diff --git "a/checkpoint-825/trainer_state.json" "b/checkpoint-825/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-825/trainer_state.json"
@@ -0,0 +1,4969 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.937799043062201,
+  "eval_steps": 500,
+  "global_step": 825,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.050505050505052e-07,
+      "loss": 2.9831,
+      "step": 1
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.0101010101010103e-06,
+      "loss": 3.0585,
+      "step": 2
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.5151515151515152e-06,
+      "loss": 2.9613,
+      "step": 3
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.0202020202020206e-06,
+      "loss": 2.9218,
+      "step": 4
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.5252525252525253e-06,
+      "loss": 3.1384,
+      "step": 5
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 3.0303030303030305e-06,
+      "loss": 2.9236,
+      "step": 6
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 3.5353535353535352e-06,
+      "loss": 2.9136,
+      "step": 7
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.040404040404041e-06,
+      "loss": 2.8794,
+      "step": 8
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 2.9476,
+      "step": 9
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.050505050505051e-06,
+      "loss": 2.8648,
+      "step": 10
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.555555555555556e-06,
+      "loss": 2.7458,
+      "step": 11
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 6.060606060606061e-06,
+      "loss": 2.995,
+      "step": 12
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 6.565656565656567e-06,
+      "loss": 2.7891,
+      "step": 13
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 7.0707070707070704e-06,
+      "loss": 2.9389,
+      "step": 14
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 7.5757575757575764e-06,
+      "loss": 2.9519,
+      "step": 15
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.080808080808082e-06,
+      "loss": 2.9985,
+      "step": 16
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.585858585858587e-06,
+      "loss": 2.8554,
+      "step": 17
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 2.9224,
+      "step": 18
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 9.595959595959595e-06,
+      "loss": 2.7839,
+      "step": 19
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.0101010101010101e-05,
+      "loss": 2.7728,
+      "step": 20
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.0606060606060607e-05,
+      "loss": 2.9102,
+      "step": 21
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.1111111111111112e-05,
+      "loss": 2.8321,
+      "step": 22
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.1616161616161616e-05,
+      "loss": 2.7018,
+      "step": 23
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.2121212121212122e-05,
+      "loss": 2.6893,
+      "step": 24
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.2626262626262628e-05,
+      "loss": 2.5533,
+      "step": 25
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.3131313131313134e-05,
+      "loss": 2.563,
+      "step": 26
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.3636363636363637e-05,
+      "loss": 2.6159,
+      "step": 27
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4141414141414141e-05,
+      "loss": 2.5117,
+      "step": 28
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.4646464646464647e-05,
+      "loss": 2.5404,
+      "step": 29
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.5151515151515153e-05,
+      "loss": 2.5025,
+      "step": 30
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.565656565656566e-05,
+      "loss": 2.4743,
+      "step": 31
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.6161616161616165e-05,
+      "loss": 2.4143,
+      "step": 32
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 2.3167,
+      "step": 33
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.7171717171717173e-05,
+      "loss": 2.3055,
+      "step": 34
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.7676767676767676e-05,
+      "loss": 2.2991,
+      "step": 35
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8181818181818182e-05,
+      "loss": 2.1632,
+      "step": 36
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8686868686868688e-05,
+      "loss": 2.2231,
+      "step": 37
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.919191919191919e-05,
+      "loss": 2.0811,
+      "step": 38
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.9696969696969697e-05,
+      "loss": 2.0147,
+      "step": 39
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.0202020202020203e-05,
+      "loss": 1.9772,
+      "step": 40
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.070707070707071e-05,
+      "loss": 1.9509,
+      "step": 41
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.1212121212121215e-05,
+      "loss": 1.88,
+      "step": 42
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.171717171717172e-05,
+      "loss": 1.7471,
+      "step": 43
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 1.7748,
+      "step": 44
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.272727272727273e-05,
+      "loss": 1.6766,
+      "step": 45
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.3232323232323232e-05,
+      "loss": 1.634,
+      "step": 46
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.3737373737373738e-05,
+      "loss": 1.5996,
+      "step": 47
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2.4242424242424244e-05,
+      "loss": 1.5091,
+      "step": 48
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2.474747474747475e-05,
+      "loss": 1.5411,
+      "step": 49
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.5252525252525256e-05,
+      "loss": 1.4753,
+      "step": 50
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.575757575757576e-05,
+      "loss": 1.4678,
+      "step": 51
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.6262626262626268e-05,
+      "loss": 1.5142,
+      "step": 52
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.676767676767677e-05,
+      "loss": 1.4128,
+      "step": 53
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.7272727272727273e-05,
+      "loss": 1.4542,
+      "step": 54
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 1.4162,
+      "step": 55
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.8282828282828282e-05,
+      "loss": 1.4502,
+      "step": 56
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.878787878787879e-05,
+      "loss": 1.295,
+      "step": 57
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.9292929292929294e-05,
+      "loss": 1.3446,
+      "step": 58
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.9797979797979796e-05,
+      "loss": 1.2936,
+      "step": 59
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.0303030303030306e-05,
+      "loss": 1.2942,
+      "step": 60
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.080808080808081e-05,
+      "loss": 1.3991,
+      "step": 61
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.131313131313132e-05,
+      "loss": 1.2571,
+      "step": 62
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.181818181818182e-05,
+      "loss": 1.4106,
+      "step": 63
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.232323232323233e-05,
+      "loss": 1.3386,
+      "step": 64
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.282828282828283e-05,
+      "loss": 1.2726,
+      "step": 65
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 1.3484,
+      "step": 66
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3838383838383844e-05,
+      "loss": 1.3355,
+      "step": 67
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.434343434343435e-05,
+      "loss": 1.3021,
+      "step": 68
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.484848484848485e-05,
+      "loss": 1.2461,
+      "step": 69
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.535353535353535e-05,
+      "loss": 1.346,
+      "step": 70
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.5858585858585855e-05,
+      "loss": 1.3081,
+      "step": 71
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.6363636363636364e-05,
+      "loss": 1.338,
+      "step": 72
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.686868686868687e-05,
+      "loss": 1.3435,
+      "step": 73
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.7373737373737376e-05,
+      "loss": 1.3025,
+      "step": 74
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.787878787878788e-05,
+      "loss": 1.2198,
+      "step": 75
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.838383838383838e-05,
+      "loss": 1.314,
+      "step": 76
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.888888888888889e-05,
+      "loss": 1.262,
+      "step": 77
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.939393939393939e-05,
+      "loss": 1.3292,
+      "step": 78
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.98989898989899e-05,
+      "loss": 1.1564,
+      "step": 79
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 4.0404040404040405e-05,
+      "loss": 1.1664,
+      "step": 80
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 4.0909090909090915e-05,
+      "loss": 1.262,
+      "step": 81
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 4.141414141414142e-05,
+      "loss": 1.2889,
+      "step": 82
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 4.191919191919192e-05,
+      "loss": 1.1571,
+      "step": 83
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 4.242424242424243e-05,
+      "loss": 1.165,
+      "step": 84
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 4.292929292929293e-05,
+      "loss": 1.1475,
+      "step": 85
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 4.343434343434344e-05,
+      "loss": 1.3032,
+      "step": 86
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 4.3939393939393944e-05,
+      "loss": 1.3794,
+      "step": 87
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 1.2761,
+      "step": 88
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.494949494949495e-05,
+      "loss": 1.1025,
+      "step": 89
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.545454545454546e-05,
+      "loss": 1.139,
+      "step": 90
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.595959595959596e-05,
+      "loss": 1.2408,
+      "step": 91
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 4.6464646464646464e-05,
+      "loss": 1.2418,
+      "step": 92
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 4.696969696969697e-05,
+      "loss": 1.089,
+      "step": 93
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 4.7474747474747476e-05,
+      "loss": 1.2089,
+      "step": 94
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 4.797979797979798e-05,
+      "loss": 1.1818,
+      "step": 95
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 4.848484848484849e-05,
+      "loss": 1.255,
+      "step": 96
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 4.898989898989899e-05,
+      "loss": 1.2591,
+      "step": 97
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 4.94949494949495e-05,
+      "loss": 1.1962,
+      "step": 98
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 1.1802,
+      "step": 99
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.050505050505051e-05,
+      "loss": 1.2168,
+      "step": 100
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5.101010101010101e-05,
+      "loss": 1.203,
+      "step": 101
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5.151515151515152e-05,
+      "loss": 1.2206,
+      "step": 102
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5.2020202020202026e-05,
+      "loss": 1.1691,
+      "step": 103
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5.2525252525252536e-05,
+      "loss": 1.1194,
+      "step": 104
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.303030303030303e-05,
+      "loss": 1.1041,
+      "step": 105
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.353535353535354e-05,
+      "loss": 1.1181,
+      "step": 106
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.4040404040404044e-05,
+      "loss": 1.2211,
+      "step": 107
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5.4545454545454546e-05,
+      "loss": 1.2155,
+      "step": 108
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5.5050505050505056e-05,
+      "loss": 1.0976,
+      "step": 109
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5.555555555555556e-05,
+      "loss": 1.27,
+      "step": 110
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5.606060606060606e-05,
+      "loss": 1.0708,
+      "step": 111
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5.6565656565656563e-05,
+      "loss": 1.1921,
+      "step": 112
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5.707070707070707e-05,
+      "loss": 1.1495,
+      "step": 113
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5.757575757575758e-05,
+      "loss": 1.113,
+      "step": 114
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5.808080808080808e-05,
+      "loss": 1.1519,
+      "step": 115
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5.858585858585859e-05,
+      "loss": 1.1364,
+      "step": 116
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5.90909090909091e-05,
+      "loss": 1.1744,
+      "step": 117
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5.959595959595959e-05,
+      "loss": 1.1322,
+      "step": 118
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 6.01010101010101e-05,
+      "loss": 1.0931,
+      "step": 119
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 6.060606060606061e-05,
+      "loss": 1.1242,
+      "step": 120
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.111111111111112e-05,
+      "loss": 1.1163,
+      "step": 121
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.161616161616162e-05,
+      "loss": 1.3074,
+      "step": 122
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.212121212121213e-05,
+      "loss": 1.213,
+      "step": 123
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.262626262626264e-05,
+      "loss": 1.0974,
+      "step": 124
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.313131313131313e-05,
+      "loss": 1.1272,
+      "step": 125
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.363636363636364e-05,
+      "loss": 1.2024,
+      "step": 126
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.414141414141415e-05,
+      "loss": 1.1247,
+      "step": 127
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 6.464646464646466e-05,
+      "loss": 1.1386,
+      "step": 128
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 6.515151515151516e-05,
+      "loss": 1.2468,
+      "step": 129
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 6.565656565656566e-05,
+      "loss": 1.1624,
+      "step": 130
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 6.616161616161617e-05,
+      "loss": 1.0378,
+      "step": 131
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 1.1028,
+      "step": 132
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 6.717171717171718e-05,
+      "loss": 1.1575,
+      "step": 133
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 6.767676767676769e-05,
+      "loss": 1.0586,
+      "step": 134
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 6.818181818181818e-05,
+      "loss": 1.0942,
+      "step": 135
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 6.86868686868687e-05,
+      "loss": 1.0912,
+      "step": 136
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.91919191919192e-05,
+      "loss": 1.0711,
+      "step": 137
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.96969696969697e-05,
+      "loss": 1.1007,
+      "step": 138
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 7.020202020202021e-05,
+      "loss": 1.2377,
+      "step": 139
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 7.07070707070707e-05,
+      "loss": 1.0448,
+      "step": 140
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 7.121212121212121e-05,
+      "loss": 1.0147,
+      "step": 141
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 7.171717171717171e-05,
+      "loss": 1.1971,
+      "step": 142
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 7.222222222222222e-05,
+      "loss": 1.0587,
+      "step": 143
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 7.272727272727273e-05,
+      "loss": 1.0985,
+      "step": 144
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 7.323232323232324e-05,
+      "loss": 1.2073,
+      "step": 145
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 7.373737373737373e-05,
+      "loss": 1.1526,
+      "step": 146
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 7.424242424242424e-05,
+      "loss": 1.0646,
+      "step": 147
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 7.474747474747475e-05,
+      "loss": 1.1812,
+      "step": 148
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 7.525252525252525e-05,
+      "loss": 1.0482,
+      "step": 149
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 7.575757575757576e-05,
+      "loss": 1.1301,
+      "step": 150
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 7.626262626262627e-05,
+      "loss": 0.9766,
+      "step": 151
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 7.676767676767676e-05,
+      "loss": 1.1234,
+      "step": 152
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 7.727272727272727e-05,
+      "loss": 1.0825,
+      "step": 153
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 7.777777777777778e-05,
+      "loss": 1.1252,
+      "step": 154
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 7.828282828282829e-05,
+      "loss": 1.0601,
+      "step": 155
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 7.878787878787879e-05,
+      "loss": 1.132,
+      "step": 156
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 7.92929292929293e-05,
+      "loss": 1.1978,
+      "step": 157
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 7.97979797979798e-05,
+      "loss": 1.0783,
+      "step": 158
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 8.03030303030303e-05,
+      "loss": 1.1149,
+      "step": 159
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 8.080808080808081e-05,
+      "loss": 1.0897,
+      "step": 160
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 8.131313131313132e-05,
+      "loss": 1.1671,
+      "step": 161
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 8.181818181818183e-05,
+      "loss": 0.9917,
+      "step": 162
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 8.232323232323233e-05,
+      "loss": 1.0764,
+      "step": 163
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 8.282828282828283e-05,
+      "loss": 1.2091,
+      "step": 164
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 8.333333333333334e-05,
+      "loss": 0.9721,
+      "step": 165
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 8.383838383838384e-05,
+      "loss": 1.0641,
+      "step": 166
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 8.434343434343435e-05,
+      "loss": 1.1268,
+      "step": 167
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 8.484848484848486e-05,
+      "loss": 1.1053,
+      "step": 168
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 8.535353535353535e-05,
+      "loss": 0.9693,
+      "step": 169
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 8.585858585858586e-05,
+      "loss": 1.2105,
+      "step": 170
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 8.636363636363637e-05,
+      "loss": 1.1085,
+      "step": 171
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 8.686868686868688e-05,
+      "loss": 1.102,
+      "step": 172
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.737373737373738e-05,
+      "loss": 1.1063,
+      "step": 173
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.787878787878789e-05,
+      "loss": 1.0495,
+      "step": 174
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 8.83838383838384e-05,
+      "loss": 1.0666,
+      "step": 175
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 8.888888888888889e-05,
+      "loss": 1.0146,
+      "step": 176
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 8.93939393939394e-05,
+      "loss": 1.0375,
+      "step": 177
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 8.98989898989899e-05,
+      "loss": 1.0957,
+      "step": 178
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 9.040404040404041e-05,
+      "loss": 0.9608,
+      "step": 179
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 9.090909090909092e-05,
+      "loss": 1.1103,
+      "step": 180
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 9.141414141414141e-05,
+      "loss": 1.0886,
+      "step": 181
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 9.191919191919192e-05,
+      "loss": 1.0851,
+      "step": 182
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 9.242424242424242e-05,
+      "loss": 1.0472,
+      "step": 183
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 9.292929292929293e-05,
+      "loss": 1.1018,
+      "step": 184
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 9.343434343434344e-05,
+      "loss": 1.1315,
+      "step": 185
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 9.393939393939395e-05,
+      "loss": 1.1845,
+      "step": 186
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 9.444444444444444e-05,
+      "loss": 1.0339,
+      "step": 187
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 9.494949494949495e-05,
+      "loss": 1.054,
+      "step": 188
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 9.545454545454546e-05,
+      "loss": 0.9609,
+      "step": 189
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 9.595959595959596e-05,
+      "loss": 1.2661,
+      "step": 190
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 9.646464646464647e-05,
+      "loss": 1.0543,
+      "step": 191
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 9.696969696969698e-05,
+      "loss": 1.0177,
+      "step": 192
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 9.747474747474747e-05,
+      "loss": 1.0964,
+      "step": 193
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 9.797979797979798e-05,
+      "loss": 1.2337,
+      "step": 194
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 9.848484848484849e-05,
+      "loss": 1.0947,
+      "step": 195
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 9.8989898989899e-05,
+      "loss": 1.2549,
+      "step": 196
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 9.94949494949495e-05,
+      "loss": 1.1435,
+      "step": 197
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 0.0001,
+      "loss": 0.9866,
+      "step": 198
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 0.0001005050505050505,
+      "loss": 1.1122,
+      "step": 199
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 0.00010101010101010102,
+      "loss": 0.9689,
+      "step": 200
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 0.00010151515151515152,
+      "loss": 0.9857,
+      "step": 201
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 0.00010202020202020202,
+      "loss": 1.0618,
+      "step": 202
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 0.00010252525252525254,
+      "loss": 0.9173,
+      "step": 203
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 0.00010303030303030303,
+      "loss": 1.0439,
+      "step": 204
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 0.00010353535353535353,
+      "loss": 0.9341,
+      "step": 205
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 0.00010404040404040405,
+      "loss": 0.9435,
+      "step": 206
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 0.00010454545454545455,
+      "loss": 0.9463,
+      "step": 207
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 0.00010505050505050507,
+      "loss": 0.9608,
+      "step": 208
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 0.00010555555555555557,
+      "loss": 0.9932,
+      "step": 209
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 0.00010606060606060606,
+      "loss": 1.0082,
+      "step": 210
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 0.00010656565656565659,
+      "loss": 0.9094,
+      "step": 211
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 0.00010707070707070708,
+      "loss": 0.926,
+      "step": 212
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 0.00010757575757575758,
+      "loss": 0.9295,
+      "step": 213
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 0.00010808080808080809,
+      "loss": 0.8823,
+      "step": 214
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 0.0001085858585858586,
+      "loss": 0.984,
+      "step": 215
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 0.00010909090909090909,
+      "loss": 1.0788,
+      "step": 216
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 0.0001095959595959596,
+      "loss": 1.0356,
+      "step": 217
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 0.00011010101010101011,
+      "loss": 0.9449,
+      "step": 218
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 0.00011060606060606061,
+      "loss": 1.0918,
+      "step": 219
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 0.00011111111111111112,
+      "loss": 0.9733,
+      "step": 220
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 0.00011161616161616161,
+      "loss": 1.0341,
+      "step": 221
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 0.00011212121212121212,
+      "loss": 0.981,
+      "step": 222
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 0.00011262626262626263,
+      "loss": 0.9898,
+      "step": 223
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 0.00011313131313131313,
+      "loss": 0.9891,
+      "step": 224
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 0.00011363636363636365,
+      "loss": 0.9681,
+      "step": 225
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 0.00011414141414141415,
+      "loss": 0.9518,
+      "step": 226
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 0.00011464646464646464,
+      "loss": 0.9959,
+      "step": 227
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 0.00011515151515151516,
+      "loss": 0.9998,
+      "step": 228
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 0.00011565656565656566,
+      "loss": 0.8444,
+      "step": 229
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 0.00011616161616161616,
+      "loss": 1.0036,
+      "step": 230
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 0.00011666666666666668,
+      "loss": 0.9296,
+      "step": 231
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 0.00011717171717171717,
+      "loss": 0.9785,
+      "step": 232
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 0.00011767676767676767,
+      "loss": 0.9004,
+      "step": 233
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 0.0001181818181818182,
+      "loss": 1.0123,
+      "step": 234
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 0.00011868686868686869,
+      "loss": 1.1656,
+      "step": 235
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 0.00011919191919191919,
+      "loss": 0.9841,
+      "step": 236
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 0.00011969696969696971,
+      "loss": 0.8851,
+      "step": 237
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 0.0001202020202020202,
+      "loss": 1.0571,
+      "step": 238
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 0.0001207070707070707,
+      "loss": 0.9695,
+      "step": 239
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 0.00012121212121212122,
+      "loss": 0.7582,
+      "step": 240
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 0.00012171717171717172,
+      "loss": 1.0994,
+      "step": 241
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 0.00012222222222222224,
+      "loss": 0.9242,
+      "step": 242
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 0.00012272727272727272,
+      "loss": 0.96,
+      "step": 243
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 0.00012323232323232323,
+      "loss": 1.0824,
+      "step": 244
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 0.00012373737373737374,
+      "loss": 1.0037,
+      "step": 245
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 0.00012424242424242425,
+      "loss": 0.8877,
+      "step": 246
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 0.00012474747474747473,
+      "loss": 0.9342,
+      "step": 247
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 0.00012525252525252527,
+      "loss": 1.0559,
+      "step": 248
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 0.00012575757575757575,
+      "loss": 1.0001,
+      "step": 249
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 0.00012626262626262626,
+      "loss": 0.873,
+      "step": 250
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 0.00012676767676767677,
+      "loss": 0.8952,
+      "step": 251
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 0.00012727272727272728,
+      "loss": 1.0204,
+      "step": 252
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 0.00012777777777777776,
+      "loss": 0.9089,
+      "step": 253
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 0.0001282828282828283,
+      "loss": 0.959,
+      "step": 254
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 0.00012878787878787878,
+      "loss": 1.0861,
+      "step": 255
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 0.00012929292929292932,
+      "loss": 0.9781,
+      "step": 256
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 0.0001297979797979798,
+      "loss": 1.0855,
+      "step": 257
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 0.0001303030303030303,
+      "loss": 0.9624,
+      "step": 258
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 0.00013080808080808082,
+      "loss": 1.0674,
+      "step": 259
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 0.00013131313131313133,
+      "loss": 1.0015,
+      "step": 260
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 0.0001318181818181818,
+      "loss": 1.0056,
+      "step": 261
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 0.00013232323232323235,
+      "loss": 0.9579,
+      "step": 262
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 0.00013282828282828283,
+      "loss": 1.0703,
+      "step": 263
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 0.00013333333333333334,
+      "loss": 1.1343,
+      "step": 264
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 0.00013383838383838385,
+      "loss": 0.9392,
+      "step": 265
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 0.00013434343434343436,
+      "loss": 0.9109,
+      "step": 266
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 0.00013484848484848484,
+      "loss": 0.8773,
+      "step": 267
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 0.00013535353535353538,
+      "loss": 0.8556,
+      "step": 268
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 0.00013585858585858586,
+      "loss": 0.7971,
+      "step": 269
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 0.00013636363636363637,
+      "loss": 0.7486,
+      "step": 270
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 0.00013686868686868688,
+      "loss": 0.7878,
+      "step": 271
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 0.0001373737373737374,
+      "loss": 0.735,
+      "step": 272
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 0.0001378787878787879,
+      "loss": 0.81,
+      "step": 273
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 0.0001383838383838384,
+      "loss": 0.8716,
+      "step": 274
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 0.0001388888888888889,
+      "loss": 0.7618,
+      "step": 275
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 0.0001393939393939394,
+      "loss": 0.8475,
+      "step": 276
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 0.0001398989898989899,
+      "loss": 0.8456,
+      "step": 277
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 0.00014040404040404042,
+      "loss": 0.8424,
+      "step": 278
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 0.00014090909090909093,
+      "loss": 0.9333,
+      "step": 279
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 0.0001414141414141414,
+      "loss": 0.8743,
+      "step": 280
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 0.00014191919191919192,
+      "loss": 0.9118,
+      "step": 281
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 0.00014242424242424243,
+      "loss": 0.8716,
+      "step": 282
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 0.00014292929292929294,
+      "loss": 0.8154,
+      "step": 283
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 0.00014343434343434342,
+      "loss": 0.7588,
+      "step": 284
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 0.00014393939393939396,
+      "loss": 0.8305,
+      "step": 285
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 0.00014444444444444444,
+      "loss": 0.7744,
+      "step": 286
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 0.00014494949494949495,
+      "loss": 0.9332,
+      "step": 287
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 0.00014545454545454546,
+      "loss": 0.8516,
+      "step": 288
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 0.00014595959595959597,
+      "loss": 0.9717,
+      "step": 289
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 0.00014646464646464648,
+      "loss": 0.8857,
+      "step": 290
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 0.00014696969696969698,
+      "loss": 0.8565,
+      "step": 291
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 0.00014747474747474747,
+      "loss": 0.8423,
+      "step": 292
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 0.000147979797979798,
+      "loss": 0.7805,
+      "step": 293
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 0.00014848484848484849,
+      "loss": 0.7872,
+      "step": 294
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 0.000148989898989899,
+      "loss": 0.7916,
+      "step": 295
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 0.0001494949494949495,
+      "loss": 0.8366,
+      "step": 296
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 0.8232,
+      "step": 297
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 0.0001505050505050505,
+      "loss": 0.849,
+      "step": 298
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 0.00015101010101010103,
+      "loss": 0.8255,
+      "step": 299
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 0.00015151515151515152,
+      "loss": 0.8112,
+      "step": 300
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 0.00015202020202020202,
+      "loss": 0.8022,
+      "step": 301
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 0.00015252525252525253,
+      "loss": 0.7779,
+      "step": 302
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 0.00015303030303030304,
+      "loss": 0.8785,
+      "step": 303
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 0.00015353535353535353,
+      "loss": 0.8064,
+      "step": 304
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 0.00015404040404040406,
+      "loss": 0.8665,
+      "step": 305
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 0.00015454545454545454,
+      "loss": 0.8263,
+      "step": 306
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 0.00015505050505050508,
+      "loss": 0.8513,
+      "step": 307
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 0.00015555555555555556,
+      "loss": 0.7776,
+      "step": 308
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 0.00015606060606060607,
+      "loss": 0.8738,
+      "step": 309
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 0.00015656565656565658,
+      "loss": 0.8859,
+      "step": 310
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 0.0001570707070707071,
+      "loss": 0.8391,
+      "step": 311
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 0.00015757575757575757,
+      "loss": 0.7208,
+      "step": 312
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 0.00015808080808080808,
+      "loss": 0.8693,
+      "step": 313
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 0.0001585858585858586,
+      "loss": 0.9205,
+      "step": 314
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 0.0001590909090909091,
+      "loss": 0.8755,
+      "step": 315
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 0.0001595959595959596,
+      "loss": 0.8039,
+      "step": 316
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 0.00016010101010101012,
+      "loss": 0.7777,
+      "step": 317
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 0.0001606060606060606,
+      "loss": 0.7352,
+      "step": 318
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 0.0001611111111111111,
+      "loss": 0.8669,
+      "step": 319
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 0.00016161616161616162,
+      "loss": 0.871,
+      "step": 320
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 0.00016212121212121213,
+      "loss": 0.9014,
+      "step": 321
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 0.00016262626262626264,
+      "loss": 0.973,
+      "step": 322
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 0.00016313131313131312,
+      "loss": 0.8274,
+      "step": 323
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 0.00016363636363636366,
+      "loss": 0.8801,
+      "step": 324
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 0.00016414141414141414,
+      "loss": 0.9279,
+      "step": 325
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 0.00016464646464646465,
+      "loss": 0.8946,
+      "step": 326
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 0.00016515151515151516,
+      "loss": 0.8768,
+      "step": 327
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 0.00016565656565656567,
+      "loss": 0.9011,
+      "step": 328
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 0.00016616161616161615,
+      "loss": 0.8496,
+      "step": 329
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 0.0001666666666666667,
+      "loss": 0.7764,
+      "step": 330
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 0.00016717171717171717,
+      "loss": 0.9669,
+      "step": 331
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 0.00016767676767676768,
+      "loss": 0.8556,
+      "step": 332
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 0.0001681818181818182,
+      "loss": 0.6294,
+      "step": 333
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 0.0001686868686868687,
+      "loss": 0.5905,
+      "step": 334
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 0.00016919191919191918,
+      "loss": 0.6959,
+      "step": 335
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 0.00016969696969696972,
+      "loss": 0.7869,
+      "step": 336
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 0.0001702020202020202,
+      "loss": 0.6294,
+      "step": 337
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 0.0001707070707070707,
+      "loss": 0.6231,
+      "step": 338
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 0.00017121212121212122,
+      "loss": 0.6398,
+      "step": 339
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 0.00017171717171717173,
+      "loss": 0.6709,
+      "step": 340
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 0.00017222222222222224,
+      "loss": 0.6276,
+      "step": 341
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 0.00017272727272727275,
+      "loss": 0.7472,
+      "step": 342
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 0.00017323232323232323,
+      "loss": 0.6593,
+      "step": 343
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 0.00017373737373737377,
+      "loss": 0.6895,
+      "step": 344
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 0.00017424242424242425,
+      "loss": 0.6663,
+      "step": 345
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 0.00017474747474747476,
+      "loss": 0.7473,
+      "step": 346
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 0.00017525252525252527,
+      "loss": 0.6482,
+      "step": 347
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 0.00017575757575757578,
+      "loss": 0.7009,
+      "step": 348
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 0.00017626262626262626,
+      "loss": 0.7085,
+      "step": 349
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 0.0001767676767676768,
+      "loss": 0.6697,
+      "step": 350
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 0.00017727272727272728,
+      "loss": 0.6169,
+      "step": 351
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 0.00017777777777777779,
+      "loss": 0.6844,
+      "step": 352
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 0.0001782828282828283,
+      "loss": 0.8004,
+      "step": 353
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 0.0001787878787878788,
+      "loss": 0.7215,
+      "step": 354
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 0.00017929292929292931,
+      "loss": 0.6962,
+      "step": 355
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 0.0001797979797979798,
+      "loss": 0.6727,
+      "step": 356
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 0.0001803030303030303,
+      "loss": 0.6453,
+      "step": 357
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 0.00018080808080808082,
+      "loss": 0.7153,
+      "step": 358
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 0.00018131313131313132,
+      "loss": 0.6596,
+      "step": 359
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 0.00018181818181818183,
+      "loss": 0.667,
+      "step": 360
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 0.00018232323232323234,
+      "loss": 0.859,
+      "step": 361
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 0.00018282828282828283,
+      "loss": 0.7127,
+      "step": 362
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 0.00018333333333333334,
+      "loss": 0.7715,
+      "step": 363
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 0.00018383838383838384,
+      "loss": 0.674,
+      "step": 364
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 0.00018434343434343435,
+      "loss": 0.716,
+      "step": 365
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 0.00018484848484848484,
+      "loss": 0.692,
+      "step": 366
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 0.00018535353535353537,
+      "loss": 0.7767,
+      "step": 367
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 0.00018585858585858586,
+      "loss": 0.7128,
+      "step": 368
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 0.00018636363636363636,
+      "loss": 0.7044,
+      "step": 369
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 0.00018686868686868687,
+      "loss": 0.7437,
+      "step": 370
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 0.00018737373737373738,
+      "loss": 0.8031,
+      "step": 371
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 0.0001878787878787879,
+      "loss": 0.7541,
+      "step": 372
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 0.0001883838383838384,
+      "loss": 0.792,
+      "step": 373
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 0.00018888888888888888,
+      "loss": 0.7628,
+      "step": 374
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 0.00018939393939393942,
+      "loss": 0.7543,
+      "step": 375
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 0.0001898989898989899,
+      "loss": 0.7325,
+      "step": 376
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 0.0001904040404040404,
+      "loss": 0.6901,
+      "step": 377
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 0.00019090909090909092,
+      "loss": 0.6723,
+      "step": 378
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 0.00019141414141414143,
+      "loss": 0.7287,
+      "step": 379
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 0.00019191919191919191,
+      "loss": 0.7178,
+      "step": 380
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 0.00019242424242424245,
+      "loss": 0.8214,
+      "step": 381
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 0.00019292929292929293,
+      "loss": 0.6586,
+      "step": 382
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 0.00019343434343434344,
+      "loss": 0.7933,
+      "step": 383
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 0.00019393939393939395,
+      "loss": 0.7253,
+      "step": 384
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 0.00019444444444444446,
+      "loss": 0.7694,
+      "step": 385
+    },
+    {
+      "epoch": 5.82,
+      "learning_rate": 0.00019494949494949494,
+      "loss": 0.795,
+      "step": 386
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 0.00019545454545454548,
+      "loss": 0.7402,
+      "step": 387
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 0.00019595959595959596,
+      "loss": 0.7058,
+      "step": 388
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 0.0001964646464646465,
+      "loss": 0.786,
+      "step": 389
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 0.00019696969696969698,
+      "loss": 0.7699,
+      "step": 390
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 0.0001974747474747475,
+      "loss": 0.8197,
+      "step": 391
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 0.000197979797979798,
+      "loss": 0.8621,
+      "step": 392
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 0.0001984848484848485,
+      "loss": 0.7168,
+      "step": 393
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 0.000198989898989899,
+      "loss": 0.7006,
+      "step": 394
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 0.0001994949494949495,
+      "loss": 0.765,
+      "step": 395
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 0.0002,
+      "loss": 0.7282,
+      "step": 396
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 0.00019999291961666908,
+      "loss": 0.8135,
+      "step": 397
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 0.0001999716794693129,
+      "loss": 0.6931,
+      "step": 398
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 0.0001999362825656992,
+      "loss": 0.5689,
+      "step": 399
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 0.0001998867339183008,
+      "loss": 0.5693,
+      "step": 400
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 0.00019982304054358614,
+      "loss": 0.5369,
+      "step": 401
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 0.00019974521146102537,
+      "loss": 0.4887,
+      "step": 402
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 0.00019965325769181325,
+      "loss": 0.5723,
+      "step": 403
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 0.00019954719225730847,
+      "loss": 0.6284,
+      "step": 404
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 0.00019942703017718975,
+      "loss": 0.5439,
+      "step": 405
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 0.00019929278846732884,
+      "loss": 0.5521,
+      "step": 406
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 0.00019914448613738106,
+      "loss": 0.5535,
+      "step": 407
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 0.0001989821441880933,
+      "loss": 0.4884,
+      "step": 408
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 0.00019880578560833016,
+      "loss": 0.549,
+      "step": 409
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 0.00019861543537181867,
+      "loss": 0.5644,
+      "step": 410
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 0.0001984111204336116,
+      "loss": 0.5865,
+      "step": 411
+    },
+    {
+      "epoch": 6.21,
+      "learning_rate": 0.00019819286972627066,
+      "loss": 0.5844,
+      "step": 412
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 0.00019796071415576925,
+      "loss": 0.6734,
+      "step": 413
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 0.00019771468659711595,
+      "loss": 0.5724,
+      "step": 414
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 0.0001974548218896993,
+      "loss": 0.5014,
+      "step": 415
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 0.00019718115683235417,
+      "loss": 0.5318,
+      "step": 416
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 0.00019689373017815073,
+      "loss": 0.574,
+      "step": 417
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 0.00019659258262890683,
+      "loss": 0.4451,
+      "step": 418
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 0.0001962777568294242,
+      "loss": 0.5648,
+      "step": 419
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 0.00019594929736144976,
+      "loss": 0.5006,
+      "step": 420
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 0.00019560725073736226,
+      "loss": 0.6189,
+      "step": 421
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 0.00019525166539358606,
+      "loss": 0.5462,
+      "step": 422
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 0.00019488259168373197,
+      "loss": 0.5923,
+      "step": 423
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 0.00019450008187146684,
+      "loss": 0.5371,
+      "step": 424
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 0.00019410419012311268,
+      "loss": 0.6752,
+      "step": 425
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 0.0001936949724999762,
+      "loss": 0.6184,
+      "step": 426
+    },
+    {
+      "epoch": 6.44,
+      "learning_rate": 0.0001932724869504101,
+      "loss": 0.6163,
+      "step": 427
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 0.00019283679330160726,
+      "loss": 0.5992,
+      "step": 428
+    },
+    {
+      "epoch": 6.47,
+      "learning_rate": 0.0001923879532511287,
+      "loss": 0.6189,
+      "step": 429
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 0.00019192603035816656,
+      "loss": 0.6301,
+      "step": 430
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 0.00019145109003454396,
+      "loss": 0.6293,
+      "step": 431
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 0.00019096319953545185,
+      "loss": 0.64,
+      "step": 432
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 0.00019046242794992538,
+      "loss": 0.6398,
+      "step": 433
+    },
+    {
+      "epoch": 6.55,
+      "learning_rate": 0.00018994884619106031,
+      "loss": 0.593,
+      "step": 434
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 0.00018942252698597113,
+      "loss": 0.6103,
+      "step": 435
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 0.00018888354486549237,
+      "loss": 0.6142,
+      "step": 436
+    },
+    {
+      "epoch": 6.59,
+      "learning_rate": 0.0001883319761536244,
+      "loss": 0.583,
+      "step": 437
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 0.00018776789895672558,
+      "loss": 0.5881,
+      "step": 438
+    },
+    {
+      "epoch": 6.62,
+      "learning_rate": 0.00018719139315245148,
+      "loss": 0.5929,
+      "step": 439
+    },
+    {
+      "epoch": 6.64,
+      "learning_rate": 0.00018660254037844388,
+      "loss": 0.6417,
+      "step": 440
+    },
+    {
+      "epoch": 6.65,
+      "learning_rate": 0.00018600142402077006,
+      "loss": 0.5681,
+      "step": 441
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 0.0001853881292021148,
+      "loss": 0.627,
+      "step": 442
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 0.00018476274276972636,
+      "loss": 0.5517,
+      "step": 443
+    },
+    {
+      "epoch": 6.7,
+      "learning_rate": 0.00018412535328311814,
+      "loss": 0.5575,
+      "step": 444
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 0.00018347605100152802,
+      "loss": 0.7175,
+      "step": 445
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 0.00018281492787113708,
+      "loss": 0.6409,
+      "step": 446
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 0.00018214207751204918,
+      "loss": 0.6176,
+      "step": 447
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 0.00018145759520503358,
+      "loss": 0.6328,
+      "step": 448
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 0.00018076157787803268,
+      "loss": 0.6437,
+      "step": 449
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 0.00018005412409243606,
+      "loss": 0.6153,
+      "step": 450
+    },
+    {
+      "epoch": 6.8,
+      "learning_rate": 0.00017933533402912354,
+      "loss": 0.639,
+      "step": 451
+    },
+    {
+      "epoch": 6.82,
+      "learning_rate": 0.00017860530947427875,
+      "loss": 0.5672,
+      "step": 452
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 0.00017786415380497553,
+      "loss": 0.5955,
+      "step": 453
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 0.00017711197197453878,
+      "loss": 0.6168,
+      "step": 454
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 0.00017634887049768237,
+      "loss": 0.694,
+      "step": 455
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 0.00017557495743542585,
+      "loss": 0.6137,
+      "step": 456
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 0.0001747903423797921,
+      "loss": 0.6617,
+      "step": 457
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 0.0001739951364382884,
+      "loss": 0.6298,
+      "step": 458
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 0.00017318945221817255,
+      "loss": 0.5342,
+      "step": 459
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 0.00017237340381050703,
+      "loss": 0.6858,
+      "step": 460
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 0.00017154710677400265,
+      "loss": 0.6101,
+      "step": 461
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 0.00017071067811865476,
+      "loss": 0.5749,
+      "step": 462
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 0.00016986423628917346,
+      "loss": 0.6245,
+      "step": 463
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 0.00016900790114821122,
+      "loss": 0.6779,
+      "step": 464
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 0.00016814179395938913,
+      "loss": 0.4573,
+      "step": 465
+    },
+    {
+      "epoch": 7.03,
+      "learning_rate": 0.00016726603737012529,
+      "loss": 0.4555,
+      "step": 466
+    },
+    {
+      "epoch": 7.04,
+      "learning_rate": 0.00016638075539426677,
+      "loss": 0.414,
+      "step": 467
+    },
+    {
+      "epoch": 7.06,
+      "learning_rate": 0.00016548607339452853,
+      "loss": 0.3871,
+      "step": 468
+    },
+    {
+      "epoch": 7.07,
+      "learning_rate": 0.00016458211806474088,
+      "loss": 0.5154,
+      "step": 469
+    },
+    {
+      "epoch": 7.09,
+      "learning_rate": 0.00016366901741190882,
+      "loss": 0.3657,
+      "step": 470
+    },
+    {
+      "epoch": 7.1,
+      "learning_rate": 0.0001627469007380852,
+      "loss": 0.4374,
+      "step": 471
+    },
+    {
+      "epoch": 7.12,
+      "learning_rate": 0.00016181589862206052,
+      "loss": 0.4454,
+      "step": 472
+    },
+    {
+      "epoch": 7.13,
+      "learning_rate": 0.00016087614290087208,
+      "loss": 0.4105,
+      "step": 473
+    },
+    {
+      "epoch": 7.15,
+      "learning_rate": 0.0001599277666511347,
+      "loss": 0.4332,
+      "step": 474
+    },
+    {
+      "epoch": 7.16,
+      "learning_rate": 0.0001589709041701962,
+      "loss": 0.4608,
+      "step": 475
+    },
+    {
+      "epoch": 7.18,
+      "learning_rate": 0.00015800569095711982,
+      "loss": 0.4211,
+      "step": 476
+    },
+    {
+      "epoch": 7.19,
+      "learning_rate": 0.0001570322636934964,
+      "loss": 0.4464,
+      "step": 477
+    },
+    {
+      "epoch": 7.21,
+      "learning_rate": 0.0001560507602240894,
+      "loss": 0.4389,
+      "step": 478
+    },
+    {
+      "epoch": 7.22,
+      "learning_rate": 0.0001550613195373149,
+      "loss": 0.4094,
+      "step": 479
+    },
+    {
+      "epoch": 7.24,
+      "learning_rate": 0.00015406408174555976,
+      "loss": 0.4297,
+      "step": 480
+    },
+    {
+      "epoch": 7.25,
+      "learning_rate": 0.00015305918806534082,
+      "loss": 0.3955,
+      "step": 481
+    },
+    {
+      "epoch": 7.27,
+      "learning_rate": 0.00015204678079730724,
+      "loss": 0.3662,
+      "step": 482
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 0.00015102700330609,
+      "loss": 0.4057,
+      "step": 483
+    },
+    {
+      "epoch": 7.3,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 0.4313,
+      "step": 484
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 0.00014896591631057912,
+      "loss": 0.4156,
+      "step": 485
+    },
+    {
+      "epoch": 7.33,
+      "learning_rate": 0.0001479248986720057,
+      "loss": 0.4419,
+      "step": 486
+    },
+    {
+      "epoch": 7.35,
+      "learning_rate": 0.00014687709450035837,
+      "loss": 0.3956,
+      "step": 487
+    },
+    {
+      "epoch": 7.36,
+      "learning_rate": 0.00014582265217274104,
+      "loss": 0.3943,
+      "step": 488
+    },
+    {
+      "epoch": 7.38,
+      "learning_rate": 0.00014476172100627127,
+      "loss": 0.364,
+      "step": 489
+    },
+    {
+      "epoch": 7.39,
+      "learning_rate": 0.00014369445123693596,
+      "loss": 0.3883,
+      "step": 490
+    },
+    {
+      "epoch": 7.41,
+      "learning_rate": 0.00014262099399831683,
+      "loss": 0.4107,
+      "step": 491
+    },
+    {
+      "epoch": 7.42,
+      "learning_rate": 0.00014154150130018866,
+      "loss": 0.4114,
+      "step": 492
+    },
+    {
+      "epoch": 7.44,
+      "learning_rate": 0.0001404561260069935,
+      "loss": 0.4161,
+      "step": 493
+    },
+    {
+      "epoch": 7.45,
+      "learning_rate": 0.00013936502181619416,
+      "loss": 0.4574,
+      "step": 494
+    },
+    {
+      "epoch": 7.47,
+      "learning_rate": 0.000138268343236509,
+      "loss": 0.404,
+      "step": 495
+    },
+    {
+      "epoch": 7.48,
+      "learning_rate": 0.00013716624556603274,
+      "loss": 0.4327,
+      "step": 496
+    },
+    {
+      "epoch": 7.5,
+      "learning_rate": 0.000136058884870245,
+      "loss": 0.3808,
+      "step": 497
+    },
+    {
+      "epoch": 7.51,
+      "learning_rate": 0.00013494641795990986,
+      "loss": 0.4414,
+      "step": 498
+    },
+    {
+      "epoch": 7.53,
+      "learning_rate": 0.00013382900236887075,
+      "loss": 0.444,
+      "step": 499
+    },
+    {
+      "epoch": 7.54,
+      "learning_rate": 0.00013270679633174218,
+      "loss": 0.4098,
+      "step": 500
+    },
+    {
+      "epoch": 7.56,
+      "learning_rate": 0.0001315799587615025,
+      "loss": 0.5061,
+      "step": 501
+    },
+    {
+      "epoch": 7.57,
+      "learning_rate": 0.0001304486492269907,
+      "loss": 0.4303,
+      "step": 502
+    },
+    {
+      "epoch": 7.59,
+      "learning_rate": 0.0001293130279303102,
+      "loss": 0.4687,
+      "step": 503
+    },
+    {
+      "epoch": 7.6,
+      "learning_rate": 0.00012817325568414297,
+      "loss": 0.4191,
+      "step": 504
+    },
+    {
+      "epoch": 7.62,
+      "learning_rate": 0.0001270294938889773,
+      "loss": 0.4531,
+      "step": 505
+    },
+    {
+      "epoch": 7.63,
+      "learning_rate": 0.00012588190451025207,
+      "loss": 0.4888,
+      "step": 506
+    },
+    {
+      "epoch": 7.65,
+      "learning_rate": 0.00012473065005542155,
+      "loss": 0.4177,
+      "step": 507
+    },
+    {
+      "epoch": 7.66,
+      "learning_rate": 0.00012357589355094275,
+      "loss": 0.5206,
+      "step": 508
+    },
+    {
+      "epoch": 7.68,
+      "learning_rate": 0.0001224177985191897,
+      "loss": 0.453,
+      "step": 509
+    },
+    {
+      "epoch": 7.69,
+      "learning_rate": 0.00012125652895529766,
+      "loss": 0.4442,
+      "step": 510
+    },
+    {
+      "epoch": 7.71,
+      "learning_rate": 0.00012009224930393988,
+      "loss": 0.4883,
+      "step": 511
+    },
+    {
+      "epoch": 7.72,
+      "learning_rate": 0.00011892512443604102,
+      "loss": 0.4499,
+      "step": 512
+    },
+    {
+      "epoch": 7.74,
+      "learning_rate": 0.00011775531962543036,
+      "loss": 0.5023,
+      "step": 513
+    },
+    {
+      "epoch": 7.75,
+      "learning_rate": 0.00011658300052543742,
+      "loss": 0.3726,
+      "step": 514
+    },
+    {
+      "epoch": 7.77,
+      "learning_rate": 0.00011540833314543458,
+      "loss": 0.4412,
+      "step": 515
+    },
+    {
+      "epoch": 7.78,
+      "learning_rate": 0.00011423148382732853,
+      "loss": 0.4233,
+      "step": 516
+    },
+    {
+      "epoch": 7.8,
+      "learning_rate": 0.00011305261922200519,
+      "loss": 0.5323,
+      "step": 517
+    },
+    {
+      "epoch": 7.81,
+      "learning_rate": 0.00011187190626573052,
+      "loss": 0.4949,
+      "step": 518
+    },
+    {
+      "epoch": 7.83,
+      "learning_rate": 0.00011068951215651132,
+      "loss": 0.3815,
+      "step": 519
+    },
+    {
+      "epoch": 7.84,
+      "learning_rate": 0.00010950560433041826,
+      "loss": 0.3834,
+      "step": 520
+    },
+    {
+      "epoch": 7.86,
+      "learning_rate": 0.00010832035043787625,
+      "loss": 0.4408,
+      "step": 521
+    },
+    {
+      "epoch": 7.87,
+      "learning_rate": 0.00010713391831992323,
+      "loss": 0.4821,
+      "step": 522
+    },
+    {
+      "epoch": 7.89,
+      "learning_rate": 0.00010594647598444312,
+      "loss": 0.382,
+      "step": 523
+    },
+    {
+      "epoch": 7.9,
+      "learning_rate": 0.00010475819158237425,
+      "loss": 0.4725,
+      "step": 524
+    },
+    {
+      "epoch": 7.92,
+      "learning_rate": 0.00010356923338389806,
+      "loss": 0.3808,
+      "step": 525
+    },
+    {
+      "epoch": 7.93,
+      "learning_rate": 0.00010237976975461075,
+      "loss": 0.421,
+      "step": 526
+    },
+    {
+      "epoch": 7.95,
+      "learning_rate": 0.00010118996913168144,
+      "loss": 0.4451,
+      "step": 527
+    },
+    {
+      "epoch": 7.96,
+      "learning_rate": 0.0001,
+      "loss": 0.433,
+      "step": 528
+    },
+    {
+      "epoch": 7.98,
+      "learning_rate": 9.881003086831859e-05,
+      "loss": 0.4279,
+      "step": 529
+    },
+    {
+      "epoch": 7.99,
+      "learning_rate": 9.762023024538926e-05,
+      "loss": 0.4398,
+      "step": 530
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 9.643076661610196e-05,
+      "loss": 0.3526,
+      "step": 531
+    },
+    {
+      "epoch": 8.02,
+      "learning_rate": 9.524180841762577e-05,
+      "loss": 0.3066,
+      "step": 532
+    },
+    {
+      "epoch": 8.04,
+      "learning_rate": 9.405352401555691e-05,
+      "loss": 0.2691,
+      "step": 533
+    },
+    {
+      "epoch": 8.05,
+      "learning_rate": 9.286608168007678e-05,
+      "loss": 0.2725,
+      "step": 534
+    },
+    {
+      "epoch": 8.07,
+      "learning_rate": 9.167964956212378e-05,
+      "loss": 0.2821,
+      "step": 535
+    },
+    {
+      "epoch": 8.08,
+      "learning_rate": 9.049439566958175e-05,
+      "loss": 0.277,
+      "step": 536
+    },
+    {
+      "epoch": 8.1,
+      "learning_rate": 8.931048784348875e-05,
+      "loss": 0.2472,
+      "step": 537
+    },
+    {
+      "epoch": 8.11,
+      "learning_rate": 8.812809373426951e-05,
+      "loss": 0.2742,
+      "step": 538
+    },
+    {
+      "epoch": 8.13,
+      "learning_rate": 8.694738077799488e-05,
+      "loss": 0.3079,
+      "step": 539
+    },
+    {
+      "epoch": 8.14,
+      "learning_rate": 8.57685161726715e-05,
+      "loss": 0.2882,
+      "step": 540
+    },
+    {
+      "epoch": 8.16,
+      "learning_rate": 8.459166685456547e-05,
+      "loss": 0.3107,
+      "step": 541
+    },
+    {
+      "epoch": 8.17,
+      "learning_rate": 8.34169994745626e-05,
+      "loss": 0.2603,
+      "step": 542
+    },
+    {
+      "epoch": 8.19,
+      "learning_rate": 8.224468037456969e-05,
+      "loss": 0.2487,
+      "step": 543
+    },
+    {
+      "epoch": 8.21,
+      "learning_rate": 8.107487556395901e-05,
+      "loss": 0.3226,
+      "step": 544
+    },
+    {
+      "epoch": 8.22,
+      "learning_rate": 7.990775069606012e-05,
+      "loss": 0.2915,
+      "step": 545
+    },
+    {
+      "epoch": 8.24,
+      "learning_rate": 7.874347104470234e-05,
+      "loss": 0.2517,
+      "step": 546
+    },
+    {
+      "epoch": 8.25,
+      "learning_rate": 7.758220148081028e-05,
+      "loss": 0.305,
+      "step": 547
+    },
+    {
+      "epoch": 8.27,
+      "learning_rate": 7.642410644905726e-05,
+      "loss": 0.3541,
+      "step": 548
+    },
+    {
+      "epoch": 8.28,
+      "learning_rate": 7.526934994457844e-05,
+      "loss": 0.2514,
+      "step": 549
+    },
+    {
+      "epoch": 8.3,
+      "learning_rate": 7.411809548974792e-05,
+      "loss": 0.3586,
+      "step": 550
+    },
+    {
+      "epoch": 8.31,
+      "learning_rate": 7.297050611102272e-05,
+      "loss": 0.2921,
+      "step": 551
+    },
+    {
+      "epoch": 8.33,
+      "learning_rate": 7.182674431585704e-05,
+      "loss": 0.2393,
+      "step": 552
+    },
+    {
+      "epoch": 8.34,
+      "learning_rate": 7.068697206968979e-05,
+      "loss": 0.2893,
+      "step": 553
+    },
+    {
+      "epoch": 8.36,
+      "learning_rate": 6.955135077300931e-05,
+      "loss": 0.3069,
+      "step": 554
+    },
+    {
+      "epoch": 8.37,
+      "learning_rate": 6.842004123849752e-05,
+      "loss": 0.3323,
+      "step": 555
+    },
+    {
+      "epoch": 8.39,
+      "learning_rate": 6.729320366825784e-05,
+      "loss": 0.283,
+      "step": 556
+    },
+    {
+      "epoch": 8.4,
+      "learning_rate": 6.617099763112929e-05,
+      "loss": 0.2652,
+      "step": 557
+    },
+    {
+      "epoch": 8.42,
+      "learning_rate": 6.505358204009017e-05,
+      "loss": 0.2642,
+      "step": 558
+    },
+    {
+      "epoch": 8.43,
+      "learning_rate": 6.394111512975504e-05,
+      "loss": 0.2811,
+      "step": 559
+    },
+    {
+      "epoch": 8.45,
+      "learning_rate": 6.283375443396726e-05,
+      "loss": 0.329,
+      "step": 560
+    },
+    {
+      "epoch": 8.46,
+      "learning_rate": 6.173165676349103e-05,
+      "loss": 0.28,
+      "step": 561
+    },
+    {
+      "epoch": 8.48,
+      "learning_rate": 6.063497818380587e-05,
+      "loss": 0.28,
+      "step": 562
+    },
+    {
+      "epoch": 8.49,
+      "learning_rate": 5.9543873993006496e-05,
+      "loss": 0.3122,
+      "step": 563
+    },
+    {
+      "epoch": 8.51,
+      "learning_rate": 5.845849869981137e-05,
+      "loss": 0.2803,
+      "step": 564
+    },
+    {
+      "epoch": 8.52,
+      "learning_rate": 5.73790060016832e-05,
+      "loss": 0.3026,
+      "step": 565
+    },
+    {
+      "epoch": 8.54,
+      "learning_rate": 5.630554876306407e-05,
+      "loss": 0.2672,
+      "step": 566
+    },
+    {
+      "epoch": 8.55,
+      "learning_rate": 5.5238278993728756e-05,
+      "loss": 0.2843,
+      "step": 567
+    },
+    {
+      "epoch": 8.57,
+      "learning_rate": 5.417734782725896e-05,
+      "loss": 0.2332,
+      "step": 568
+    },
+    {
+      "epoch": 8.58,
+      "learning_rate": 5.3122905499641615e-05,
+      "loss": 0.2727,
+      "step": 569
+    },
+    {
+      "epoch": 8.6,
+      "learning_rate": 5.207510132799436e-05,
+      "loss": 0.3152,
+      "step": 570
+    },
+    {
+      "epoch": 8.61,
+      "learning_rate": 5.1034083689420905e-05,
+      "loss": 0.2891,
+      "step": 571
+    },
+    {
+      "epoch": 8.63,
+      "learning_rate": 5.000000000000002e-05,
+      "loss": 0.3056,
+      "step": 572
+    },
+    {
+      "epoch": 8.64,
+      "learning_rate": 4.8972996693910054e-05,
+      "loss": 0.2962,
+      "step": 573
+    },
+    {
+      "epoch": 8.66,
+      "learning_rate": 4.795321920269279e-05,
+      "loss": 0.3021,
+      "step": 574
+    },
+    {
+      "epoch": 8.67,
+      "learning_rate": 4.694081193465921e-05,
+      "loss": 0.2582,
+      "step": 575
+    },
+    {
+      "epoch": 8.69,
+      "learning_rate": 4.593591825444028e-05,
+      "loss": 0.2658,
+      "step": 576
+    },
+    {
+      "epoch": 8.7,
+      "learning_rate": 4.493868046268514e-05,
+      "loss": 0.3138,
+      "step": 577
+    },
+    {
+      "epoch": 8.72,
+      "learning_rate": 4.394923977591059e-05,
+      "loss": 0.2751,
+      "step": 578
+    },
+    {
+      "epoch": 8.73,
+      "learning_rate": 4.296773630650358e-05,
+      "loss": 0.2633,
+      "step": 579
+    },
+    {
+      "epoch": 8.75,
+      "learning_rate": 4.19943090428802e-05,
+      "loss": 0.3056,
+      "step": 580
+    },
+    {
+      "epoch": 8.76,
+      "learning_rate": 4.10290958298038e-05,
+      "loss": 0.2989,
+      "step": 581
+    },
+    {
+      "epoch": 8.78,
+      "learning_rate": 4.007223334886531e-05,
+      "loss": 0.258,
+      "step": 582
+    },
+    {
+      "epoch": 8.79,
+      "learning_rate": 3.9123857099127936e-05,
+      "loss": 0.3036,
+      "step": 583
+    },
+    {
+      "epoch": 8.81,
+      "learning_rate": 3.8184101377939476e-05,
+      "loss": 0.3249,
+      "step": 584
+    },
+    {
+      "epoch": 8.82,
+      "learning_rate": 3.725309926191479e-05,
+      "loss": 0.2593,
+      "step": 585
+    },
+    {
+      "epoch": 8.84,
+      "learning_rate": 3.6330982588091186e-05,
+      "loss": 0.2858,
+      "step": 586
+    },
+    {
+      "epoch": 8.85,
+      "learning_rate": 3.541788193525913e-05,
+      "loss": 0.2881,
+      "step": 587
+    },
+    {
+      "epoch": 8.87,
+      "learning_rate": 3.45139266054715e-05,
+      "loss": 0.244,
+      "step": 588
+    },
+    {
+      "epoch": 8.88,
+      "learning_rate": 3.361924460573325e-05,
+      "loss": 0.2691,
+      "step": 589
+    },
+    {
+      "epoch": 8.9,
+      "learning_rate": 3.273396262987475e-05,
+      "loss": 0.2898,
+      "step": 590
+    },
+    {
+      "epoch": 8.91,
+      "learning_rate": 3.185820604061088e-05,
+      "loss": 0.3022,
+      "step": 591
+    },
+    {
+      "epoch": 8.93,
+      "learning_rate": 3.099209885178882e-05,
+      "loss": 0.2656,
+      "step": 592
+    },
+    {
+      "epoch": 8.94,
+      "learning_rate": 3.013576371082655e-05,
+      "loss": 0.2368,
+      "step": 593
+    },
+    {
+      "epoch": 8.96,
+      "learning_rate": 2.9289321881345254e-05,
+      "loss": 0.328,
+      "step": 594
+    },
+    {
+      "epoch": 8.97,
+      "learning_rate": 2.8452893225997346e-05,
+      "loss": 0.3093,
+      "step": 595
+    },
+    {
+      "epoch": 8.99,
+      "learning_rate": 2.7626596189492983e-05,
+      "loss": 0.2982,
+      "step": 596
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 2.681054778182748e-05,
+      "loss": 0.2176,
+      "step": 597
+    },
+    {
+      "epoch": 9.02,
+      "learning_rate": 2.6004863561711635e-05,
+      "loss": 0.2075,
+      "step": 598
+    },
+    {
+      "epoch": 9.03,
+      "learning_rate": 2.5209657620207915e-05,
+      "loss": 0.2337,
+      "step": 599
+    },
+    {
+      "epoch": 9.05,
+      "learning_rate": 2.4425042564574184e-05,
+      "loss": 0.2136,
+      "step": 600
+    },
+    {
+      "epoch": 9.06,
+      "learning_rate": 2.3651129502317647e-05,
+      "loss": 0.1918,
+      "step": 601
+    },
+    {
+      "epoch": 9.08,
+      "learning_rate": 2.288802802546124e-05,
+      "loss": 0.1969,
+      "step": 602
+    },
+    {
+      "epoch": 9.1,
+      "learning_rate": 2.2135846195024513e-05,
+      "loss": 0.1933,
+      "step": 603
+    },
+    {
+      "epoch": 9.11,
+      "learning_rate": 2.139469052572127e-05,
+      "loss": 0.2262,
+      "step": 604
+    },
+    {
+      "epoch": 9.13,
+      "learning_rate": 2.0664665970876496e-05,
+      "loss": 0.192,
+      "step": 605
+    },
+    {
+      "epoch": 9.14,
+      "learning_rate": 1.994587590756397e-05,
+      "loss": 0.1849,
+      "step": 606
+    },
+    {
+      "epoch": 9.16,
+      "learning_rate": 1.923842212196735e-05,
+      "loss": 0.213,
+      "step": 607
+    },
+    {
+      "epoch": 9.17,
+      "learning_rate": 1.854240479496643e-05,
+      "loss": 0.1907,
+      "step": 608
+    },
+    {
+      "epoch": 9.19,
+      "learning_rate": 1.7857922487950874e-05,
+      "loss": 0.2197,
+      "step": 609
+    },
+    {
+      "epoch": 9.2,
+      "learning_rate": 1.7185072128862933e-05,
+      "loss": 0.226,
+      "step": 610
+    },
+    {
+      "epoch": 9.22,
+      "learning_rate": 1.6523948998471973e-05,
+      "loss": 0.1816,
+      "step": 611
+    },
+    {
+      "epoch": 9.23,
+      "learning_rate": 1.587464671688187e-05,
+      "loss": 0.2111,
+      "step": 612
+    },
+    {
+      "epoch": 9.25,
+      "learning_rate": 1.523725723027365e-05,
+      "loss": 0.19,
+      "step": 613
+    },
+    {
+      "epoch": 9.26,
+      "learning_rate": 1.4611870797885197e-05,
+      "loss": 0.2407,
+      "step": 614
+    },
+    {
+      "epoch": 9.28,
+      "learning_rate": 1.3998575979229944e-05,
+      "loss": 0.1747,
+      "step": 615
+    },
+    {
+      "epoch": 9.29,
+      "learning_rate": 1.339745962155613e-05,
+      "loss": 0.1693,
+      "step": 616
+    },
+    {
+      "epoch": 9.31,
+      "learning_rate": 1.280860684754852e-05,
+      "loss": 0.2539,
+      "step": 617
+    },
+    {
+      "epoch": 9.32,
+      "learning_rate": 1.2232101043274436e-05,
+      "loss": 0.1913,
+      "step": 618
+    },
+    {
+      "epoch": 9.34,
+      "learning_rate": 1.166802384637561e-05,
+      "loss": 0.2216,
+      "step": 619
+    },
+    {
+      "epoch": 9.35,
+      "learning_rate": 1.1116455134507664e-05,
+      "loss": 0.2111,
+      "step": 620
+    },
+    {
+      "epoch": 9.37,
+      "learning_rate": 1.057747301402887e-05,
+      "loss": 0.2153,
+      "step": 621
+    },
+    {
+      "epoch": 9.38,
+      "learning_rate": 1.0051153808939685e-05,
+      "loss": 0.1951,
+      "step": 622
+    },
+    {
+      "epoch": 9.4,
+      "learning_rate": 9.537572050074618e-06,
+      "loss": 0.1915,
+      "step": 623
+    },
+    {
+      "epoch": 9.41,
+      "learning_rate": 9.036800464548157e-06,
+      "loss": 0.2099,
+      "step": 624
+    },
+    {
+      "epoch": 9.43,
+      "learning_rate": 8.548909965456065e-06,
+      "loss": 0.1805,
+      "step": 625
+    },
+    {
+      "epoch": 9.44,
+      "learning_rate": 8.073969641833445e-06,
+      "loss": 0.2221,
+      "step": 626
+    },
+    {
+      "epoch": 9.46,
+      "learning_rate": 7.612046748871327e-06,
+      "loss": 0.2165,
+      "step": 627
+    },
+    {
+      "epoch": 9.47,
+      "learning_rate": 7.163206698392744e-06,
+      "loss": 0.199,
+      "step": 628
+    },
+    {
+      "epoch": 9.49,
+      "learning_rate": 6.7275130495899175e-06,
+      "loss": 0.1879,
+      "step": 629
+    },
+    {
+      "epoch": 9.5,
+      "learning_rate": 6.3050275000238414e-06,
+      "loss": 0.2181,
+      "step": 630
+    },
+    {
+      "epoch": 9.52,
+      "learning_rate": 5.895809876887326e-06,
+      "loss": 0.1951,
+      "step": 631
+    },
+    {
+      "epoch": 9.53,
+      "learning_rate": 5.499918128533155e-06,
+      "loss": 0.1942,
+      "step": 632
+    },
+    {
+      "epoch": 9.55,
+      "learning_rate": 5.1174083162680465e-06,
+      "loss": 0.2205,
+      "step": 633
+    },
+    {
+      "epoch": 9.56,
+      "learning_rate": 4.748334606413951e-06,
+      "loss": 0.2113,
+      "step": 634
+    },
+    {
+      "epoch": 9.58,
+      "learning_rate": 4.392749262637752e-06,
+      "loss": 0.2094,
+      "step": 635
+    },
+    {
+      "epoch": 9.59,
+      "learning_rate": 4.050702638550275e-06,
+      "loss": 0.1757,
+      "step": 636
+    },
+    {
+      "epoch": 9.61,
+      "learning_rate": 3.722243170575801e-06,
+      "loss": 0.2251,
+      "step": 637
+    },
+    {
+      "epoch": 9.62,
+      "learning_rate": 3.40741737109318e-06,
+      "loss": 0.1949,
+      "step": 638
+    },
+    {
+      "epoch": 9.64,
+      "learning_rate": 3.1062698218492724e-06,
+      "loss": 0.2234,
+      "step": 639
+    },
+    {
+      "epoch": 9.65,
+      "learning_rate": 2.818843167645835e-06,
+      "loss": 0.2287,
+      "step": 640
+    },
+    {
+      "epoch": 9.67,
+      "learning_rate": 2.5451781103006944e-06,
+      "loss": 0.1959,
+      "step": 641
+    },
+    {
+      "epoch": 9.68,
+      "learning_rate": 2.2853134028840594e-06,
+      "loss": 0.1696,
+      "step": 642
+    },
+    {
+      "epoch": 9.7,
+      "learning_rate": 2.039285844230765e-06,
+      "loss": 0.1919,
+      "step": 643
+    },
+    {
+      "epoch": 9.71,
+      "learning_rate": 1.8071302737293295e-06,
+      "loss": 0.1944,
+      "step": 644
+    },
+    {
+      "epoch": 9.73,
+      "learning_rate": 1.5888795663883904e-06,
+      "loss": 0.1871,
+      "step": 645
+    },
+    {
+      "epoch": 9.74,
+      "learning_rate": 1.3845646281813507e-06,
+      "loss": 0.2535,
+      "step": 646
+    },
+    {
+      "epoch": 9.76,
+      "learning_rate": 1.1942143916698457e-06,
+      "loss": 0.1789,
+      "step": 647
+    },
+    {
+      "epoch": 9.77,
+      "learning_rate": 1.0178558119067315e-06,
+      "loss": 0.1722,
+      "step": 648
+    },
+    {
+      "epoch": 9.79,
+      "learning_rate": 8.555138626189618e-07,
+      "loss": 0.2148,
+      "step": 649
+    },
+    {
+      "epoch": 9.8,
+      "learning_rate": 7.072115326711704e-07,
+      "loss": 0.1889,
+      "step": 650
+    },
+    {
+      "epoch": 9.82,
+      "learning_rate": 5.729698228102653e-07,
+      "loss": 0.1746,
+      "step": 651
+    },
+    {
+      "epoch": 9.83,
+      "learning_rate": 4.5280774269154115e-07,
+      "loss": 0.1984,
+      "step": 652
+    },
+    {
+      "epoch": 9.85,
+      "learning_rate": 3.467423081867649e-07,
+      "loss": 0.2096,
+      "step": 653
+    },
+    {
+      "epoch": 9.86,
+      "learning_rate": 2.547885389746485e-07,
+      "loss": 0.2252,
+      "step": 654
+    },
+    {
+      "epoch": 9.88,
+      "learning_rate": 1.7695945641386102e-07,
+      "loss": 0.1878,
+      "step": 655
+    },
+    {
+      "epoch": 9.89,
+      "learning_rate": 1.1326608169920372e-07,
+      "loss": 0.1954,
+      "step": 656
+    },
+    {
+      "epoch": 9.91,
+      "learning_rate": 6.37174343008251e-08,
+      "loss": 0.1886,
+      "step": 657
+    },
+    {
+      "epoch": 9.92,
+      "learning_rate": 2.8320530687098166e-08,
+      "loss": 0.251,
+      "step": 658
+    },
+    {
+      "epoch": 9.94,
+      "learning_rate": 7.080383330915208e-09,
+      "loss": 0.2084,
+      "step": 659
+    },
+    {
+      "epoch": 9.95,
+      "learning_rate": 0.0,
+      "loss": 0.2066,
+      "step": 660
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 0.00010657942003278107,
+      "loss": 1.6314,
+      "step": 661
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 0.00010564058429494808,
+      "loss": 1.5509,
+      "step": 662
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 0.00010470124952590977,
+      "loss": 1.6615,
+      "step": 663
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 0.00010376149883006406,
+      "loss": 1.3416,
+      "step": 664
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 0.0001028214153486066,
+      "loss": 1.4959,
+      "step": 665
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 0.00010188108225217516,
+      "loss": 1.2929,
+      "step": 666
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 0.00010094058273349125,
+      "loss": 1.2884,
+      "step": 667
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 0.0001,
+      "loss": 1.272,
+      "step": 668
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 9.90594172665088e-05,
+      "loss": 1.2148,
+      "step": 669
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 9.811891774782484e-05,
+      "loss": 1.2181,
+      "step": 670
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 9.717858465139342e-05,
+      "loss": 1.2137,
+      "step": 671
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 9.623850116993596e-05,
+      "loss": 1.1983,
+      "step": 672
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 9.529875047409027e-05,
+      "loss": 1.1848,
+      "step": 673
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 9.43594157050519e-05,
+      "loss": 1.0902,
+      "step": 674
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 9.342057996721894e-05,
+      "loss": 1.1571,
+      "step": 675
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 9.248232632083923e-05,
+      "loss": 0.9283,
+      "step": 676
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 9.15447377746621e-05,
+      "loss": 1.1645,
+      "step": 677
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 9.060789727859457e-05,
+      "loss": 1.0975,
+      "step": 678
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 8.967188771636236e-05,
+      "loss": 0.9576,
+      "step": 679
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 8.873679189817732e-05,
+      "loss": 1.1544,
+      "step": 680
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 8.78026925534108e-05,
+      "loss": 1.0777,
+      "step": 681
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 8.686967232327478e-05,
+      "loss": 1.0782,
+      "step": 682
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 8.59378137535102e-05,
+      "loss": 1.1202,
+      "step": 683
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 8.500719928708405e-05,
+      "loss": 1.1844,
+      "step": 684
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 8.407791125689578e-05,
+      "loss": 1.0473,
+      "step": 685
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 8.315003187849277e-05,
+      "loss": 1.0613,
+      "step": 686
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 8.222364324279689e-05,
+      "loss": 1.1134,
+      "step": 687
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 8.129882730884168e-05,
+      "loss": 1.0871,
+      "step": 688
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 8.037566589652141e-05,
+      "loss": 1.1047,
+      "step": 689
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 7.945424067935217e-05,
+      "loss": 1.024,
+      "step": 690
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 7.853463317724614e-05,
+      "loss": 1.2049,
+      "step": 691
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 7.761692474929961e-05,
+      "loss": 1.0818,
+      "step": 692
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 7.67011965865947e-05,
+      "loss": 1.2091,
+      "step": 693
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 7.578752970501642e-05,
+      "loss": 1.1569,
+      "step": 694
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 7.487600493808513e-05,
+      "loss": 1.0476,
+      "step": 695
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 7.396670292980492e-05,
+      "loss": 0.8989,
+      "step": 696
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 7.305970412752909e-05,
+      "loss": 1.0927,
+      "step": 697
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 7.215508877484258e-05,
+      "loss": 1.1766,
+      "step": 698
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 7.125293690446306e-05,
+      "loss": 1.2596,
+      "step": 699
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 7.035332833115997e-05,
+      "loss": 1.1233,
+      "step": 700
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 6.945634264469339e-05,
+      "loss": 0.997,
+      "step": 701
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 6.85620592027725e-05,
+      "loss": 1.0831,
+      "step": 702
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 6.76705571240348e-05,
+      "loss": 1.0424,
+      "step": 703
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 6.678191528104629e-05,
+      "loss": 1.2312,
+      "step": 704
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 6.58962122933234e-05,
+      "loss": 1.0623,
+      "step": 705
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 6.501352652037764e-05,
+      "loss": 1.0416,
+      "step": 706
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 6.413393605478275e-05,
+      "loss": 0.9957,
+      "step": 707
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 6.325751871526588e-05,
+      "loss": 0.9729,
+      "step": 708
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 6.238435203982278e-05,
+      "loss": 1.124,
+      "step": 709
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 6.151451327885805e-05,
+      "loss": 1.0767,
+      "step": 710
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 6.0648079388350466e-05,
+      "loss": 1.0175,
+      "step": 711
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 5.978512702304461e-05,
+      "loss": 0.9669,
+      "step": 712
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 5.892573252966926e-05,
+      "loss": 1.1609,
+      "step": 713
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 5.806997194018271e-05,
+      "loss": 1.0129,
+      "step": 714
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 5.721792096504611e-05,
+      "loss": 1.0527,
+      "step": 715
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 5.636965498652544e-05,
+      "loss": 1.0049,
+      "step": 716
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 5.5525249052022076e-05,
+      "loss": 1.0176,
+      "step": 717
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 5.4684777867433335e-05,
+      "loss": 0.9946,
+      "step": 718
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 5.3848315790543126e-05,
+      "loss": 0.9892,
+      "step": 719
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 5.301593682444352e-05,
+      "loss": 1.0541,
+      "step": 720
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 5.218771461098733e-05,
+      "loss": 1.0958,
+      "step": 721
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 5.1363722424273156e-05,
+      "loss": 0.9798,
+      "step": 722
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 5.054403316416247e-05,
+      "loss": 1.0182,
+      "step": 723
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 4.972871934983032e-05,
+      "loss": 1.2027,
+      "step": 724
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 4.891785311334923e-05,
+      "loss": 0.9431,
+      "step": 725
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 4.811150619330759e-05,
+      "loss": 1.0192,
+      "step": 726
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 4.7309749928463035e-05,
+      "loss": 1.0675,
+      "step": 727
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 4.6512655251430695e-05,
+      "loss": 1.0731,
+      "step": 728
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 4.5720292682407874e-05,
+      "loss": 0.9864,
+      "step": 729
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 4.493273232293506e-05,
+      "loss": 1.1571,
+      "step": 730
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 4.41500438496937e-05,
+      "loss": 0.9996,
+      "step": 731
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 4.3372296508342026e-05,
+      "loss": 1.0229,
+      "step": 732
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 4.2599559107388645e-05,
+      "loss": 0.9749,
+      "step": 733
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 4.1831900012105105e-05,
+      "loss": 0.9654,
+      "step": 734
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 4.10693871384773e-05,
+      "loss": 1.157,
+      "step": 735
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 4.031208794719695e-05,
+      "loss": 1.1395,
+      "step": 736
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 3.95600694376933e-05,
+      "loss": 0.9787,
+      "step": 737
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 3.881339814220555e-05,
+      "loss": 1.1573,
+      "step": 738
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 3.8072140119896504e-05,
+      "loss": 1.0817,
+      "step": 739
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 3.733636095100838e-05,
+      "loss": 1.0108,
+      "step": 740
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 3.660612573106081e-05,
+      "loss": 1.0674,
+      "step": 741
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 3.588149906509166e-05,
+      "loss": 1.0425,
+      "step": 742
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 3.5162545061941335e-05,
+      "loss": 0.9176,
+      "step": 743
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 3.444932732858107e-05,
+      "loss": 0.949,
+      "step": 744
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 3.3741908964485414e-05,
+      "loss": 1.0735,
+      "step": 745
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 3.304035255604974e-05,
+      "loss": 1.0489,
+      "step": 746
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 3.234472017105313e-05,
+      "loss": 1.014,
+      "step": 747
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 3.1655073353167285e-05,
+      "loss": 1.0575,
+      "step": 748
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 3.0971473116511394e-05,
+      "loss": 1.0651,
+      "step": 749
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 3.0293979940254314e-05,
+      "loss": 1.038,
+      "step": 750
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 2.9622653763263874e-05,
+      "loss": 1.0974,
+      "step": 751
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 2.8957553978803898e-05,
+      "loss": 1.087,
+      "step": 752
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 2.8298739429279707e-05,
+      "loss": 1.1009,
+      "step": 753
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 2.764626840103208e-05,
+      "loss": 1.0945,
+      "step": 754
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 2.7000198619180794e-05,
+      "loss": 1.0436,
+      "step": 755
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 2.636058724251739e-05,
+      "loss": 1.1178,
+      "step": 756
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 2.5727490858448288e-05,
+      "loss": 1.0239,
+      "step": 757
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 2.510096547798857e-05,
+      "loss": 1.0082,
+      "step": 758
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 2.4481066530806395e-05,
+      "loss": 0.9967,
+      "step": 759
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 2.3867848860319144e-05,
+      "loss": 1.0141,
+      "step": 760
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 2.3261366718841305e-05,
+      "loss": 1.0257,
+      "step": 761
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 2.2661673762784807e-05,
+      "loss": 1.0973,
+      "step": 762
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 2.206882304791176e-05,
+      "loss": 0.9521,
+      "step": 763
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 2.148286702464063e-05,
+      "loss": 1.0711,
+      "step": 764
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 2.0903857533405958e-05,
+      "loss": 0.9501,
+      "step": 765
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 2.033184580007177e-05,
+      "loss": 1.016,
+      "step": 766
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 1.9766882431399646e-05,
+      "loss": 1.0498,
+      "step": 767
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 1.920901741057153e-05,
+      "loss": 1.1671,
+      "step": 768
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 1.8658300092767544e-05,
+      "loss": 1.0246,
+      "step": 769
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 1.8114779200799437e-05,
+      "loss": 0.9781,
+      "step": 770
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 1.7578502820800045e-05,
+      "loss": 1.0746,
+      "step": 771
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 1.7049518397969132e-05,
+      "loss": 1.0822,
+      "step": 772
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 1.652787273237565e-05,
+      "loss": 1.0571,
+      "step": 773
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 1.601361197481741e-05,
+      "loss": 0.9481,
+      "step": 774
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 1.5506781622737942e-05,
+      "loss": 1.0077,
+      "step": 775
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 1.500742651620145e-05,
+      "loss": 1.0027,
+      "step": 776
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 1.4515590833925507e-05,
+      "loss": 0.9245,
+      "step": 777
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 1.4031318089372615e-05,
+      "loss": 1.2058,
+      "step": 778
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 1.3554651126900564e-05,
+      "loss": 1.1016,
+      "step": 779
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 1.308563211797179e-05,
+      "loss": 1.1456,
+      "step": 780
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 1.2624302557422473e-05,
+      "loss": 1.0946,
+      "step": 781
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 1.2170703259791471e-05,
+      "loss": 1.2343,
+      "step": 782
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 1.172487435570926e-05,
+      "loss": 1.006,
+      "step": 783
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 1.1286855288347675e-05,
+      "loss": 1.1386,
+      "step": 784
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 1.0856684809930151e-05,
+      "loss": 0.9418,
+      "step": 785
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 1.0434400978303426e-05,
+      "loss": 1.1213,
+      "step": 786
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 1.0020041153570347e-05,
+      "loss": 0.9317,
+      "step": 787
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 9.613641994784651e-06,
+      "loss": 1.0073,
+      "step": 788
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 9.215239456707635e-06,
+      "loss": 1.0513,
+      "step": 789
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 8.824868786627304e-06,
+      "loss": 1.1849,
+      "step": 790
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 8.442564521239782e-06,
+      "loss": 1.0606,
+      "step": 791
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 8.06836048359394e-06,
+      "loss": 0.9558,
+      "step": 792
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 7.70228978009907e-06,
+      "loss": 1.0529,
+      "step": 793
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 7.344384797595716e-06,
+      "loss": 0.9983,
+      "step": 794
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 6.994677200490507e-06,
+      "loss": 0.9064,
+      "step": 795
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 6.653197927954791e-06,
+      "loss": 1.0392,
+      "step": 796
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 6.319977191187232e-06,
+      "loss": 1.0478,
+      "step": 797
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 5.995044470741151e-06,
+      "loss": 1.0528,
+      "step": 798
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 5.678428513916212e-06,
+      "loss": 1.0008,
+      "step": 799
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 5.370157332215209e-06,
+      "loss": 1.0729,
+      "step": 800
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 5.07025819886574e-06,
+      "loss": 1.0285,
+      "step": 801
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 4.778757646407362e-06,
+      "loss": 1.0473,
+      "step": 802
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 4.495681464344259e-06,
+      "loss": 1.0021,
+      "step": 803
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 4.22105469686348e-06,
+      "loss": 1.0449,
+      "step": 804
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 3.954901640619368e-06,
+      "loss": 0.9546,
+      "step": 805
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 3.6972458425838962e-06,
+      "loss": 1.0409,
+      "step": 806
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 3.4481100979635306e-06,
+      "loss": 1.0518,
+      "step": 807
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 3.207516448182435e-06,
+      "loss": 0.9276,
+      "step": 808
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 2.9754861789324073e-06,
+      "loss": 1.0274,
+      "step": 809
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 2.752039818289809e-06,
+      "loss": 1.1427,
+      "step": 810
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 2.537197134899294e-06,
+      "loss": 0.957,
+      "step": 811
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 2.330977136224932e-06,
+      "loss": 1.0667,
+      "step": 812
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 2.1333980668685414e-06,
+      "loss": 1.0124,
+      "step": 813
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 1.9444774069555694e-06,
+      "loss": 1.012,
+      "step": 814
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 1.7642318705886286e-06,
+      "loss": 0.9649,
+      "step": 815
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 1.5926774043687365e-06,
+      "loss": 1.1151,
+      "step": 816
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 1.4298291859845214e-06,
+      "loss": 1.0556,
+      "step": 817
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 1.2757016228693964e-06,
+      "loss": 1.0976,
+      "step": 818
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 1.1303083509269452e-06,
+      "loss": 1.0273,
+      "step": 819
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 9.936622333245104e-07,
+      "loss": 1.0017,
+      "step": 820
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 8.657753593552143e-07,
+      "loss": 1.1785,
+      "step": 821
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 7.466590433683251e-07,
+      "loss": 1.12,
+      "step": 822
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 6.363238237683033e-07,
+      "loss": 1.0588,
+      "step": 823
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 5.347794620824576e-07,
+      "loss": 1.1339,
+      "step": 824
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 4.4203494209733576e-07,
+      "loss": 0.9589,
+      "step": 825
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 835,
+  "num_train_epochs": 5,
+  "save_steps": 15,
+  "total_flos": 1.2359419310859878e+17,
+  "trial_name": null,
+  "trial_params": null
+}