e-05,
+      "loss": 0.636,
+      "step": 176
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8477846794205258e-05,
+      "loss": 0.6591,
+      "step": 177
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8458023863740667e-05,
+      "loss": 0.6613,
+      "step": 178
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8438083467054444e-05,
+      "loss": 0.6332,
+      "step": 179
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8418025881081612e-05,
+      "loss": 0.6661,
+      "step": 180
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.839785138438474e-05,
+      "loss": 0.6209,
+      "step": 181
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.837756025715006e-05,
+      "loss": 0.6876,
+      "step": 182
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8357152781183606e-05,
+      "loss": 0.6718,
+      "step": 183
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8336629239907255e-05,
+      "loss": 0.6477,
+      "step": 184
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8315989918354835e-05,
+      "loss": 0.6508,
+      "step": 185
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.829523510316813e-05,
+      "loss": 0.6494,
+      "step": 186
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8274365082592932e-05,
+      "loss": 0.6417,
+      "step": 187
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.825338014647501e-05,
+      "loss": 0.6745,
+      "step": 188
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.82322805862561e-05,
+      "loss": 0.6554,
+      "step": 189
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8211066694969855e-05,
+      "loss": 0.6445,
+      "step": 190
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8189738767237768e-05,
+      "loss": 0.6611,
+      "step": 191
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8168297099265094e-05,
+      "loss": 0.6561,
+      "step": 192
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8146741988836728e-05,
+      "loss": 0.6457,
+      "step": 193
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.812507373531306e-05,
+      "loss": 0.6381,
+      "step": 194
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.810329263962584e-05,
+      "loss": 0.6554,
+      "step": 195
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.8081399004273972e-05,
+      "loss": 0.6648,
+      "step": 196
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.805939313331934e-05,
+      "loss": 0.6563,
+      "step": 197
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.803727533238257e-05,
+      "loss": 0.6573,
+      "step": 198
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.801504590863877e-05,
+      "loss": 0.6864,
+      "step": 199
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7992705170813306e-05,
+      "loss": 0.649,
+      "step": 200
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7970253429177477e-05,
+      "loss": 0.6605,
+      "step": 201
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7947690995544217e-05,
+      "loss": 0.6444,
+      "step": 202
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7925018183263772e-05,
+      "loss": 0.6734,
+      "step": 203
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7902235307219333e-05,
+      "loss": 0.6564,
+      "step": 204
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.7879342683822683e-05,
+      "loss": 0.6558,
+      "step": 205
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.7856340631009784e-05,
+      "loss": 0.6572,
+      "step": 206
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.7833229468236367e-05,
+      "loss": 0.6278,
+      "step": 207
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.781000951647351e-05,
+      "loss": 0.631,
+      "step": 208
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.778668109820316e-05,
+      "loss": 0.6468,
+      "step": 209
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.776324453741365e-05,
+      "loss": 0.6491,
+      "step": 210
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7739700159595236e-05,
+      "loss": 0.6562,
+      "step": 211
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7716048291735534e-05,
+      "loss": 0.6611,
+      "step": 212
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7692289262315e-05,
+      "loss": 0.6484,
+      "step": 213
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7668423401302372e-05,
+      "loss": 0.6434,
+      "step": 214
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.764445104015007e-05,
+      "loss": 0.6358,
+      "step": 215
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7620372511789607e-05,
+      "loss": 0.6532,
+      "step": 216
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.7596188150626964e-05,
+      "loss": 0.6462,
+      "step": 217
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.7571898292537944e-05,
+      "loss": 0.6561,
+      "step": 218
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.75475032748635e-05,
+      "loss": 0.6509,
+      "step": 219
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.7523003436405055e-05,
+      "loss": 0.6572,
+      "step": 220
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.749839911741981e-05,
+      "loss": 0.6723,
+      "step": 221
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.747369065961599e-05,
+      "loss": 0.6534,
+      "step": 222
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.7448878406148122e-05,
+      "loss": 0.6338,
+      "step": 223
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.7423962701612268e-05,
+      "loss": 0.6636,
+      "step": 224
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7398943892041223e-05,
+      "loss": 0.661,
+      "step": 225
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7373822324899724e-05,
+      "loss": 0.6592,
+      "step": 226
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7348598349079618e-05,
+      "loss": 0.6575,
+      "step": 227
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7323272314895022e-05,
+      "loss": 0.6465,
+      "step": 228
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7297844574077447e-05,
+      "loss": 0.6513,
+      "step": 229
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7272315479770925e-05,
+      "loss": 0.6443,
+      "step": 230
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7246685386527098e-05,
+      "loss": 0.6521,
+      "step": 231
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7220954650300293e-05,
+      "loss": 0.6566,
+      "step": 232
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.719512362844259e-05,
+      "loss": 0.6599,
+      "step": 233
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7169192679698837e-05,
+      "loss": 0.6464,
+      "step": 234
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.714316216420169e-05,
+      "loss": 0.6291,
+      "step": 235
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7117032443466593e-05,
+      "loss": 0.6723,
+      "step": 236
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7090803880386784e-05,
+      "loss": 0.647,
+      "step": 237
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7064476839228213e-05,
+      "loss": 0.6645,
+      "step": 238
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.703805168562453e-05,
+      "loss": 0.6333,
+      "step": 239
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.701152878657197e-05,
+      "loss": 0.6281,
+      "step": 240
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.6984908510424272e-05,
+      "loss": 0.6167,
+      "step": 241
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.695819122688757e-05,
+      "loss": 0.6484,
+      "step": 242
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.693137730701524e-05,
+      "loss": 0.678,
+      "step": 243
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6904467123202762e-05,
+      "loss": 0.6557,
+      "step": 244
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.687746104918255e-05,
+      "loss": 0.6528,
+      "step": 245
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6850359460018737e-05,
+      "loss": 0.624,
+      "step": 246
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6823162732102002e-05,
+      "loss": 0.6407,
+      "step": 247
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.679587124314431e-05,
+      "loss": 0.6495,
+      "step": 248
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.6768485372173696e-05,
+      "loss": 0.6313,
+      "step": 249
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.674100549952897e-05,
+      "loss": 0.6578,
+      "step": 250
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.671343200685446e-05,
+      "loss": 0.6448,
+      "step": 251
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6685765277094702e-05,
+      "loss": 0.6423,
+      "step": 252
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.665800569448911e-05,
+      "loss": 0.668,
+      "step": 253
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.663015364456667e-05,
+      "loss": 0.6554,
+      "step": 254
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6602209514140552e-05,
+      "loss": 0.6569,
+      "step": 255
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.657417369130276e-05,
+      "loss": 0.6464,
+      "step": 256
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6546046565418734e-05,
+      "loss": 0.6534,
+      "step": 257
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6517828527121942e-05,
+      "loss": 0.6583,
+      "step": 258
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.648951996830846e-05,
+      "loss": 0.6354,
+      "step": 259
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6461121282131518e-05,
+      "loss": 0.6341,
+      "step": 260
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6432632862996056e-05,
+      "loss": 0.6481,
+      "step": 261
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6404055106553236e-05,
+      "loss": 0.6414,
+      "step": 262
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6375388409694942e-05,
+      "loss": 0.6423,
+      "step": 263
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.634663317054829e-05,
+      "loss": 0.6433,
+      "step": 264
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6317789788470063e-05,
+      "loss": 0.6434,
+      "step": 265
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.62888586640412e-05,
+      "loss": 0.6203,
+      "step": 266
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6259840199061215e-05,
+      "loss": 0.6372,
+      "step": 267
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6230734796542628e-05,
+      "loss": 0.6167,
+      "step": 268
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.620154286070534e-05,
+      "loss": 0.6569,
+      "step": 269
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.617226479697105e-05,
+      "loss": 0.6411,
+      "step": 270
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6142901011957615e-05,
+      "loss": 0.6556,
+      "step": 271
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.611345191347341e-05,
+      "loss": 0.6574,
+      "step": 272
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.608391791051163e-05,
+      "loss": 0.6435,
+      "step": 273
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6054299413244655e-05,
+      "loss": 0.6384,
+      "step": 274
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.6024596833018335e-05,
+      "loss": 0.6288,
+      "step": 275
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.599481058234626e-05,
+      "loss": 0.6172,
+      "step": 276
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.5964941074904064e-05,
+      "loss": 0.6401,
+      "step": 277
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.5934988725523655e-05,
+      "loss": 0.6363,
+      "step": 278
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.5904953950187458e-05,
+      "loss": 0.6343,
+      "step": 279
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.587483716602265e-05,
+      "loss": 0.6223,
+      "step": 280
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5844638791295348e-05,
+      "loss": 0.6508,
+      "step": 281
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5814359245404818e-05,
+      "loss": 0.6267,
+      "step": 282
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5783998948877642e-05,
+      "loss": 0.6133,
+      "step": 283
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.575355832336187e-05,
+      "loss": 0.6556,
+      "step": 284
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5723037791621193e-05,
+      "loss": 0.6492,
+      "step": 285
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.569243777752902e-05,
+      "loss": 0.634,
+      "step": 286
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.566175870606264e-05,
+      "loss": 0.6598,
+      "step": 287
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.563100100329731e-05,
+      "loss": 0.642,
+      "step": 288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.560016509640031e-05,
+      "loss": 0.6429,
+      "step": 289
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5569251413625043e-05,
+      "loss": 0.6209,
+      "step": 290
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5538260384305076e-05,
+      "loss": 0.4244,
+      "step": 291
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5507192438848164e-05,
+      "loss": 0.4075,
+      "step": 292
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.547604800873029e-05,
+      "loss": 0.3895,
+      "step": 293
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5444827526489675e-05,
+      "loss": 0.3953,
+      "step": 294
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5413531425720747e-05,
+      "loss": 0.3954,
+      "step": 295
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5382160141068147e-05,
+      "loss": 0.3929,
+      "step": 296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5350714108220673e-05,
+      "loss": 0.3973,
+      "step": 297
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5319193763905247e-05,
+      "loss": 0.4219,
+      "step": 298
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.528759954588082e-05,
+      "loss": 0.3942,
+      "step": 299
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5255931892932333e-05,
+      "loss": 0.3954,
+      "step": 300
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5224191244864583e-05,
+      "loss": 0.4059,
+      "step": 301
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5192378042496148e-05,
+      "loss": 0.387,
+      "step": 302
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5160492727653241e-05,
+      "loss": 0.3922,
+      "step": 303
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.512853574316359e-05,
+      "loss": 0.4027,
+      "step": 304
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5096507532850275e-05,
+      "loss": 0.4016,
+      "step": 305
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.5064408541525573e-05,
+      "loss": 0.403,
+      "step": 306
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.5032239214984773e-05,
+      "loss": 0.3767,
+      "step": 307
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.3907,
+      "step": 308
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4967691344313995e-05,
+      "loss": 0.3833,
+      "step": 309
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.493531369663389e-05,
+      "loss": 0.3966,
+      "step": 310
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4902867506625002e-05,
+      "loss": 0.3824,
+      "step": 311
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4870353224904572e-05,
+      "loss": 0.3703,
+      "step": 312
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.48377713030355e-05,
+      "loss": 0.4073,
+      "step": 313
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4805122193520088e-05,
+      "loss": 0.4308,
+      "step": 314
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4772406349793744e-05,
+      "loss": 0.4018,
+      "step": 315
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4739624226218701e-05,
+      "loss": 0.3731,
+      "step": 316
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4706776278077681e-05,
+      "loss": 0.3876,
+      "step": 317
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4673862961567602e-05,
+      "loss": 0.3862,
+      "step": 318
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4640884733793216e-05,
+      "loss": 0.398,
+      "step": 319
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4607842052760777e-05,
+      "loss": 0.3783,
+      "step": 320
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.457473537737167e-05,
+      "loss": 0.3871,
+      "step": 321
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.454156516741605e-05,
+      "loss": 0.386,
+      "step": 322
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4508331883566445e-05,
+      "loss": 0.3768,
+      "step": 323
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4475035987371355e-05,
+      "loss": 0.4102,
+      "step": 324
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4441677941248862e-05,
+      "loss": 0.3924,
+      "step": 325
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.440825820848019e-05,
+      "loss": 0.3906,
+      "step": 326
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4374777253203273e-05,
+      "loss": 0.3799,
+      "step": 327
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4341235540406314e-05,
+      "loss": 0.3914,
+      "step": 328
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4307633535921319e-05,
+      "loss": 0.3819,
+      "step": 329
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4273971706417648e-05,
+      "loss": 0.3754,
+      "step": 330
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4240250519395504e-05,
+      "loss": 0.3792,
+      "step": 331
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4206470443179466e-05,
+      "loss": 0.402,
+      "step": 332
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4172631946911964e-05,
+      "loss": 0.3916,
+      "step": 333
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4138735500546781e-05,
+      "loss": 0.3739,
+      "step": 334
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4104781574842524e-05,
+      "loss": 0.3731,
+      "step": 335
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.407077064135607e-05,
+      "loss": 0.3842,
+      "step": 336
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4036703172436037e-05,
+      "loss": 0.4027,
+      "step": 337
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.4002579641216207e-05,
+      "loss": 0.3844,
+      "step": 338
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3968400521608969e-05,
+      "loss": 0.4135,
+      "step": 339
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3934166288298729e-05,
+      "loss": 0.4006,
+      "step": 340
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3899877416735322e-05,
+      "loss": 0.3826,
+      "step": 341
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3865534383127406e-05,
+      "loss": 0.388,
+      "step": 342
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3831137664435846e-05,
+      "loss": 0.3848,
+      "step": 343
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3796687738367095e-05,
+      "loss": 0.3857,
+      "step": 344
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3762185083366557e-05,
+      "loss": 0.4133,
+      "step": 345
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3727630178611948e-05,
+      "loss": 0.3842,
+      "step": 346
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3693023504006629e-05,
+      "loss": 0.3872,
+      "step": 347
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3658365540172948e-05,
+      "loss": 0.3829,
+      "step": 348
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3623656768445572e-05,
+      "loss": 0.3966,
+      "step": 349
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3588897670864787e-05,
+      "loss": 0.3855,
+      "step": 350
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3554088730169814e-05,
+      "loss": 0.4032,
+      "step": 351
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3519230429792107e-05,
+      "loss": 0.3778,
+      "step": 352
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.348432325384863e-05,
+      "loss": 0.3827,
+      "step": 353
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3449367687135134e-05,
+      "loss": 0.3737,
+      "step": 354
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3414364215119436e-05,
+      "loss": 0.4059,
+      "step": 355
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3379313323934664e-05,
+      "loss": 0.3931,
+      "step": 356
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.334421550037251e-05,
+      "loss": 0.3639,
+      "step": 357
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3309071231876469e-05,
+      "loss": 0.3757,
+      "step": 358
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3273881006535074e-05,
+      "loss": 0.3772,
+      "step": 359
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3238645313075104e-05,
+      "loss": 0.4002,
+      "step": 360
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3203364640854819e-05,
+      "loss": 0.3933,
+      "step": 361
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3168039479857138e-05,
+      "loss": 0.396,
+      "step": 362
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.313267032068285e-05,
+      "loss": 0.3961,
+      "step": 363
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.30972576545438e-05,
+      "loss": 0.4078,
+      "step": 364
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3061801973256066e-05,
+      "loss": 0.4048,
+      "step": 365
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3026303769233112e-05,
+      "loss": 0.4148,
+      "step": 366
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2990763535478986e-05,
+      "loss": 0.3872,
+      "step": 367
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2955181765581428e-05,
+      "loss": 0.4021,
+      "step": 368
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2919558953705055e-05,
+      "loss": 0.394,
+      "step": 369
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2883895594584474e-05,
+      "loss": 0.4144,
+      "step": 370
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2848192183517415e-05,
+      "loss": 0.4045,
+      "step": 371
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2812449216357863e-05,
+      "loss": 0.3992,
+      "step": 372
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2776667189509155e-05,
+      "loss": 0.3934,
+      "step": 373
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.27408465999171e-05,
+      "loss": 0.3903,
+      "step": 374
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.270498794506307e-05,
+      "loss": 0.389,
+      "step": 375
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.266909172295709e-05,
+      "loss": 0.3722,
+      "step": 376
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.263315843213093e-05,
+      "loss": 0.4058,
+      "step": 377
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.259718857163117e-05,
+      "loss": 0.3984,
+      "step": 378
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.256118264101228e-05,
+      "loss": 0.383,
+      "step": 379
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2525141140329673e-05,
+      "loss": 0.4,
+      "step": 380
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2489064570132764e-05,
+      "loss": 0.3781,
+      "step": 381
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2452953431458021e-05,
+      "loss": 0.3964,
+      "step": 382
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2416808225822004e-05,
+      "loss": 0.4305,
+      "step": 383
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2380629455214392e-05,
+      "loss": 0.3889,
+      "step": 384
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2344417622091032e-05,
+      "loss": 0.3863,
+      "step": 385
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.230817322936694e-05,
+      "loss": 0.3903,
+      "step": 386
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2271896780409321e-05,
+      "loss": 0.3982,
+      "step": 387
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2235588779030587e-05,
+      "loss": 0.3915,
+      "step": 388
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2199249729481356e-05,
+      "loss": 0.3944,
+      "step": 389
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2162880136443447e-05,
+      "loss": 0.401,
+      "step": 390
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2126480505022861e-05,
+      "loss": 0.4051,
+      "step": 391
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.209005134074279e-05,
+      "loss": 0.3936,
+      "step": 392
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.2053593149536576e-05,
+      "loss": 0.3585,
+      "step": 393
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.2017106437740688e-05,
+      "loss": 0.3726,
+      "step": 394
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.19805917120877e-05,
+      "loss": 0.388,
+      "step": 395
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.1944049479699244e-05,
+      "loss": 0.4049,
+      "step": 396
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.1907480248078965e-05,
+      "loss": 0.3855,
+      "step": 397
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.187088452510548e-05,
+      "loss": 0.3706,
+      "step": 398
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1834262819025326e-05,
+      "loss": 0.3846,
+      "step": 399
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.179761563844589e-05,
+      "loss": 0.3816,
+      "step": 400
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1760943492328358e-05,
+      "loss": 0.3999,
+      "step": 401
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1724246889980638e-05,
+      "loss": 0.3882,
+      "step": 402
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1687526341050287e-05,
+      "loss": 0.385,
+      "step": 403
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.165078235551744e-05,
+      "loss": 0.3706,
+      "step": 404
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1614015443687723e-05,
+      "loss": 0.395,
+      "step": 405
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1577226116185162e-05,
+      "loss": 0.3986,
+      "step": 406
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1540414883945099e-05,
+      "loss": 0.3999,
+      "step": 407
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.150358225820709e-05,
+      "loss": 0.3849,
+      "step": 408
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1466728750507805e-05,
+      "loss": 0.3934,
+      "step": 409
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1429854872673927e-05,
+      "loss": 0.3943,
+      "step": 410
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1392961136815046e-05,
+      "loss": 0.3928,
+      "step": 411
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1356048055316538e-05,
+      "loss": 0.3829,
+      "step": 412
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1319116140832452e-05,
+      "loss": 0.392,
+      "step": 413
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1282165906278402e-05,
+      "loss": 0.4141,
+      "step": 414
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1245197864824426e-05,
+      "loss": 0.4017,
+      "step": 415
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1208212529887869e-05,
+      "loss": 0.3996,
+      "step": 416
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1171210415126248e-05,
+      "loss": 0.398,
+      "step": 417
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1134192034430123e-05,
+      "loss": 0.3795,
+      "step": 418
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1097157901915954e-05,
+      "loss": 0.3887,
+      "step": 419
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.1060108531918972e-05,
+      "loss": 0.3746,
+      "step": 420
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.102304443898602e-05,
+      "loss": 0.3891,
+      "step": 421
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.098596613786842e-05,
+      "loss": 0.3797,
+      "step": 422
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.094887414351482e-05,
+      "loss": 0.3978,
+      "step": 423
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0911768971064036e-05,
+      "loss": 0.3972,
+      "step": 424
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0874651135837912e-05,
+      "loss": 0.3999,
+      "step": 425
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0837521153334143e-05,
+      "loss": 0.3944,
+      "step": 426
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0800379539219136e-05,
+      "loss": 0.3905,
+      "step": 427
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.076322680932084e-05,
+      "loss": 0.378,
+      "step": 428
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0726063479621574e-05,
+      "loss": 0.3914,
+      "step": 429
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0688890066250876e-05,
+      "loss": 0.4029,
+      "step": 430
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0651707085478322e-05,
+      "loss": 0.3765,
+      "step": 431
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0614515053706367e-05,
+      "loss": 0.405,
+      "step": 432
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.057731448746316e-05,
+      "loss": 0.39,
+      "step": 433
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.0540105903395383e-05,
+      "loss": 0.3945,
+      "step": 434
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.0502889818261075e-05,
+      "loss": 0.372,
+      "step": 435
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.0465666748922438e-05,
+      "loss": 0.3839,
+      "step": 436
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0428437212338686e-05,
+      "loss": 0.4005,
+      "step": 437
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0391201725558842e-05,
+      "loss": 0.3842,
+      "step": 438
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0353960805714567e-05,
+      "loss": 0.3813,
+      "step": 439
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.031671497001298e-05,
+      "loss": 0.3843,
+      "step": 440
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0279464735729472e-05,
+      "loss": 0.4104,
+      "step": 441
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.024221062020052e-05,
+      "loss": 0.3968,
+      "step": 442
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0204953140816505e-05,
+      "loss": 0.3805,
+      "step": 443
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0167692815014527e-05,
+      "loss": 0.4006,
+      "step": 444
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0130430160271215e-05,
+      "loss": 0.3819,
+      "step": 445
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.0093165694095546e-05,
+      "loss": 0.3903,
+      "step": 446
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.0055899934021649e-05,
+      "loss": 0.3904,
+      "step": 447
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.0018633397601632e-05,
+      "loss": 0.3809,
+      "step": 448
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.98136660239837e-06,
+      "loss": 0.3817,
+      "step": 449
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.944100065978351e-06,
+      "loss": 0.3914,
+      "step": 450
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.906834305904456e-06,
+      "loss": 0.3731,
+      "step": 451
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.869569839728785e-06,
+      "loss": 0.4,
+      "step": 452
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.832307184985475e-06,
+      "loss": 0.374,
+      "step": 453
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.795046859183495e-06,
+      "loss": 0.3845,
+      "step": 454
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.757789379799482e-06,
+      "loss": 0.3969,
+      "step": 455
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.720535264270529e-06,
+      "loss": 0.403,
+      "step": 456
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.683285029987021e-06,
+      "loss": 0.3919,
+      "step": 457
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.646039194285437e-06,
+      "loss": 0.3939,
+      "step": 458
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.60879827444116e-06,
+      "loss": 0.3924,
+      "step": 459
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.571562787661316e-06,
+      "loss": 0.4056,
+      "step": 460
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.534333251077564e-06,
+      "loss": 0.381,
+      "step": 461
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.497110181738928e-06,
+      "loss": 0.4035,
+      "step": 462
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.459894096604618e-06,
+      "loss": 0.3812,
+      "step": 463
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.422685512536844e-06,
+      "loss": 0.3749,
+      "step": 464
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.385484946293636e-06,
+      "loss": 0.3956,
+      "step": 465
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.348292914521681e-06,
+      "loss": 0.3932,
+      "step": 466
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.311109933749127e-06,
+      "loss": 0.3799,
+      "step": 467
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.273936520378428e-06,
+      "loss": 0.3853,
+      "step": 468
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.236773190679163e-06,
+      "loss": 0.3836,
+      "step": 469
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.199620460780867e-06,
+      "loss": 0.3972,
+      "step": 470
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.16247884666586e-06,
+      "loss": 0.4047,
+      "step": 471
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.125348864162093e-06,
+      "loss": 0.3944,
+      "step": 472
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.088231028935965e-06,
+      "loss": 0.3889,
+      "step": 473
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.051125856485183e-06,
+      "loss": 0.3978,
+      "step": 474
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.014033862131583e-06,
+      "loss": 0.3909,
+      "step": 475
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 8.976955561013984e-06,
+      "loss": 0.3976,
+      "step": 476
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 8.939891468081033e-06,
+      "loss": 0.3943,
+      "step": 477
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.902842098084047e-06,
+      "loss": 0.3819,
+      "step": 478
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.865807965569879e-06,
+      "loss": 0.3843,
+      "step": 479
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.828789584873754e-06,
+      "loss": 0.3929,
+      "step": 480
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.791787470112133e-06,
+      "loss": 0.3779,
+      "step": 481
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.754802135175574e-06,
+      "loss": 0.3781,
+      "step": 482
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.717834093721598e-06,
+      "loss": 0.3973,
+      "step": 483
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.680883859167548e-06,
+      "loss": 0.3961,
+      "step": 484
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.643951944683466e-06,
+      "loss": 0.3879,
+      "step": 485
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.607038863184957e-06,
+      "loss": 0.3715,
+      "step": 486
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.570145127326075e-06,
+      "loss": 0.393,
+      "step": 487
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.533271249492198e-06,
+      "loss": 0.3794,
+      "step": 488
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.496417741792912e-06,
+      "loss": 0.3799,
+      "step": 489
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.459585116054903e-06,
+      "loss": 0.3844,
+      "step": 490
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.422773883814841e-06,
+      "loss": 0.3941,
+      "step": 491
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.385984556312282e-06,
+      "loss": 0.391,
+      "step": 492
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.349217644482563e-06,
+      "loss": 0.4106,
+      "step": 493
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.312473658949717e-06,
+      "loss": 0.4004,
+      "step": 494
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.275753110019367e-06,
+      "loss": 0.369,
+      "step": 495
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.239056507671645e-06,
+      "loss": 0.3954,
+      "step": 496
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.202384361554112e-06,
+      "loss": 0.3871,
+      "step": 497
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.165737180974678e-06,
+      "loss": 0.3811,
+      "step": 498
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.129115474894523e-06,
+      "loss": 0.3912,
+      "step": 499
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.09251975192104e-06,
+      "loss": 0.3737,
+      "step": 500
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.05595052030076e-06,
+      "loss": 0.3901,
+      "step": 501
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.019408287912304e-06,
+      "loss": 0.4104,
+      "step": 502
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 7.982893562259317e-06,
+      "loss": 0.4047,
+      "step": 503
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.94640685046343e-06,
+      "loss": 0.4104,
+      "step": 504
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.909948659257213e-06,
+      "loss": 0.3734,
+      "step": 505
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.873519494977144e-06,
+      "loss": 0.3961,
+      "step": 506
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.837119863556554e-06,
+      "loss": 0.3923,
+      "step": 507
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.800750270518644e-06,
+      "loss": 0.3824,
+      "step": 508
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.764411220969413e-06,
+      "loss": 0.4126,
+      "step": 509
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.72810321959068e-06,
+      "loss": 0.3915,
+      "step": 510
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.691826770633063e-06,
+      "loss": 0.3775,
+      "step": 511
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.65558237790897e-06,
+      "loss": 0.3723,
+      "step": 512
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.619370544785608e-06,
+      "loss": 0.4012,
+      "step": 513
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.583191774177999e-06,
+      "loss": 0.3672,
+      "step": 514
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.54704656854198e-06,
+      "loss": 0.3813,
+      "step": 515
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.510935429867237e-06,
+      "loss": 0.3933,
+      "step": 516
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.47485885967033e-06,
+      "loss": 0.4066,
+      "step": 517
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.438817358987722e-06,
+      "loss": 0.3866,
+      "step": 518
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.402811428368832e-06,
+      "loss": 0.3859,
+      "step": 519
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.3668415678690745e-06,
+      "loss": 0.3862,
+      "step": 520
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.330908277042914e-06,
+      "loss": 0.3837,
+      "step": 521
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.295012054936934e-06,
+      "loss": 0.3841,
+      "step": 522
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.259153400082904e-06,
+      "loss": 0.4005,
+      "step": 523
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.223332810490848e-06,
+      "loss": 0.4087,
+      "step": 524
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.187550783642141e-06,
+      "loss": 0.3907,
+      "step": 525
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.151807816482588e-06,
+      "loss": 0.3783,
+      "step": 526
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.116104405415529e-06,
+      "loss": 0.3917,
+      "step": 527
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.080441046294948e-06,
+      "loss": 0.3955,
+      "step": 528
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.0448182344185755e-06,
+      "loss": 0.3671,
+      "step": 529
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.009236464521021e-06,
+      "loss": 0.3937,
+      "step": 530
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.973696230766891e-06,
+      "loss": 0.3964,
+      "step": 531
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.938198026743941e-06,
+      "loss": 0.4003,
+      "step": 532
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.902742345456204e-06,
+      "loss": 0.3891,
+      "step": 533
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.8673296793171555e-06,
+      "loss": 0.369,
+      "step": 534
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.831960520142868e-06,
+      "loss": 0.3655,
+      "step": 535
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.796635359145183e-06,
+      "loss": 0.3867,
+      "step": 536
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.761354686924895e-06,
+      "loss": 0.3595,
+      "step": 537
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.7261189934649275e-06,
+      "loss": 0.3889,
+      "step": 538
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.69092876812353e-06,
+      "loss": 0.372,
+      "step": 539
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.655784499627491e-06,
+      "loss": 0.3687,
+      "step": 540
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.620686676065338e-06,
+      "loss": 0.3725,
+      "step": 541
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.585635784880566e-06,
+      "loss": 0.3673,
+      "step": 542
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.550632312864869e-06,
+      "loss": 0.3794,
+      "step": 543
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.5156767461513735e-06,
+      "loss": 0.3808,
+      "step": 544
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.480769570207897e-06,
+      "loss": 0.3846,
+      "step": 545
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.445911269830189e-06,
+      "loss": 0.3764,
+      "step": 546
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.411102329135216e-06,
+      "loss": 0.3721,
+      "step": 547
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.376343231554431e-06,
+      "loss": 0.3649,
+      "step": 548
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.341634459827053e-06,
+      "loss": 0.3876,
+      "step": 549
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.306976495993373e-06,
+      "loss": 0.395,
+      "step": 550
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.272369821388054e-06,
+      "loss": 0.389,
+      "step": 551
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.237814916633444e-06,
+      "loss": 0.389,
+      "step": 552
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.203312261632908e-06,
+      "loss": 0.3687,
+      "step": 553
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.1688623355641584e-06,
+      "loss": 0.3875,
+      "step": 554
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.134465616872598e-06,
+      "loss": 0.372,
+      "step": 555
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.1001225832646806e-06,
+      "loss": 0.3813,
+      "step": 556
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.065833711701274e-06,
+      "loss": 0.4067,
+      "step": 557
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.0315994783910345e-06,
+      "loss": 0.3742,
+      "step": 558
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.9974203587837985e-06,
+      "loss": 0.386,
+      "step": 559
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.963296827563969e-06,
+      "loss": 0.3815,
+      "step": 560
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.929229358643932e-06,
+      "loss": 0.4034,
+      "step": 561
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.895218425157479e-06,
+      "loss": 0.3774,
+      "step": 562
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.861264499453221e-06,
+      "loss": 0.4018,
+      "step": 563
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.827368053088043e-06,
+      "loss": 0.3788,
+      "step": 564
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.793529556820538e-06,
+      "loss": 0.3851,
+      "step": 565
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.759749480604499e-06,
+      "loss": 0.3823,
+      "step": 566
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.726028293582355e-06,
+      "loss": 0.3898,
+      "step": 567
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.692366464078683e-06,
+      "loss": 0.3833,
+      "step": 568
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.658764459593691e-06,
+      "loss": 0.3852,
+      "step": 569
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.62522274679673e-06,
+      "loss": 0.3718,
+      "step": 570
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.591741791519812e-06,
+      "loss": 0.3816,
+      "step": 571
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.55832205875114e-06,
+      "loss": 0.365,
+      "step": 572
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.524964012628648e-06,
+      "loss": 0.371,
+      "step": 573
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.4916681164335575e-06,
+      "loss": 0.3937,
+      "step": 574
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.45843483258395e-06,
+      "loss": 0.3626,
+      "step": 575
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.42526462262833e-06,
+      "loss": 0.3867,
+      "step": 576
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.392157947239224e-06,
+      "loss": 0.3888,
+      "step": 577
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.359115266206785e-06,
+      "loss": 0.3816,
+      "step": 578
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.326137038432399e-06,
+      "loss": 0.3826,
+      "step": 579
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.293223721922319e-06,
+      "loss": 0.3723,
+      "step": 580
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.260375773781302e-06,
+      "loss": 0.2175,
+      "step": 581
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.227593650206258e-06,
+      "loss": 0.2223,
+      "step": 582
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.194877806479916e-06,
+      "loss": 0.2026,
+      "step": 583
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.162228696964504e-06,
+      "loss": 0.2036,
+      "step": 584
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.129646775095432e-06,
+      "loss": 0.1993,
+      "step": 585
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.097132493375e-06,
+      "loss": 0.2149,
+      "step": 586
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.064686303366113e-06,
+      "loss": 0.2141,
+      "step": 587
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.032308655686011e-06,
+      "loss": 0.2261,
+      "step": 588
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.000000000000003e-06,
+      "loss": 0.2082,
+      "step": 589
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 4.96776078501523e-06,
+      "loss": 0.2224,
+      "step": 590
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.935591458474433e-06,
+      "loss": 0.215,
+      "step": 591
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.90349246714973e-06,
+      "loss": 0.2069,
+      "step": 592
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.871464256836414e-06,
+      "loss": 0.1989,
+      "step": 593
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.8395072723467585e-06,
+      "loss": 0.2022,
+      "step": 594
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.807621957503853e-06,
+      "loss": 0.2134,
+      "step": 595
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.775808755135418e-06,
+      "loss": 0.2254,
+      "step": 596
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.74406810706767e-06,
+      "loss": 0.213,
+      "step": 597
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.712400454119181e-06,
+      "loss": 0.2005,
+      "step": 598
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.680806236094756e-06,
+      "loss": 0.2121,
+      "step": 599
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.649285891779327e-06,
+      "loss": 0.2209,
+      "step": 600
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.617839858931855e-06,
+      "loss": 0.2037,
+      "step": 601
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.5864685742792555e-06,
+      "loss": 0.182,
+      "step": 602
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.5551724735103285e-06,
+      "loss": 0.212,
+      "step": 603
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.523951991269712e-06,
+      "loss": 0.1969,
+      "step": 604
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.492807561151841e-06,
+      "loss": 0.2178,
+      "step": 605
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.461739615694929e-06,
+      "loss": 0.2075,
+      "step": 606
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.4307485863749566e-06,
+      "loss": 0.2073,
+      "step": 607
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.399834903599692e-06,
+      "loss": 0.1934,
+      "step": 608
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.368998996702694e-06,
+      "loss": 0.1983,
+      "step": 609
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.338241293937362e-06,
+      "loss": 0.2127,
+      "step": 610
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.307562222470985e-06,
+      "loss": 0.2045,
+      "step": 611
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.276962208378811e-06,
+      "loss": 0.2069,
+      "step": 612
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.246441676638128e-06,
+      "loss": 0.2015,
+      "step": 613
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.2160010511223595e-06,
+      "loss": 0.1905,
+      "step": 614
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.185640754595183e-06,
+      "loss": 0.2114,
+      "step": 615
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.155361208704654e-06,
+      "loss": 0.1984,
+      "step": 616
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.125162833977354e-06,
+      "loss": 0.1895,
+      "step": 617
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.095046049812545e-06,
+      "loss": 0.2022,
+      "step": 618
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.065011274476349e-06,
+      "loss": 0.1976,
+      "step": 619
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.03505892509594e-06,
+      "loss": 0.2049,
+      "step": 620
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.005189417653743e-06,
+      "loss": 0.2072,
+      "step": 621
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 3.975403166981671e-06,
+      "loss": 0.217,
+      "step": 622
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.9457005867553454e-06,
+      "loss": 0.1999,
+      "step": 623
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.916082089488372e-06,
+      "loss": 0.1898,
+      "step": 624
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.886548086526594e-06,
+      "loss": 0.2055,
+      "step": 625
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.857098988042384e-06,
+      "loss": 0.2165,
+      "step": 626
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.827735203028953e-06,
+      "loss": 0.2054,
+      "step": 627
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.7984571392946643e-06,
+      "loss": 0.1916,
+      "step": 628
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.769265203457375e-06,
+      "loss": 0.2208,
+      "step": 629
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.740159800938784e-06,
+      "loss": 0.1872,
+      "step": 630
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.7111413359588024e-06,
+      "loss": 0.208,
+      "step": 631
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.6822102115299406e-06,
+      "loss": 0.205,
+      "step": 632
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.6533668294517154e-06,
+      "loss": 0.1824,
+      "step": 633
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.6246115903050592e-06,
+      "loss": 0.2104,
+      "step": 634
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.595944893446768e-06,
+      "loss": 0.1995,
+      "step": 635
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.5673671370039464e-06,
+      "loss": 0.1975,
+      "step": 636
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.5388787178684856e-06,
+      "loss": 0.2058,
+      "step": 637
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.5104800316915454e-06,
+      "loss": 0.2012,
+      "step": 638
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.482171472878062e-06,
+      "loss": 0.1969,
+      "step": 639
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.453953434581271e-06,
+      "loss": 0.18,
+      "step": 640
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.425826308697242e-06,
+      "loss": 0.209,
+      "step": 641
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.39779048585945e-06,
+      "loss": 0.2026,
+      "step": 642
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.3698463554333328e-06,
+      "loss": 0.1941,
+      "step": 643
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.3419943055108917e-06,
+      "loss": 0.201,
+      "step": 644
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.314234722905302e-06,
+      "loss": 0.2069,
+      "step": 645
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.286567993145541e-06,
+      "loss": 0.1914,
+      "step": 646
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.2589945004710312e-06,
+      "loss": 0.2062,
+      "step": 647
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.2315146278263053e-06,
+      "loss": 0.2041,
+      "step": 648
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.20412875685569e-06,
+      "loss": 0.2023,
+      "step": 649
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.176837267898002e-06,
+      "loss": 0.1981,
+      "step": 650
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.149640539981267e-06,
+      "loss": 0.2078,
+      "step": 651
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.122538950817452e-06,
+      "loss": 0.194,
+      "step": 652
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.095532876797237e-06,
+      "loss": 0.2012,
+      "step": 653
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.0686226929847617e-06,
+      "loss": 0.1979,
+      "step": 654
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.0418087731124324e-06,
+      "loss": 0.227,
+      "step": 655
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.015091489575729e-06,
+      "loss": 0.1859,
+      "step": 656
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.9884712134280324e-06,
+      "loss": 0.1976,
+      "step": 657
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.9619483143754703e-06,
+      "loss": 0.2009,
+      "step": 658
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.935523160771786e-06,
+      "loss": 0.1861,
+      "step": 659
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.909196119613218e-06,
+      "loss": 0.1989,
+      "step": 660
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.882967556533407e-06,
+      "loss": 0.221,
+      "step": 661
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.8568378357983138e-06,
+      "loss": 0.1986,
+      "step": 662
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.8308073203011667e-06,
+      "loss": 0.1985,
+      "step": 663
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.8048763715574133e-06,
+      "loss": 0.2207,
+      "step": 664
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.779045349699708e-06,
+      "loss": 0.2095,
+      "step": 665
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.753314613472906e-06,
+      "loss": 0.212,
+      "step": 666
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.727684520229079e-06,
+      "loss": 0.1969,
+      "step": 667
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.7021554259225567e-06,
+      "loss": 0.2045,
+      "step": 668
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.6767276851049818e-06,
+      "loss": 0.1982,
+      "step": 669
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.6514016509203843e-06,
+      "loss": 0.1855,
+      "step": 670
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.6261776751002798e-06,
+      "loss": 0.1903,
+      "step": 671
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.6010561079587817e-06,
+      "loss": 0.2167,
+      "step": 672
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.576037298387737e-06,
+      "loss": 0.2044,
+      "step": 673
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.551121593851883e-06,
+      "loss": 0.1963,
+      "step": 674
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.5263093403840145e-06,
+      "loss": 0.2001,
+      "step": 675
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.501600882580194e-06,
+      "loss": 0.1867,
+      "step": 676
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.4769965635949454e-06,
+      "loss": 0.1993,
+      "step": 677
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.452496725136503e-06,
+      "loss": 0.2088,
+      "step": 678
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.428101707462058e-06,
+      "loss": 0.2075,
+      "step": 679
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.4038118493730366e-06,
+      "loss": 0.202,
+      "step": 680
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.3796274882103964e-06,
+      "loss": 0.1953,
+      "step": 681
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.3555489598499327e-06,
+      "loss": 0.2028,
+      "step": 682
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.33157659869763e-06,
+      "loss": 0.2115,
+      "step": 683
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.3077107376850005e-06,
+      "loss": 0.1942,
+      "step": 684
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.283951708264468e-06,
+      "loss": 0.202,
+      "step": 685
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.260299840404765e-06,
+      "loss": 0.1918,
+      "step": 686
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.2367554625863496e-06,
+      "loss": 0.2337,
+      "step": 687
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.2133189017968436e-06,
+      "loss": 0.181,
+      "step": 688
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.1899904835264895e-06,
+      "loss": 0.2138,
+      "step": 689
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.1667705317636333e-06,
+      "loss": 0.1897,
+      "step": 690
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.1436593689902195e-06,
+      "loss": 0.1918,
+      "step": 691
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.120657316177319e-06,
+      "loss": 0.2129,
+      "step": 692
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.0977646927806682e-06,
+      "loss": 0.1906,
+      "step": 693
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.07498181673623e-06,
+      "loss": 0.2017,
+      "step": 694
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0523090044557824e-06,
+      "loss": 0.221,
+      "step": 695
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.029746570822524e-06,
+      "loss": 0.1876,
+      "step": 696
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0072948291866945e-06,
+      "loss": 0.2122,
+      "step": 697
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.984954091361232e-06,
+      "loss": 0.1951,
+      "step": 698
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.9627246676174363e-06,
+      "loss": 0.2019,
+      "step": 699
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.940606866680663e-06,
+      "loss": 0.2023,
+      "step": 700
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.9186009957260322e-06,
+      "loss": 0.1938,
+      "step": 701
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.896707360374167e-06,
+      "loss": 0.2184,
+      "step": 702
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.8749262646869437e-06,
+      "loss": 0.1986,
+      "step": 703
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.853258011163277e-06,
+      "loss": 0.1993,
+      "step": 704
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.8317029007349086e-06,
+      "loss": 0.1894,
+      "step": 705
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.8102612327622359e-06,
+      "loss": 0.2029,
+      "step": 706
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7889333050301505e-06,
+      "loss": 0.1976,
+      "step": 707
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7677194137439036e-06,
+      "loss": 0.2021,
+      "step": 708
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7466198535249913e-06,
+      "loss": 0.2072,
+      "step": 709
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.7256349174070685e-06,
+      "loss": 0.193,
+      "step": 710
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.7047648968318697e-06,
+      "loss": 0.2005,
+      "step": 711
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6840100816451676e-06,
+      "loss": 0.1883,
+      "step": 712
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6633707600927461e-06,
+      "loss": 0.1983,
+      "step": 713
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.642847218816398e-06,
+      "loss": 0.1983,
+      "step": 714
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.6224397428499394e-06,
+      "loss": 0.1828,
+      "step": 715
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.6021486156152633e-06,
+      "loss": 0.1969,
+      "step": 716
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.5819741189183902e-06,
+      "loss": 0.2048,
+      "step": 717
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.5619165329455578e-06,
+      "loss": 0.2085,
+      "step": 718
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.5419761362593334e-06,
+      "loss": 0.1954,
+      "step": 719
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.522153205794742e-06,
+      "loss": 0.203,
+      "step": 720
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.5024480168554246e-06,
+      "loss": 0.209,
+      "step": 721
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.4828608431098057e-06,
+      "loss": 0.1933,
+      "step": 722
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.4633919565873033e-06,
+      "loss": 0.1993,
+      "step": 723
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.444041627674544e-06,
+      "loss": 0.192,
+      "step": 724
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.4248101251116075e-06,
+      "loss": 0.1987,
+      "step": 725
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.4056977159883011e-06,
+      "loss": 0.1984,
+      "step": 726
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3867046657404416e-06,
+      "loss": 0.2119,
+      "step": 727
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3678312381461755e-06,
+      "loss": 0.1977,
+      "step": 728
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3490776953223107e-06,
+      "loss": 0.2,
+      "step": 729
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.3304442977206822e-06,
+      "loss": 0.2081,
+      "step": 730
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.3119313041245263e-06,
+      "loss": 0.1947,
+      "step": 731
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.2935389716448976e-06,
+      "loss": 0.1896,
+      "step": 732
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.275267555717088e-06,
+      "loss": 0.2032,
+      "step": 733
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.2571173100970824e-06,
+      "loss": 0.194,
+      "step": 734
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.23908848685804e-06,
+      "loss": 0.2073,
+      "step": 735
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.2211813363867852e-06,
+      "loss": 0.2023,
+      "step": 736
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.2033961073803368e-06,
+      "loss": 0.2039,
+      "step": 737
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.1857330468424466e-06,
+      "loss": 0.1933,
+      "step": 738
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.1681924000801791e-06,
+      "loss": 0.2037,
+      "step": 739
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.1507744107004937e-06,
+      "loss": 0.1888,
+      "step": 740
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.1334793206068739e-06,
+      "loss": 0.2021,
+      "step": 741
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.1163073699959548e-06,
+      "loss": 0.1918,
+      "step": 742
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.0992587973541935e-06,
+      "loss": 0.2013,
+      "step": 743
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.082333839454559e-06,
+      "loss": 0.192,
+      "step": 744
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.0655327313532393e-06,
+      "loss": 0.1881,
+      "step": 745
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.0488557063863814e-06,
+      "loss": 0.2068,
+      "step": 746
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.0323029961668463e-06,
+      "loss": 0.1786,
+      "step": 747
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.0158748305809952e-06,
+      "loss": 0.1897,
+      "step": 748
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 9.995714377854936e-07,
+      "loss": 0.2039,
+      "step": 749
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.833930442041506e-07,
+      "loss": 0.2147,
+      "step": 750
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.673398745247643e-07,
+      "loss": 0.1869,
+      "step": 751
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.514121516960051e-07,
+      "loss": 0.1991,
+      "step": 752
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.356100969243231e-07,
+      "loss": 0.2027,
+      "step": 753
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.199339296708698e-07,
+      "loss": 0.1988,
+      "step": 754
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.043838676484562e-07,
+      "loss": 0.1997,
+      "step": 755
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.889601268185233e-07,
+      "loss": 0.2049,
+      "step": 756
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.736629213881465e-07,
+      "loss": 0.2032,
+      "step": 757
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.584924638070614e-07,
+      "loss": 0.1965,
+      "step": 758
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.434489647647093e-07,
+      "loss": 0.1972,
+      "step": 759
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.285326331873156e-07,
+      "loss": 0.1974,
+      "step": 760
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.13743676234986e-07,
+      "loss": 0.2021,
+      "step": 761
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.990822992988267e-07,
+      "loss": 0.1896,
+      "step": 762
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.845487059980972e-07,
+      "loss": 0.1972,
+      "step": 763
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.701430981773816e-07,
+      "loss": 0.2011,
+      "step": 764
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.558656759037796e-07,
+      "loss": 0.2044,
+      "step": 765
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.417166374641349e-07,
+      "loss": 0.1986,
+      "step": 766
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.276961793622772e-07,
+      "loss": 0.1974,
+      "step": 767
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 7.13804496316296e-07,
+      "loss": 0.1935,
+      "step": 768
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 7.000417812558302e-07,
+      "loss": 0.2018,
+      "step": 769
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.864082253194027e-07,
+      "loss": 0.2064,
+      "step": 770
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.729040178517454e-07,
+      "loss": 0.2024,
+      "step": 771
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.595293464011887e-07,
+      "loss": 0.208,
+      "step": 772
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.46284396717044e-07,
+      "loss": 0.1946,
+      "step": 773
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.331693527470306e-07,
+      "loss": 0.197,
+      "step": 774
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.201843966347176e-07,
+      "loss": 0.1987,
+      "step": 775
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 6.073297087169982e-07,
+      "loss": 0.2027,
+      "step": 776
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.946054675215785e-07,
+      "loss": 0.1972,
+      "step": 777
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.820118497645078e-07,
+      "loss": 0.1921,
+      "step": 778
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.695490303477125e-07,
+      "loss": 0.1921,
+      "step": 779
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.572171823565797e-07,
+      "loss": 0.196,
+      "step": 780
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.450164770575417e-07,
+      "loss": 0.1867,
+      "step": 781
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.329470838957052e-07,
+      "loss": 0.1998,
+      "step": 782
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.210091704924947e-07,
+      "loss": 0.1938,
+      "step": 783
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.092029026433265e-07,
+      "loss": 0.2067,
+      "step": 784
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.975284443153062e-07,
+      "loss": 0.1963,
+      "step": 785
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.859859576449444e-07,
+      "loss": 0.1937,
+      "step": 786
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.745756029359183e-07,
+      "loss": 0.2356,
+      "step": 787
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.632975386568328e-07,
+      "loss": 0.2061,
+      "step": 788
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.5215192143902577e-07,
+      "loss": 0.2027,
+      "step": 789
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.4113890607439315e-07,
+      "loss": 0.1991,
+      "step": 790
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.3025864551323627e-07,
+      "loss": 0.1834,
+      "step": 791
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.1951129086214015e-07,
+      "loss": 0.1926,
+      "step": 792
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.088969913818741e-07,
+      "loss": 0.1896,
+      "step": 793
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.984158944853167e-07,
+      "loss": 0.2112,
+      "step": 794
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.8806814573541185e-07,
+      "loss": 0.1917,
+      "step": 795
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.778538888431471e-07,
+      "loss": 0.1893,
+      "step": 796
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.6777326566555196e-07,
+      "loss": 0.2103,
+      "step": 797
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.578264162037348e-07,
+      "loss": 0.1939,
+      "step": 798
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.4801347860093924e-07,
+      "loss": 0.1988,
+      "step": 799
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.383345891406176e-07,
+      "loss": 0.1959,
+      "step": 800
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.2878988224454346e-07,
+      "loss": 0.2014,
+      "step": 801
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.1937949047094465e-07,
+      "loss": 0.2018,
+      "step": 802
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.1010354451266213e-07,
+      "loss": 0.1961,
+      "step": 803
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.0096217319533386e-07,
+      "loss": 0.2146,
+      "step": 804
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.919555034756083e-07,
+      "loss": 0.1896,
+      "step": 805
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.8308366043937606e-07,
+      "loss": 0.2014,
+      "step": 806
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.7434676730003886e-07,
+      "loss": 0.2012,
+      "step": 807
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.657449453967931e-07,
+      "loss": 0.2054,
+      "step": 808
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.572783141929469e-07,
+      "loss": 0.204,
+      "step": 809
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.489469912742637e-07,
+      "loss": 0.2003,
+      "step": 810
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.407510923473233e-07,
+      "loss": 0.1976,
+      "step": 811
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.3269073123792007e-07,
+      "loss": 0.2064,
+      "step": 812
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.2476601988947965e-07,
+      "loss": 0.2098,
+      "step": 813
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.1697706836150356e-07,
+      "loss": 0.2013,
+      "step": 814
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.0932398482804374e-07,
+      "loss": 0.1835,
+      "step": 815
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.0180687557619816e-07,
+      "loss": 0.1963,
+      "step": 816
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.9442584500463323e-07,
+      "loss": 0.1835,
+      "step": 817
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.871809956221382e-07,
+      "loss": 0.2055,
+      "step": 818
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.800724280461963e-07,
+      "loss": 0.1901,
+      "step": 819
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.731002410015914e-07,
+      "loss": 0.1968,
+      "step": 820
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.6626453131903475e-07,
+      "loss": 0.1861,
+      "step": 821
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.5956539393382043e-07,
+      "loss": 0.2106,
+      "step": 822
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.5300292188450972e-07,
+      "loss": 0.2004,
+      "step": 823
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.465772063116333e-07,
+      "loss": 0.206,
+      "step": 824
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.4028833645643113e-07,
+      "loss": 0.2007,
+      "step": 825
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.34136399659609e-07,
+      "loss": 0.1743,
+      "step": 826
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.2812148136012726e-07,
+      "loss": 0.2041,
+      "step": 827
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.2224366509401732e-07,
+      "loss": 0.1903,
+      "step": 828
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1650303249321371e-07,
+      "loss": 0.2249,
+      "step": 829
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1089966328442836e-07,
+      "loss": 0.2008,
+      "step": 830
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.0543363528803696e-07,
+      "loss": 0.1973,
+      "step": 831
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.0010502441700543e-07,
+      "loss": 0.2076,
+      "step": 832
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 9.49139046758274e-08,
+      "loss": 0.1947,
+      "step": 833
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.986034815950173e-08,
+      "loss": 0.182,
+      "step": 834
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.494442505252998e-08,
+      "loss": 0.2141,
+      "step": 835
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.016620362794159e-08,
+      "loss": 0.1962,
+      "step": 836
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 7.55257502463469e-08,
+      "loss": 0.1924,
+      "step": 837
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 7.102312935501343e-08,
+      "loss": 0.1794,
+      "step": 838
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 6.665840348696994e-08,
+      "loss": 0.1926,
+      "step": 839
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 6.243163326014268e-08,
+      "loss": 0.1969,
+      "step": 840
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.8342877376508235e-08,
+      "loss": 0.1939,
+      "step": 841
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.4392192621283146e-08,
+      "loss": 0.1945,
+      "step": 842
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5.057963386213116e-08,
+      "loss": 0.1937,
+      "step": 843
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.690525404840163e-08,
+      "loss": 0.2145,
+      "step": 844
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.336910421039897e-08,
+      "loss": 0.2057,
+      "step": 845
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.9971233458665495e-08,
+      "loss": 0.1935,
+      "step": 846
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.671168898330857e-08,
+      "loss": 0.2108,
+      "step": 847
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.359051605333896e-08,
+      "loss": 0.203,
+      "step": 848
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 3.0607758016043546e-08,
+      "loss": 0.2071,
+      "step": 849
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.776345629638355e-08,
+      "loss": 0.2002,
+      "step": 850
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.5057650396420607e-08,
+      "loss": 0.2104,
+      "step": 851
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.2490377894768266e-08,
+      "loss": 0.1932,
+      "step": 852
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.006167444606466e-08,
+      "loss": 0.2055,
+      "step": 853
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.7771573780485107e-08,
+      "loss": 0.1887,
+      "step": 854
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.562010770326916e-08,
+      "loss": 0.2167,
+      "step": 855
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.3607306094277628e-08,
+      "loss": 0.1945,
+      "step": 856
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.173319690758401e-08,
+      "loss": 0.1936,
+      "step": 857
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 9.99780617107815e-09,
+      "loss": 0.2018,
+      "step": 858
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 8.401157986109853e-09,
+      "loss": 0.185,
+      "step": 859
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 6.9432745271535986e-09,
+      "loss": 0.1919,
+      "step": 860
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5.6241760414987856e-09,
+      "loss": 0.1976,
+      "step": 861
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 4.443880848971072e-09,
+      "loss": 0.2027,
+      "step": 862
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 3.402405341674797e-09,
+      "loss": 0.1989,
+      "step": 863
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.4997639837687217e-09,
+      "loss": 0.197,
+      "step": 864
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.7359693112606323e-09,
+      "loss": 0.2111,
+      "step": 865
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.1110319318408114e-09,
+      "loss": 0.2005,
+      "step": 866
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 6.24960524725493e-10,
+      "loss": 0.2101,
+      "step": 867
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 2.77761840542512e-10,
+      "loss": 0.199,
+      "step": 868
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 6.944070123582336e-11,
+      "loss": 0.1985,
+      "step": 869
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 0.0,
+      "loss": 0.2074,
+      "step": 870
+    },
+    {
+      "epoch": 3.0,
+      "step": 870,
+      "total_flos": 2.3051979663291187e+18,
+      "train_loss": 0.4218829972134239,
+      "train_runtime": 30029.0147,
+      "train_samples_per_second": 3.708,
+      "train_steps_per_second": 0.029
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 870,
+  "num_train_epochs": 3,
+  "save_steps": 2000,
+  "total_flos": 2.3051979663291187e+18,
+  "trial_name": null,
+  "trial_params": null
+}