diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4900 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9986154911160678,
+  "eval_steps": 500,
+  "global_step": 812,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0,
+      "loss": 1.5218,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.306765580733931e-06,
+      "loss": 1.4152,
+      "step": 2
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.826061944859854e-06,
+      "loss": 1.6324,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.613531161467863e-06,
+      "loss": 1.4134,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1e-05,
+      "loss": 1.3908,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1132827525593786e-05,
+      "loss": 1.3776,
+      "step": 6
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2090619551221676e-05,
+      "loss": 1.4819,
+      "step": 7
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2920296742201793e-05,
+      "loss": 1.4535,
+      "step": 8
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3652123889719709e-05,
+      "loss": 1.5043,
+      "step": 9
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4306765580733935e-05,
+      "loss": 1.378,
+      "step": 10
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.4898961024049785e-05,
+      "loss": 1.4313,
+      "step": 11
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.543959310632772e-05,
+      "loss": 1.4704,
+      "step": 12
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.5936926411670824e-05,
+      "loss": 1.4381,
+      "step": 13
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.639738513195561e-05,
+      "loss": 1.491,
+      "step": 14
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.6826061944859853e-05,
+      "loss": 1.4661,
+      "step": 15
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.7227062322935725e-05,
+      "loss": 1.505,
+      "step": 16
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.7603744277225883e-05,
+      "loss": 1.4474,
+      "step": 17
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.7958889470453637e-05,
+      "loss": 1.4598,
+      "step": 18
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.8294828004351506e-05,
+      "loss": 1.4767,
+      "step": 19
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.8613531161467863e-05,
+      "loss": 1.4241,
+      "step": 20
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.891668149608153e-05,
+      "loss": 1.5017,
+      "step": 21
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.9205726604783716e-05,
+      "loss": 1.4871,
+      "step": 22
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.94819209346638e-05,
+      "loss": 1.5035,
+      "step": 23
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9746358687061647e-05,
+      "loss": 1.4263,
+      "step": 24
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2e-05,
+      "loss": 1.3643,
+      "step": 25
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2e-05,
+      "loss": 1.4116,
+      "step": 26
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.997458703939009e-05,
+      "loss": 1.4073,
+      "step": 27
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.994917407878018e-05,
+      "loss": 1.3866,
+      "step": 28
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.992376111817027e-05,
+      "loss": 1.3919,
+      "step": 29
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.9898348157560356e-05,
+      "loss": 1.3814,
+      "step": 30
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9872935196950447e-05,
+      "loss": 1.3982,
+      "step": 31
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9847522236340535e-05,
+      "loss": 1.4264,
+      "step": 32
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9822109275730623e-05,
+      "loss": 1.431,
+      "step": 33
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9796696315120714e-05,
+      "loss": 1.4176,
+      "step": 34
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9771283354510802e-05,
+      "loss": 1.3248,
+      "step": 35
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.974587039390089e-05,
+      "loss": 1.4244,
+      "step": 36
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.972045743329098e-05,
+      "loss": 1.3757,
+      "step": 37
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.969504447268107e-05,
+      "loss": 1.2918,
+      "step": 38
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.966963151207116e-05,
+      "loss": 1.3476,
+      "step": 39
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9644218551461248e-05,
+      "loss": 1.3144,
+      "step": 40
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9618805590851335e-05,
+      "loss": 1.3452,
+      "step": 41
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9593392630241426e-05,
+      "loss": 1.3085,
+      "step": 42
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9567979669631514e-05,
+      "loss": 1.3217,
+      "step": 43
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9542566709021602e-05,
+      "loss": 1.3127,
+      "step": 44
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9517153748411693e-05,
+      "loss": 1.2837,
+      "step": 45
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.949174078780178e-05,
+      "loss": 1.265,
+      "step": 46
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.946632782719187e-05,
+      "loss": 1.3112,
+      "step": 47
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.944091486658196e-05,
+      "loss": 1.2877,
+      "step": 48
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9415501905972048e-05,
+      "loss": 1.247,
+      "step": 49
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9390088945362135e-05,
+      "loss": 1.2566,
+      "step": 50
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9364675984752227e-05,
+      "loss": 1.2455,
+      "step": 51
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9339263024142314e-05,
+      "loss": 1.2432,
+      "step": 52
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9313850063532402e-05,
+      "loss": 1.2218,
+      "step": 53
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9288437102922493e-05,
+      "loss": 1.2129,
+      "step": 54
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.926302414231258e-05,
+      "loss": 1.2442,
+      "step": 55
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.923761118170267e-05,
+      "loss": 1.2742,
+      "step": 56
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.921219822109276e-05,
+      "loss": 1.2181,
+      "step": 57
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9186785260482848e-05,
+      "loss": 1.2066,
+      "step": 58
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9161372299872936e-05,
+      "loss": 1.2478,
+      "step": 59
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9135959339263027e-05,
+      "loss": 1.2087,
+      "step": 60
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9110546378653115e-05,
+      "loss": 1.2431,
+      "step": 61
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9085133418043206e-05,
+      "loss": 1.1892,
+      "step": 62
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9059720457433294e-05,
+      "loss": 1.1854,
+      "step": 63
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.903430749682338e-05,
+      "loss": 1.2251,
+      "step": 64
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9008894536213473e-05,
+      "loss": 1.2441,
+      "step": 65
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.898348157560356e-05,
+      "loss": 1.2214,
+      "step": 66
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.8958068614993648e-05,
+      "loss": 1.1892,
+      "step": 67
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.893265565438374e-05,
+      "loss": 1.2177,
+      "step": 68
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.8907242693773827e-05,
+      "loss": 1.1768,
+      "step": 69
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.8881829733163915e-05,
+      "loss": 1.1923,
+      "step": 70
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.8856416772554006e-05,
+      "loss": 1.1698,
+      "step": 71
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.8831003811944094e-05,
+      "loss": 1.0962,
+      "step": 72
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.880559085133418e-05,
+      "loss": 1.2324,
+      "step": 73
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.8780177890724273e-05,
+      "loss": 1.1119,
+      "step": 74
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.875476493011436e-05,
+      "loss": 1.1775,
+      "step": 75
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.8729351969504448e-05,
+      "loss": 1.177,
+      "step": 76
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.870393900889454e-05,
+      "loss": 1.1707,
+      "step": 77
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.8678526048284627e-05,
+      "loss": 1.2316,
+      "step": 78
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.8653113087674715e-05,
+      "loss": 1.1577,
+      "step": 79
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8627700127064806e-05,
+      "loss": 1.1703,
+      "step": 80
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8602287166454894e-05,
+      "loss": 1.1594,
+      "step": 81
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8576874205844982e-05,
+      "loss": 1.1533,
+      "step": 82
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8551461245235073e-05,
+      "loss": 1.1837,
+      "step": 83
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.852604828462516e-05,
+      "loss": 1.1109,
+      "step": 84
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8500635324015252e-05,
+      "loss": 1.1358,
+      "step": 85
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.847522236340534e-05,
+      "loss": 1.1166,
+      "step": 86
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8449809402795427e-05,
+      "loss": 1.1473,
+      "step": 87
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.842439644218552e-05,
+      "loss": 1.0967,
+      "step": 88
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.8398983481575606e-05,
+      "loss": 1.1847,
+      "step": 89
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.8373570520965694e-05,
+      "loss": 1.0904,
+      "step": 90
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.8348157560355785e-05,
+      "loss": 1.1095,
+      "step": 91
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8322744599745873e-05,
+      "loss": 1.1249,
+      "step": 92
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.829733163913596e-05,
+      "loss": 1.1636,
+      "step": 93
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8271918678526052e-05,
+      "loss": 1.1173,
+      "step": 94
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.824650571791614e-05,
+      "loss": 1.107,
+      "step": 95
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8221092757306228e-05,
+      "loss": 1.1126,
+      "step": 96
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.819567979669632e-05,
+      "loss": 1.076,
+      "step": 97
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8170266836086407e-05,
+      "loss": 1.0646,
+      "step": 98
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8144853875476494e-05,
+      "loss": 1.0878,
+      "step": 99
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8119440914866586e-05,
+      "loss": 1.0689,
+      "step": 100
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8094027954256673e-05,
+      "loss": 1.0592,
+      "step": 101
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.806861499364676e-05,
+      "loss": 1.0954,
+      "step": 102
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8043202033036852e-05,
+      "loss": 1.1227,
+      "step": 103
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.801778907242694e-05,
+      "loss": 1.1568,
+      "step": 104
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7992376111817028e-05,
+      "loss": 1.0799,
+      "step": 105
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.796696315120712e-05,
+      "loss": 1.0771,
+      "step": 106
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7941550190597207e-05,
+      "loss": 1.0117,
+      "step": 107
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.7916137229987295e-05,
+      "loss": 1.1364,
+      "step": 108
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.7890724269377386e-05,
+      "loss": 1.047,
+      "step": 109
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.7865311308767473e-05,
+      "loss": 1.1267,
+      "step": 110
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.783989834815756e-05,
+      "loss": 1.0693,
+      "step": 111
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.7814485387547652e-05,
+      "loss": 1.0369,
+      "step": 112
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.778907242693774e-05,
+      "loss": 1.0769,
+      "step": 113
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.7763659466327828e-05,
+      "loss": 1.0858,
+      "step": 114
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.773824650571792e-05,
+      "loss": 1.061,
+      "step": 115
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7712833545108007e-05,
+      "loss": 1.1051,
+      "step": 116
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7687420584498095e-05,
+      "loss": 1.1002,
+      "step": 117
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7662007623888186e-05,
+      "loss": 1.0675,
+      "step": 118
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7636594663278274e-05,
+      "loss": 1.1439,
+      "step": 119
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.761118170266836e-05,
+      "loss": 1.0421,
+      "step": 120
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7585768742058453e-05,
+      "loss": 1.0826,
+      "step": 121
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.756035578144854e-05,
+      "loss": 1.0945,
+      "step": 122
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7534942820838628e-05,
+      "loss": 1.0928,
+      "step": 123
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.750952986022872e-05,
+      "loss": 1.0491,
+      "step": 124
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7484116899618807e-05,
+      "loss": 1.0479,
+      "step": 125
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7458703939008895e-05,
+      "loss": 1.0354,
+      "step": 126
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7433290978398986e-05,
+      "loss": 1.0821,
+      "step": 127
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7407878017789074e-05,
+      "loss": 1.0827,
+      "step": 128
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.738246505717916e-05,
+      "loss": 1.0352,
+      "step": 129
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7357052096569253e-05,
+      "loss": 1.0851,
+      "step": 130
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.733163913595934e-05,
+      "loss": 1.0679,
+      "step": 131
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.730622617534943e-05,
+      "loss": 1.0779,
+      "step": 132
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.728081321473952e-05,
+      "loss": 1.0398,
+      "step": 133
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7255400254129607e-05,
+      "loss": 1.0511,
+      "step": 134
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7229987293519695e-05,
+      "loss": 1.0495,
+      "step": 135
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7204574332909786e-05,
+      "loss": 1.0217,
+      "step": 136
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.7179161372299874e-05,
+      "loss": 1.0381,
+      "step": 137
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.7153748411689962e-05,
+      "loss": 1.0974,
+      "step": 138
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.7128335451080053e-05,
+      "loss": 1.0635,
+      "step": 139
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.710292249047014e-05,
+      "loss": 1.0068,
+      "step": 140
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.707750952986023e-05,
+      "loss": 1.1153,
+      "step": 141
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.705209656925032e-05,
+      "loss": 1.0512,
+      "step": 142
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.7026683608640408e-05,
+      "loss": 1.0265,
+      "step": 143
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.7001270648030495e-05,
+      "loss": 1.0824,
+      "step": 144
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6975857687420586e-05,
+      "loss": 0.9982,
+      "step": 145
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6950444726810674e-05,
+      "loss": 1.1003,
+      "step": 146
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6925031766200762e-05,
+      "loss": 1.0906,
+      "step": 147
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6899618805590853e-05,
+      "loss": 1.0872,
+      "step": 148
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.687420584498094e-05,
+      "loss": 1.0271,
+      "step": 149
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.684879288437103e-05,
+      "loss": 1.0468,
+      "step": 150
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.682337992376112e-05,
+      "loss": 1.0058,
+      "step": 151
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.6797966963151208e-05,
+      "loss": 1.0341,
+      "step": 152
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.6772554002541296e-05,
+      "loss": 1.0171,
+      "step": 153
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.6747141041931387e-05,
+      "loss": 1.1015,
+      "step": 154
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.6721728081321474e-05,
+      "loss": 1.0363,
+      "step": 155
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.6696315120711562e-05,
+      "loss": 1.0537,
+      "step": 156
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.6670902160101653e-05,
+      "loss": 1.0612,
+      "step": 157
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.664548919949174e-05,
+      "loss": 1.0243,
+      "step": 158
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.662007623888183e-05,
+      "loss": 1.0702,
+      "step": 159
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.659466327827192e-05,
+      "loss": 1.0049,
+      "step": 160
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.6569250317662008e-05,
+      "loss": 1.002,
+      "step": 161
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.6543837357052096e-05,
+      "loss": 1.0191,
+      "step": 162
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.6518424396442187e-05,
+      "loss": 1.0291,
+      "step": 163
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.6493011435832275e-05,
+      "loss": 1.0008,
+      "step": 164
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.6467598475222362e-05,
+      "loss": 1.0225,
+      "step": 165
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.6442185514612454e-05,
+      "loss": 0.9738,
+      "step": 166
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.641677255400254e-05,
+      "loss": 1.0618,
+      "step": 167
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.639135959339263e-05,
+      "loss": 1.0028,
+      "step": 168
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.636594663278272e-05,
+      "loss": 0.9994,
+      "step": 169
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.6340533672172808e-05,
+      "loss": 1.0041,
+      "step": 170
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.6315120711562896e-05,
+      "loss": 1.0078,
+      "step": 171
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.6289707750952987e-05,
+      "loss": 0.9461,
+      "step": 172
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.6264294790343075e-05,
+      "loss": 1.0427,
+      "step": 173
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.6238881829733166e-05,
+      "loss": 0.9973,
+      "step": 174
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.6213468869123254e-05,
+      "loss": 1.0302,
+      "step": 175
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.618805590851334e-05,
+      "loss": 1.0448,
+      "step": 176
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.6162642947903433e-05,
+      "loss": 0.9953,
+      "step": 177
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.613722998729352e-05,
+      "loss": 1.0202,
+      "step": 178
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.6111817026683608e-05,
+      "loss": 1.0182,
+      "step": 179
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.60864040660737e-05,
+      "loss": 1.0448,
+      "step": 180
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.6060991105463787e-05,
+      "loss": 1.0158,
+      "step": 181
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.6035578144853875e-05,
+      "loss": 1.0064,
+      "step": 182
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.6010165184243966e-05,
+      "loss": 0.9783,
+      "step": 183
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.5984752223634054e-05,
+      "loss": 1.0154,
+      "step": 184
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.5959339263024142e-05,
+      "loss": 1.0082,
+      "step": 185
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.5933926302414233e-05,
+      "loss": 1.0565,
+      "step": 186
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.590851334180432e-05,
+      "loss": 0.9971,
+      "step": 187
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.588310038119441e-05,
+      "loss": 0.9979,
+      "step": 188
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.58576874205845e-05,
+      "loss": 0.9809,
+      "step": 189
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.5832274459974587e-05,
+      "loss": 0.9936,
+      "step": 190
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.5806861499364675e-05,
+      "loss": 1.07,
+      "step": 191
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.5781448538754766e-05,
+      "loss": 1.0522,
+      "step": 192
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.5756035578144854e-05,
+      "loss": 1.0442,
+      "step": 193
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.5730622617534942e-05,
+      "loss": 1.004,
+      "step": 194
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.5705209656925033e-05,
+      "loss": 0.9988,
+      "step": 195
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.567979669631512e-05,
+      "loss": 1.0105,
+      "step": 196
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.5654383735705212e-05,
+      "loss": 0.9885,
+      "step": 197
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.56289707750953e-05,
+      "loss": 1.0738,
+      "step": 198
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.5603557814485388e-05,
+      "loss": 1.0308,
+      "step": 199
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.557814485387548e-05,
+      "loss": 1.0519,
+      "step": 200
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.5552731893265567e-05,
+      "loss": 0.9842,
+      "step": 201
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.5527318932655654e-05,
+      "loss": 1.0244,
+      "step": 202
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.5501905972045746e-05,
+      "loss": 1.0619,
+      "step": 203
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.5476493011435833e-05,
+      "loss": 1.0231,
+      "step": 204
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.545108005082592e-05,
+      "loss": 1.0014,
+      "step": 205
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.5425667090216012e-05,
+      "loss": 0.9918,
+      "step": 206
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.54002541296061e-05,
+      "loss": 0.9899,
+      "step": 207
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.5374841168996188e-05,
+      "loss": 0.9915,
+      "step": 208
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.534942820838628e-05,
+      "loss": 0.9876,
+      "step": 209
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.5324015247776367e-05,
+      "loss": 0.9486,
+      "step": 210
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.5298602287166455e-05,
+      "loss": 1.044,
+      "step": 211
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.5273189326556546e-05,
+      "loss": 0.9792,
+      "step": 212
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.5247776365946634e-05,
+      "loss": 0.9666,
+      "step": 213
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.5222363405336721e-05,
+      "loss": 1.0225,
+      "step": 214
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.5196950444726812e-05,
+      "loss": 0.93,
+      "step": 215
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.51715374841169e-05,
+      "loss": 1.0136,
+      "step": 216
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.5146124523506988e-05,
+      "loss": 1.0384,
+      "step": 217
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.512071156289708e-05,
+      "loss": 1.0185,
+      "step": 218
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.5095298602287167e-05,
+      "loss": 1.0078,
+      "step": 219
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.5069885641677258e-05,
+      "loss": 1.0217,
+      "step": 220
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.5044472681067346e-05,
+      "loss": 0.9563,
+      "step": 221
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.5019059720457434e-05,
+      "loss": 0.9469,
+      "step": 222
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.4993646759847525e-05,
+      "loss": 1.0142,
+      "step": 223
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.4968233799237613e-05,
+      "loss": 0.9781,
+      "step": 224
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.49428208386277e-05,
+      "loss": 0.9935,
+      "step": 225
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.4917407878017792e-05,
+      "loss": 1.01,
+      "step": 226
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.489199491740788e-05,
+      "loss": 1.0315,
+      "step": 227
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.4866581956797967e-05,
+      "loss": 0.9434,
+      "step": 228
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.4841168996188058e-05,
+      "loss": 0.9923,
+      "step": 229
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.4815756035578146e-05,
+      "loss": 0.9665,
+      "step": 230
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.4790343074968234e-05,
+      "loss": 1.031,
+      "step": 231
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.4764930114358325e-05,
+      "loss": 0.9442,
+      "step": 232
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.4739517153748413e-05,
+      "loss": 0.9624,
+      "step": 233
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.47141041931385e-05,
+      "loss": 0.9832,
+      "step": 234
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.4688691232528592e-05,
+      "loss": 1.0166,
+      "step": 235
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.466327827191868e-05,
+      "loss": 1.0162,
+      "step": 236
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.4637865311308767e-05,
+      "loss": 1.0158,
+      "step": 237
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.4612452350698859e-05,
+      "loss": 0.969,
+      "step": 238
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.4587039390088946e-05,
+      "loss": 1.0131,
+      "step": 239
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.4561626429479036e-05,
+      "loss": 1.0364,
+      "step": 240
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.4536213468869125e-05,
+      "loss": 0.9988,
+      "step": 241
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.4510800508259213e-05,
+      "loss": 0.967,
+      "step": 242
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.4485387547649303e-05,
+      "loss": 1.0422,
+      "step": 243
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.4459974587039392e-05,
+      "loss": 1.0578,
+      "step": 244
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.443456162642948e-05,
+      "loss": 0.9673,
+      "step": 245
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.440914866581957e-05,
+      "loss": 0.9412,
+      "step": 246
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.4383735705209659e-05,
+      "loss": 0.9383,
+      "step": 247
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.4358322744599746e-05,
+      "loss": 0.994,
+      "step": 248
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.4332909783989836e-05,
+      "loss": 0.9997,
+      "step": 249
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.4307496823379925e-05,
+      "loss": 0.9917,
+      "step": 250
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.4282083862770013e-05,
+      "loss": 0.9974,
+      "step": 251
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.4256670902160103e-05,
+      "loss": 0.9607,
+      "step": 252
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.4231257941550192e-05,
+      "loss": 0.9606,
+      "step": 253
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.420584498094028e-05,
+      "loss": 0.9918,
+      "step": 254
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.418043202033037e-05,
+      "loss": 0.9764,
+      "step": 255
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.4155019059720459e-05,
+      "loss": 1.0094,
+      "step": 256
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.4129606099110547e-05,
+      "loss": 0.9528,
+      "step": 257
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.4104193138500636e-05,
+      "loss": 1.0268,
+      "step": 258
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.4078780177890726e-05,
+      "loss": 0.9968,
+      "step": 259
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.4053367217280813e-05,
+      "loss": 1.0086,
+      "step": 260
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.4027954256670903e-05,
+      "loss": 0.9582,
+      "step": 261
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.4002541296060992e-05,
+      "loss": 0.968,
+      "step": 262
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.3977128335451082e-05,
+      "loss": 1.016,
+      "step": 263
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.395171537484117e-05,
+      "loss": 0.9361,
+      "step": 264
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.3926302414231259e-05,
+      "loss": 0.9922,
+      "step": 265
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.3900889453621349e-05,
+      "loss": 0.9913,
+      "step": 266
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.3875476493011436e-05,
+      "loss": 1.0117,
+      "step": 267
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.3850063532401526e-05,
+      "loss": 1.0134,
+      "step": 268
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.3824650571791615e-05,
+      "loss": 1.0612,
+      "step": 269
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.3799237611181703e-05,
+      "loss": 1.0274,
+      "step": 270
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.3773824650571793e-05,
+      "loss": 1.0323,
+      "step": 271
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.3748411689961882e-05,
+      "loss": 0.9686,
+      "step": 272
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.372299872935197e-05,
+      "loss": 0.9759,
+      "step": 273
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.369758576874206e-05,
+      "loss": 0.9758,
+      "step": 274
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.3672172808132149e-05,
+      "loss": 0.984,
+      "step": 275
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.3646759847522237e-05,
+      "loss": 0.9578,
+      "step": 276
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.3621346886912326e-05,
+      "loss": 0.9805,
+      "step": 277
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.3595933926302415e-05,
+      "loss": 0.962,
+      "step": 278
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.3570520965692503e-05,
+      "loss": 0.9359,
+      "step": 279
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.3545108005082593e-05,
+      "loss": 0.9394,
+      "step": 280
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.3519695044472682e-05,
+      "loss": 0.9663,
+      "step": 281
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.349428208386277e-05,
+      "loss": 0.9542,
+      "step": 282
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.346886912325286e-05,
+      "loss": 0.9816,
+      "step": 283
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.3443456162642949e-05,
+      "loss": 0.9832,
+      "step": 284
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.3418043202033037e-05,
+      "loss": 0.9679,
+      "step": 285
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.3392630241423128e-05,
+      "loss": 0.9745,
+      "step": 286
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.3367217280813216e-05,
+      "loss": 0.9674,
+      "step": 287
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.3341804320203303e-05,
+      "loss": 1.005,
+      "step": 288
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.3316391359593395e-05,
+      "loss": 1.0021,
+      "step": 289
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.3290978398983482e-05,
+      "loss": 1.0045,
+      "step": 290
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.326556543837357e-05,
+      "loss": 0.9755,
+      "step": 291
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.3240152477763661e-05,
+      "loss": 0.9835,
+      "step": 292
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.3214739517153749e-05,
+      "loss": 0.9675,
+      "step": 293
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.3189326556543837e-05,
+      "loss": 0.9747,
+      "step": 294
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.3163913595933928e-05,
+      "loss": 0.9955,
+      "step": 295
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.3138500635324016e-05,
+      "loss": 0.9177,
+      "step": 296
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.3113087674714104e-05,
+      "loss": 0.989,
+      "step": 297
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.3087674714104195e-05,
+      "loss": 0.975,
+      "step": 298
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.3062261753494283e-05,
+      "loss": 0.9778,
+      "step": 299
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.303684879288437e-05,
+      "loss": 0.9824,
+      "step": 300
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.3011435832274462e-05,
+      "loss": 0.9465,
+      "step": 301
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.298602287166455e-05,
+      "loss": 0.9677,
+      "step": 302
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2960609911054637e-05,
+      "loss": 0.9724,
+      "step": 303
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2935196950444728e-05,
+      "loss": 0.9715,
+      "step": 304
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2909783989834816e-05,
+      "loss": 1.0021,
+      "step": 305
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2884371029224904e-05,
+      "loss": 1.0122,
+      "step": 306
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2858958068614995e-05,
+      "loss": 0.9257,
+      "step": 307
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2833545108005083e-05,
+      "loss": 0.9902,
+      "step": 308
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2808132147395174e-05,
+      "loss": 1.0042,
+      "step": 309
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2782719186785262e-05,
+      "loss": 0.9734,
+      "step": 310
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.275730622617535e-05,
+      "loss": 1.0252,
+      "step": 311
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.273189326556544e-05,
+      "loss": 0.9737,
+      "step": 312
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2706480304955528e-05,
+      "loss": 1.0046,
+      "step": 313
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2681067344345616e-05,
+      "loss": 0.9398,
+      "step": 314
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2655654383735707e-05,
+      "loss": 0.9513,
+      "step": 315
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2630241423125795e-05,
+      "loss": 0.9746,
+      "step": 316
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2604828462515883e-05,
+      "loss": 1.0342,
+      "step": 317
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2579415501905974e-05,
+      "loss": 0.9126,
+      "step": 318
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.2554002541296062e-05,
+      "loss": 0.9317,
+      "step": 319
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.252858958068615e-05,
+      "loss": 0.9586,
+      "step": 320
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.2503176620076241e-05,
+      "loss": 0.9955,
+      "step": 321
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.2477763659466329e-05,
+      "loss": 0.9629,
+      "step": 322
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.2452350698856416e-05,
+      "loss": 0.9597,
+      "step": 323
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.2426937738246508e-05,
+      "loss": 0.9521,
+      "step": 324
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.2401524777636595e-05,
+      "loss": 0.9956,
+      "step": 325
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.2376111817026683e-05,
+      "loss": 0.9297,
+      "step": 326
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.2350698856416774e-05,
+      "loss": 0.9846,
+      "step": 327
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.2325285895806862e-05,
+      "loss": 0.9609,
+      "step": 328
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.229987293519695e-05,
+      "loss": 0.9936,
+      "step": 329
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.2274459974587041e-05,
+      "loss": 0.9665,
+      "step": 330
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.2249047013977129e-05,
+      "loss": 0.9467,
+      "step": 331
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.222363405336722e-05,
+      "loss": 0.9862,
+      "step": 332
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.2198221092757308e-05,
+      "loss": 0.9242,
+      "step": 333
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.2172808132147396e-05,
+      "loss": 0.9865,
+      "step": 334
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.2147395171537487e-05,
+      "loss": 0.965,
+      "step": 335
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.2121982210927575e-05,
+      "loss": 0.9514,
+      "step": 336
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.2096569250317662e-05,
+      "loss": 0.9771,
+      "step": 337
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.2071156289707754e-05,
+      "loss": 0.9881,
+      "step": 338
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.2045743329097841e-05,
+      "loss": 0.9393,
+      "step": 339
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.2020330368487929e-05,
+      "loss": 0.9873,
+      "step": 340
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.199491740787802e-05,
+      "loss": 0.9837,
+      "step": 341
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.1969504447268108e-05,
+      "loss": 0.9493,
+      "step": 342
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.1944091486658196e-05,
+      "loss": 0.9564,
+      "step": 343
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1918678526048287e-05,
+      "loss": 0.9185,
+      "step": 344
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1893265565438375e-05,
+      "loss": 0.9294,
+      "step": 345
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1867852604828463e-05,
+      "loss": 0.9647,
+      "step": 346
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1842439644218554e-05,
+      "loss": 0.9193,
+      "step": 347
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.1817026683608641e-05,
+      "loss": 0.937,
+      "step": 348
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.179161372299873e-05,
+      "loss": 0.9998,
+      "step": 349
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.176620076238882e-05,
+      "loss": 0.906,
+      "step": 350
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.1740787801778908e-05,
+      "loss": 0.9507,
+      "step": 351
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.1715374841168996e-05,
+      "loss": 0.9787,
+      "step": 352
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.1689961880559087e-05,
+      "loss": 0.9426,
+      "step": 353
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.1664548919949175e-05,
+      "loss": 0.9246,
+      "step": 354
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.1639135959339264e-05,
+      "loss": 0.902,
+      "step": 355
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.1613722998729354e-05,
+      "loss": 0.9401,
+      "step": 356
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.1588310038119442e-05,
+      "loss": 0.9716,
+      "step": 357
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.1562897077509531e-05,
+      "loss": 0.9306,
+      "step": 358
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.153748411689962e-05,
+      "loss": 0.9665,
+      "step": 359
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.1512071156289708e-05,
+      "loss": 0.9744,
+      "step": 360
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.1486658195679798e-05,
+      "loss": 1.025,
+      "step": 361
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.1461245235069887e-05,
+      "loss": 1.0,
+      "step": 362
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.1435832274459975e-05,
+      "loss": 0.9639,
+      "step": 363
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.1410419313850065e-05,
+      "loss": 0.9715,
+      "step": 364
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.1385006353240154e-05,
+      "loss": 0.9995,
+      "step": 365
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.1359593392630242e-05,
+      "loss": 0.9575,
+      "step": 366
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.1334180432020331e-05,
+      "loss": 0.9676,
+      "step": 367
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.130876747141042e-05,
+      "loss": 0.9775,
+      "step": 368
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.1283354510800509e-05,
+      "loss": 0.9493,
+      "step": 369
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.1257941550190598e-05,
+      "loss": 0.9692,
+      "step": 370
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.1232528589580688e-05,
+      "loss": 0.9937,
+      "step": 371
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.1207115628970775e-05,
+      "loss": 0.966,
+      "step": 372
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.1181702668360865e-05,
+      "loss": 0.9173,
+      "step": 373
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.1156289707750954e-05,
+      "loss": 0.9281,
+      "step": 374
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.1130876747141044e-05,
+      "loss": 0.9438,
+      "step": 375
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.1105463786531132e-05,
+      "loss": 1.0017,
+      "step": 376
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.1080050825921221e-05,
+      "loss": 0.9927,
+      "step": 377
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.105463786531131e-05,
+      "loss": 0.9686,
+      "step": 378
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.1029224904701398e-05,
+      "loss": 0.9651,
+      "step": 379
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.1003811944091488e-05,
+      "loss": 0.9015,
+      "step": 380
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.0978398983481577e-05,
+      "loss": 0.9524,
+      "step": 381
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.0952986022871665e-05,
+      "loss": 0.9467,
+      "step": 382
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.0927573062261754e-05,
+      "loss": 0.9642,
+      "step": 383
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0902160101651844e-05,
+      "loss": 1.0092,
+      "step": 384
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0876747141041932e-05,
+      "loss": 0.9284,
+      "step": 385
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0851334180432021e-05,
+      "loss": 0.9508,
+      "step": 386
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.082592121982211e-05,
+      "loss": 1.0035,
+      "step": 387
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.0800508259212198e-05,
+      "loss": 0.9583,
+      "step": 388
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.0775095298602288e-05,
+      "loss": 0.9493,
+      "step": 389
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.0749682337992377e-05,
+      "loss": 0.9676,
+      "step": 390
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.0724269377382465e-05,
+      "loss": 0.9979,
+      "step": 391
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.0698856416772555e-05,
+      "loss": 0.9663,
+      "step": 392
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.0673443456162644e-05,
+      "loss": 0.9771,
+      "step": 393
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.0648030495552732e-05,
+      "loss": 0.946,
+      "step": 394
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.0622617534942821e-05,
+      "loss": 1.0092,
+      "step": 395
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.0597204574332911e-05,
+      "loss": 1.0279,
+      "step": 396
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.0571791613722999e-05,
+      "loss": 0.9364,
+      "step": 397
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.054637865311309e-05,
+      "loss": 0.9338,
+      "step": 398
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.0520965692503178e-05,
+      "loss": 0.9627,
+      "step": 399
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.0495552731893265e-05,
+      "loss": 1.0029,
+      "step": 400
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.0470139771283357e-05,
+      "loss": 0.8818,
+      "step": 401
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.0444726810673444e-05,
+      "loss": 0.9991,
+      "step": 402
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.0419313850063532e-05,
+      "loss": 0.9433,
+      "step": 403
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.0393900889453623e-05,
+      "loss": 0.9828,
+      "step": 404
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.0368487928843711e-05,
+      "loss": 0.9828,
+      "step": 405
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.0343074968233799e-05,
+      "loss": 0.9312,
+      "step": 406
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.031766200762389e-05,
+      "loss": 0.9585,
+      "step": 407
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.0292249047013978e-05,
+      "loss": 0.9446,
+      "step": 408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0266836086404066e-05,
+      "loss": 0.8982,
+      "step": 409
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0241423125794157e-05,
+      "loss": 0.909,
+      "step": 410
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0216010165184245e-05,
+      "loss": 0.9715,
+      "step": 411
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0190597204574332e-05,
+      "loss": 0.9471,
+      "step": 412
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0165184243964423e-05,
+      "loss": 0.9523,
+      "step": 413
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0139771283354511e-05,
+      "loss": 0.9511,
+      "step": 414
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0114358322744599e-05,
+      "loss": 0.957,
+      "step": 415
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.008894536213469e-05,
+      "loss": 0.9526,
+      "step": 416
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.0063532401524778e-05,
+      "loss": 0.9803,
+      "step": 417
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.0038119440914866e-05,
+      "loss": 0.9094,
+      "step": 418
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.0012706480304957e-05,
+      "loss": 0.9259,
+      "step": 419
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 9.987293519695045e-06,
+      "loss": 0.9676,
+      "step": 420
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.961880559085134e-06,
+      "loss": 0.9604,
+      "step": 421
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.936467598475224e-06,
+      "loss": 0.8967,
+      "step": 422
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.911054637865311e-06,
+      "loss": 1.0105,
+      "step": 423
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.885641677255401e-06,
+      "loss": 0.9664,
+      "step": 424
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 9.86022871664549e-06,
+      "loss": 0.944,
+      "step": 425
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 9.83481575603558e-06,
+      "loss": 0.9611,
+      "step": 426
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 9.809402795425668e-06,
+      "loss": 0.9506,
+      "step": 427
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 9.783989834815757e-06,
+      "loss": 0.9545,
+      "step": 428
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.758576874205847e-06,
+      "loss": 0.975,
+      "step": 429
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.733163913595934e-06,
+      "loss": 0.9443,
+      "step": 430
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.707750952986024e-06,
+      "loss": 0.9435,
+      "step": 431
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.682337992376113e-06,
+      "loss": 0.9145,
+      "step": 432
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 9.656925031766201e-06,
+      "loss": 0.9638,
+      "step": 433
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 9.63151207115629e-06,
+      "loss": 0.9475,
+      "step": 434
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 9.60609911054638e-06,
+      "loss": 0.9619,
+      "step": 435
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 9.580686149936468e-06,
+      "loss": 0.9599,
+      "step": 436
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 9.555273189326557e-06,
+      "loss": 0.9452,
+      "step": 437
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 9.529860228716647e-06,
+      "loss": 0.9427,
+      "step": 438
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 9.504447268106736e-06,
+      "loss": 0.9609,
+      "step": 439
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 9.479034307496824e-06,
+      "loss": 0.9418,
+      "step": 440
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 9.453621346886914e-06,
+      "loss": 0.966,
+      "step": 441
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 9.428208386277003e-06,
+      "loss": 0.941,
+      "step": 442
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 9.40279542566709e-06,
+      "loss": 0.9128,
+      "step": 443
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 9.37738246505718e-06,
+      "loss": 0.971,
+      "step": 444
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 9.35196950444727e-06,
+      "loss": 0.9399,
+      "step": 445
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 9.326556543837358e-06,
+      "loss": 0.8958,
+      "step": 446
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 9.301143583227447e-06,
+      "loss": 0.9843,
+      "step": 447
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 9.275730622617536e-06,
+      "loss": 0.9076,
+      "step": 448
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.250317662007626e-06,
+      "loss": 0.932,
+      "step": 449
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.224904701397714e-06,
+      "loss": 0.9099,
+      "step": 450
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.199491740787803e-06,
+      "loss": 0.9531,
+      "step": 451
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.174078780177893e-06,
+      "loss": 0.982,
+      "step": 452
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.14866581956798e-06,
+      "loss": 0.9371,
+      "step": 453
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 9.12325285895807e-06,
+      "loss": 0.9661,
+      "step": 454
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 9.09783989834816e-06,
+      "loss": 0.9949,
+      "step": 455
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 9.072426937738247e-06,
+      "loss": 0.9759,
+      "step": 456
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 9.047013977128337e-06,
+      "loss": 0.9005,
+      "step": 457
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 9.021601016518426e-06,
+      "loss": 0.9221,
+      "step": 458
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 8.996188055908514e-06,
+      "loss": 0.9768,
+      "step": 459
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 8.970775095298603e-06,
+      "loss": 0.9415,
+      "step": 460
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 8.945362134688693e-06,
+      "loss": 0.9471,
+      "step": 461
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 8.91994917407878e-06,
+      "loss": 0.9752,
+      "step": 462
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 8.89453621346887e-06,
+      "loss": 0.9277,
+      "step": 463
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 8.86912325285896e-06,
+      "loss": 0.8969,
+      "step": 464
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 8.843710292249047e-06,
+      "loss": 0.9713,
+      "step": 465
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 8.818297331639137e-06,
+      "loss": 0.92,
+      "step": 466
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 8.792884371029226e-06,
+      "loss": 0.9759,
+      "step": 467
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 8.767471410419314e-06,
+      "loss": 0.9458,
+      "step": 468
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 8.742058449809404e-06,
+      "loss": 0.9255,
+      "step": 469
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 8.716645489199493e-06,
+      "loss": 0.9493,
+      "step": 470
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 8.69123252858958e-06,
+      "loss": 0.9195,
+      "step": 471
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 8.66581956797967e-06,
+      "loss": 1.0031,
+      "step": 472
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 8.64040660736976e-06,
+      "loss": 0.9862,
+      "step": 473
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 8.614993646759848e-06,
+      "loss": 0.9288,
+      "step": 474
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 8.589580686149937e-06,
+      "loss": 0.9687,
+      "step": 475
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 8.564167725540027e-06,
+      "loss": 0.9512,
+      "step": 476
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 8.538754764930114e-06,
+      "loss": 0.8972,
+      "step": 477
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 8.513341804320204e-06,
+      "loss": 0.8961,
+      "step": 478
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 8.487928843710293e-06,
+      "loss": 0.967,
+      "step": 479
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 8.462515883100381e-06,
+      "loss": 0.963,
+      "step": 480
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 8.43710292249047e-06,
+      "loss": 0.9501,
+      "step": 481
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 8.41168996188056e-06,
+      "loss": 0.9803,
+      "step": 482
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 8.386277001270648e-06,
+      "loss": 0.9668,
+      "step": 483
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 8.360864040660737e-06,
+      "loss": 0.9123,
+      "step": 484
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 8.335451080050827e-06,
+      "loss": 0.9611,
+      "step": 485
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 8.310038119440914e-06,
+      "loss": 0.9579,
+      "step": 486
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 8.284625158831004e-06,
+      "loss": 0.9169,
+      "step": 487
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 8.259212198221093e-06,
+      "loss": 0.9201,
+      "step": 488
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 8.233799237611181e-06,
+      "loss": 0.9572,
+      "step": 489
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.20838627700127e-06,
+      "loss": 0.9543,
+      "step": 490
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.18297331639136e-06,
+      "loss": 0.9609,
+      "step": 491
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.157560355781448e-06,
+      "loss": 1.0179,
+      "step": 492
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.132147395171537e-06,
+      "loss": 0.922,
+      "step": 493
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 8.106734434561627e-06,
+      "loss": 0.9554,
+      "step": 494
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 8.081321473951716e-06,
+      "loss": 0.9575,
+      "step": 495
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 8.055908513341804e-06,
+      "loss": 0.9114,
+      "step": 496
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 8.030495552731894e-06,
+      "loss": 0.9815,
+      "step": 497
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 8.005082592121983e-06,
+      "loss": 0.936,
+      "step": 498
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 7.979669631512071e-06,
+      "loss": 0.9422,
+      "step": 499
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 7.95425667090216e-06,
+      "loss": 0.9759,
+      "step": 500
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 7.92884371029225e-06,
+      "loss": 0.9343,
+      "step": 501
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 7.903430749682338e-06,
+      "loss": 0.9691,
+      "step": 502
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 7.878017789072427e-06,
+      "loss": 0.9585,
+      "step": 503
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 7.852604828462517e-06,
+      "loss": 0.9495,
+      "step": 504
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 7.827191867852606e-06,
+      "loss": 0.9205,
+      "step": 505
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 7.801778907242694e-06,
+      "loss": 0.9316,
+      "step": 506
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 7.776365946632783e-06,
+      "loss": 0.9338,
+      "step": 507
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 7.750952986022873e-06,
+      "loss": 0.963,
+      "step": 508
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 7.72554002541296e-06,
+      "loss": 0.9234,
+      "step": 509
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 7.70012706480305e-06,
+      "loss": 0.9528,
+      "step": 510
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 7.67471410419314e-06,
+      "loss": 0.9431,
+      "step": 511
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 7.649301143583227e-06,
+      "loss": 0.9388,
+      "step": 512
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 7.623888182973317e-06,
+      "loss": 0.9423,
+      "step": 513
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 7.598475222363406e-06,
+      "loss": 0.9848,
+      "step": 514
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 7.573062261753494e-06,
+      "loss": 0.948,
+      "step": 515
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 7.5476493011435835e-06,
+      "loss": 0.9716,
+      "step": 516
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 7.522236340533673e-06,
+      "loss": 0.9888,
+      "step": 517
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 7.4968233799237624e-06,
+      "loss": 0.9656,
+      "step": 518
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 7.47141041931385e-06,
+      "loss": 0.9986,
+      "step": 519
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 7.44599745870394e-06,
+      "loss": 0.9386,
+      "step": 520
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 7.420584498094029e-06,
+      "loss": 0.9615,
+      "step": 521
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 7.395171537484117e-06,
+      "loss": 0.8949,
+      "step": 522
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 7.369758576874206e-06,
+      "loss": 0.9483,
+      "step": 523
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 7.344345616264296e-06,
+      "loss": 0.9417,
+      "step": 524
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 7.318932655654384e-06,
+      "loss": 0.9381,
+      "step": 525
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 7.293519695044473e-06,
+      "loss": 0.9824,
+      "step": 526
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 7.268106734434563e-06,
+      "loss": 0.9453,
+      "step": 527
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 7.242693773824651e-06,
+      "loss": 0.9279,
+      "step": 528
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 7.21728081321474e-06,
+      "loss": 0.9425,
+      "step": 529
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 7.191867852604829e-06,
+      "loss": 0.9534,
+      "step": 530
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 7.166454891994918e-06,
+      "loss": 0.8887,
+      "step": 531
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 7.141041931385007e-06,
+      "loss": 0.9785,
+      "step": 532
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 7.115628970775096e-06,
+      "loss": 0.9495,
+      "step": 533
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 7.090216010165185e-06,
+      "loss": 0.9063,
+      "step": 534
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 7.064803049555273e-06,
+      "loss": 0.9407,
+      "step": 535
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 7.039390088945363e-06,
+      "loss": 0.8544,
+      "step": 536
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 7.0139771283354514e-06,
+      "loss": 0.9321,
+      "step": 537
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 6.988564167725541e-06,
+      "loss": 0.8922,
+      "step": 538
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 6.9631512071156295e-06,
+      "loss": 0.9402,
+      "step": 539
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 6.937738246505718e-06,
+      "loss": 0.9573,
+      "step": 540
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 6.912325285895808e-06,
+      "loss": 0.9679,
+      "step": 541
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 6.886912325285896e-06,
+      "loss": 0.8885,
+      "step": 542
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 6.861499364675985e-06,
+      "loss": 0.9345,
+      "step": 543
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 6.836086404066074e-06,
+      "loss": 0.9294,
+      "step": 544
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 6.810673443456163e-06,
+      "loss": 0.9567,
+      "step": 545
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 6.785260482846252e-06,
+      "loss": 0.946,
+      "step": 546
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 6.759847522236341e-06,
+      "loss": 0.9389,
+      "step": 547
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 6.73443456162643e-06,
+      "loss": 0.928,
+      "step": 548
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 6.709021601016518e-06,
+      "loss": 0.9052,
+      "step": 549
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 6.683608640406608e-06,
+      "loss": 0.9452,
+      "step": 550
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 6.658195679796697e-06,
+      "loss": 0.9438,
+      "step": 551
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 6.632782719186785e-06,
+      "loss": 0.9627,
+      "step": 552
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 6.6073697585768746e-06,
+      "loss": 0.9326,
+      "step": 553
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 6.581956797966964e-06,
+      "loss": 0.9337,
+      "step": 554
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 6.556543837357052e-06,
+      "loss": 0.9822,
+      "step": 555
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 6.531130876747141e-06,
+      "loss": 0.9405,
+      "step": 556
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 6.505717916137231e-06,
+      "loss": 0.9622,
+      "step": 557
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 6.4803049555273186e-06,
+      "loss": 0.8943,
+      "step": 558
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 6.454891994917408e-06,
+      "loss": 0.9429,
+      "step": 559
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 6.4294790343074975e-06,
+      "loss": 0.9177,
+      "step": 560
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 6.404066073697587e-06,
+      "loss": 0.9127,
+      "step": 561
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 6.378653113087675e-06,
+      "loss": 0.9273,
+      "step": 562
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 6.353240152477764e-06,
+      "loss": 0.9668,
+      "step": 563
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 6.327827191867854e-06,
+      "loss": 0.9435,
+      "step": 564
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 6.3024142312579415e-06,
+      "loss": 0.9239,
+      "step": 565
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 6.277001270648031e-06,
+      "loss": 0.9636,
+      "step": 566
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 6.2515883100381204e-06,
+      "loss": 0.951,
+      "step": 567
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 6.226175349428208e-06,
+      "loss": 0.9619,
+      "step": 568
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 6.200762388818298e-06,
+      "loss": 0.952,
+      "step": 569
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 6.175349428208387e-06,
+      "loss": 0.9586,
+      "step": 570
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 6.149936467598475e-06,
+      "loss": 0.9743,
+      "step": 571
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 6.1245235069885644e-06,
+      "loss": 0.9572,
+      "step": 572
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 6.099110546378654e-06,
+      "loss": 0.9653,
+      "step": 573
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 6.073697585768743e-06,
+      "loss": 0.9267,
+      "step": 574
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 6.048284625158831e-06,
+      "loss": 0.8834,
+      "step": 575
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 6.022871664548921e-06,
+      "loss": 0.9369,
+      "step": 576
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.99745870393901e-06,
+      "loss": 0.8974,
+      "step": 577
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.972045743329098e-06,
+      "loss": 0.927,
+      "step": 578
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.946632782719187e-06,
+      "loss": 0.9216,
+      "step": 579
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.921219822109277e-06,
+      "loss": 0.9815,
+      "step": 580
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.895806861499365e-06,
+      "loss": 0.9627,
+      "step": 581
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.870393900889454e-06,
+      "loss": 0.9475,
+      "step": 582
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.844980940279544e-06,
+      "loss": 0.9799,
+      "step": 583
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.819567979669632e-06,
+      "loss": 0.9077,
+      "step": 584
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.794155019059721e-06,
+      "loss": 0.9036,
+      "step": 585
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.76874205844981e-06,
+      "loss": 0.9371,
+      "step": 586
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.743329097839899e-06,
+      "loss": 0.9371,
+      "step": 587
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.7179161372299876e-06,
+      "loss": 0.9181,
+      "step": 588
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.692503176620077e-06,
+      "loss": 0.9686,
+      "step": 589
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.667090216010166e-06,
+      "loss": 0.9537,
+      "step": 590
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.641677255400254e-06,
+      "loss": 0.9862,
+      "step": 591
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.616264294790344e-06,
+      "loss": 0.8973,
+      "step": 592
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.590851334180432e-06,
+      "loss": 0.9619,
+      "step": 593
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.565438373570522e-06,
+      "loss": 0.8699,
+      "step": 594
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.5400254129606105e-06,
+      "loss": 0.9863,
+      "step": 595
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.514612452350699e-06,
+      "loss": 0.9261,
+      "step": 596
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.489199491740789e-06,
+      "loss": 0.8896,
+      "step": 597
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.463786531130877e-06,
+      "loss": 0.9857,
+      "step": 598
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.438373570520966e-06,
+      "loss": 1.0084,
+      "step": 599
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.412960609911055e-06,
+      "loss": 0.9579,
+      "step": 600
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.387547649301144e-06,
+      "loss": 0.93,
+      "step": 601
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.362134688691233e-06,
+      "loss": 0.9735,
+      "step": 602
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.336721728081322e-06,
+      "loss": 0.8658,
+      "step": 603
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.311308767471411e-06,
+      "loss": 0.9362,
+      "step": 604
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.285895806861499e-06,
+      "loss": 0.9172,
+      "step": 605
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.260482846251589e-06,
+      "loss": 0.9787,
+      "step": 606
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.235069885641678e-06,
+      "loss": 0.9249,
+      "step": 607
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.209656925031766e-06,
+      "loss": 0.9552,
+      "step": 608
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.1842439644218555e-06,
+      "loss": 0.8881,
+      "step": 609
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.158831003811945e-06,
+      "loss": 0.8667,
+      "step": 610
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.133418043202033e-06,
+      "loss": 0.9652,
+      "step": 611
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.108005082592122e-06,
+      "loss": 0.933,
+      "step": 612
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.082592121982212e-06,
+      "loss": 0.9431,
+      "step": 613
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.0571791613722995e-06,
+      "loss": 0.9729,
+      "step": 614
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.031766200762389e-06,
+      "loss": 0.9525,
+      "step": 615
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5.0063532401524785e-06,
+      "loss": 0.9727,
+      "step": 616
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.980940279542567e-06,
+      "loss": 0.9068,
+      "step": 617
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.955527318932656e-06,
+      "loss": 0.9558,
+      "step": 618
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.930114358322745e-06,
+      "loss": 0.9144,
+      "step": 619
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.904701397712834e-06,
+      "loss": 0.9567,
+      "step": 620
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.879288437102923e-06,
+      "loss": 0.9562,
+      "step": 621
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.853875476493012e-06,
+      "loss": 0.9111,
+      "step": 622
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.8284625158831006e-06,
+      "loss": 0.9068,
+      "step": 623
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.80304955527319e-06,
+      "loss": 0.9412,
+      "step": 624
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.777636594663279e-06,
+      "loss": 0.9489,
+      "step": 625
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.752223634053368e-06,
+      "loss": 0.9248,
+      "step": 626
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.726810673443457e-06,
+      "loss": 0.9117,
+      "step": 627
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.701397712833545e-06,
+      "loss": 0.9437,
+      "step": 628
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.675984752223635e-06,
+      "loss": 0.9474,
+      "step": 629
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.6505717916137235e-06,
+      "loss": 0.9001,
+      "step": 630
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.625158831003813e-06,
+      "loss": 0.9602,
+      "step": 631
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.599745870393902e-06,
+      "loss": 0.9573,
+      "step": 632
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.57433290978399e-06,
+      "loss": 0.968,
+      "step": 633
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.54891994917408e-06,
+      "loss": 0.9173,
+      "step": 634
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.523506988564168e-06,
+      "loss": 0.9659,
+      "step": 635
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.498094027954257e-06,
+      "loss": 0.8949,
+      "step": 636
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.4726810673443464e-06,
+      "loss": 0.9259,
+      "step": 637
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.447268106734435e-06,
+      "loss": 0.9443,
+      "step": 638
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.421855146124524e-06,
+      "loss": 0.9227,
+      "step": 639
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.396442185514613e-06,
+      "loss": 0.9246,
+      "step": 640
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.371029224904702e-06,
+      "loss": 0.928,
+      "step": 641
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.34561626429479e-06,
+      "loss": 0.9027,
+      "step": 642
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.32020330368488e-06,
+      "loss": 0.9739,
+      "step": 643
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.2947903430749685e-06,
+      "loss": 0.953,
+      "step": 644
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.269377382465057e-06,
+      "loss": 0.91,
+      "step": 645
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.243964421855147e-06,
+      "loss": 0.976,
+      "step": 646
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.218551461245235e-06,
+      "loss": 0.8777,
+      "step": 647
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.193138500635324e-06,
+      "loss": 0.9029,
+      "step": 648
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 4.167725540025413e-06,
+      "loss": 1.0268,
+      "step": 649
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 4.142312579415502e-06,
+      "loss": 0.9461,
+      "step": 650
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 4.116899618805591e-06,
+      "loss": 0.955,
+      "step": 651
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 4.09148665819568e-06,
+      "loss": 0.9116,
+      "step": 652
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.066073697585769e-06,
+      "loss": 0.932,
+      "step": 653
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.040660736975858e-06,
+      "loss": 0.8921,
+      "step": 654
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.015247776365947e-06,
+      "loss": 0.9454,
+      "step": 655
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 3.9898348157560354e-06,
+      "loss": 0.9553,
+      "step": 656
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 3.964421855146125e-06,
+      "loss": 0.9083,
+      "step": 657
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 3.9390088945362135e-06,
+      "loss": 0.9774,
+      "step": 658
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 3.913595933926303e-06,
+      "loss": 0.9757,
+      "step": 659
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 3.888182973316392e-06,
+      "loss": 0.9616,
+      "step": 660
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 3.86277001270648e-06,
+      "loss": 0.9346,
+      "step": 661
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 3.83735705209657e-06,
+      "loss": 0.9411,
+      "step": 662
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 3.8119440914866584e-06,
+      "loss": 0.9285,
+      "step": 663
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 3.786531130876747e-06,
+      "loss": 0.983,
+      "step": 664
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.7611181702668365e-06,
+      "loss": 0.9383,
+      "step": 665
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.735705209656925e-06,
+      "loss": 0.9483,
+      "step": 666
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.7102922490470146e-06,
+      "loss": 0.9264,
+      "step": 667
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.684879288437103e-06,
+      "loss": 0.9179,
+      "step": 668
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.659466327827192e-06,
+      "loss": 0.9538,
+      "step": 669
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.6340533672172813e-06,
+      "loss": 0.9147,
+      "step": 670
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.60864040660737e-06,
+      "loss": 0.9144,
+      "step": 671
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.583227445997459e-06,
+      "loss": 0.9706,
+      "step": 672
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.557814485387548e-06,
+      "loss": 0.9414,
+      "step": 673
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.5324015247776367e-06,
+      "loss": 1.002,
+      "step": 674
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.5069885641677257e-06,
+      "loss": 0.9388,
+      "step": 675
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.4815756035578148e-06,
+      "loss": 0.9318,
+      "step": 676
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.456162642947904e-06,
+      "loss": 0.9591,
+      "step": 677
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.4307496823379925e-06,
+      "loss": 0.916,
+      "step": 678
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.4053367217280815e-06,
+      "loss": 0.8992,
+      "step": 679
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.3799237611181706e-06,
+      "loss": 0.9031,
+      "step": 680
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.354510800508259e-06,
+      "loss": 0.9213,
+      "step": 681
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.3290978398983487e-06,
+      "loss": 0.9685,
+      "step": 682
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.3036848792884373e-06,
+      "loss": 0.9147,
+      "step": 683
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.278271918678526e-06,
+      "loss": 0.9178,
+      "step": 684
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.2528589580686154e-06,
+      "loss": 0.9727,
+      "step": 685
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.227445997458704e-06,
+      "loss": 0.965,
+      "step": 686
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.2020330368487935e-06,
+      "loss": 0.9061,
+      "step": 687
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.176620076238882e-06,
+      "loss": 0.95,
+      "step": 688
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.1512071156289707e-06,
+      "loss": 0.9279,
+      "step": 689
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.1257941550190602e-06,
+      "loss": 0.8828,
+      "step": 690
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.100381194409149e-06,
+      "loss": 0.9453,
+      "step": 691
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.0749682337992375e-06,
+      "loss": 0.9456,
+      "step": 692
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.049555273189327e-06,
+      "loss": 0.9553,
+      "step": 693
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.0241423125794156e-06,
+      "loss": 0.8854,
+      "step": 694
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.998729351969505e-06,
+      "loss": 0.9213,
+      "step": 695
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.9733163913595937e-06,
+      "loss": 0.9328,
+      "step": 696
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.9479034307496823e-06,
+      "loss": 0.9652,
+      "step": 697
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.922490470139772e-06,
+      "loss": 0.9753,
+      "step": 698
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.8970775095298604e-06,
+      "loss": 0.9745,
+      "step": 699
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.8716645489199495e-06,
+      "loss": 0.9716,
+      "step": 700
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.8462515883100385e-06,
+      "loss": 0.9484,
+      "step": 701
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.820838627700127e-06,
+      "loss": 0.9651,
+      "step": 702
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.795425667090216e-06,
+      "loss": 0.909,
+      "step": 703
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.7700127064803053e-06,
+      "loss": 0.9381,
+      "step": 704
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.7445997458703943e-06,
+      "loss": 0.9284,
+      "step": 705
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.719186785260483e-06,
+      "loss": 0.9336,
+      "step": 706
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.693773824650572e-06,
+      "loss": 0.9793,
+      "step": 707
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.668360864040661e-06,
+      "loss": 0.9369,
+      "step": 708
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.6429479034307497e-06,
+      "loss": 0.9662,
+      "step": 709
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.617534942820839e-06,
+      "loss": 0.9326,
+      "step": 710
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.5921219822109278e-06,
+      "loss": 0.9697,
+      "step": 711
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.5667090216010164e-06,
+      "loss": 0.9255,
+      "step": 712
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.541296060991106e-06,
+      "loss": 0.9559,
+      "step": 713
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.5158831003811945e-06,
+      "loss": 0.937,
+      "step": 714
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.4904701397712835e-06,
+      "loss": 1.0028,
+      "step": 715
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.4650571791613726e-06,
+      "loss": 0.9311,
+      "step": 716
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.4396442185514616e-06,
+      "loss": 0.9385,
+      "step": 717
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.4142312579415503e-06,
+      "loss": 1.0194,
+      "step": 718
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.3888182973316393e-06,
+      "loss": 0.9342,
+      "step": 719
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.3634053367217284e-06,
+      "loss": 0.9348,
+      "step": 720
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.3379923761118174e-06,
+      "loss": 0.9554,
+      "step": 721
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.3125794155019065e-06,
+      "loss": 0.9584,
+      "step": 722
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.287166454891995e-06,
+      "loss": 0.9461,
+      "step": 723
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.261753494282084e-06,
+      "loss": 0.9833,
+      "step": 724
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.2363405336721732e-06,
+      "loss": 0.9317,
+      "step": 725
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.210927573062262e-06,
+      "loss": 0.9249,
+      "step": 726
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.185514612452351e-06,
+      "loss": 0.9459,
+      "step": 727
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.16010165184244e-06,
+      "loss": 0.9154,
+      "step": 728
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.1346886912325286e-06,
+      "loss": 0.8959,
+      "step": 729
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.1092757306226176e-06,
+      "loss": 0.8956,
+      "step": 730
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.0838627700127067e-06,
+      "loss": 0.965,
+      "step": 731
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.0584498094027953e-06,
+      "loss": 0.9312,
+      "step": 732
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.0330368487928844e-06,
+      "loss": 0.9422,
+      "step": 733
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.0076238881829734e-06,
+      "loss": 0.923,
+      "step": 734
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.9822109275730625e-06,
+      "loss": 0.9581,
+      "step": 735
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.9567979669631515e-06,
+      "loss": 0.9292,
+      "step": 736
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.93138500635324e-06,
+      "loss": 0.931,
+      "step": 737
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.9059720457433292e-06,
+      "loss": 0.9273,
+      "step": 738
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.8805590851334182e-06,
+      "loss": 0.9724,
+      "step": 739
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.8551461245235073e-06,
+      "loss": 0.9535,
+      "step": 740
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.829733163913596e-06,
+      "loss": 0.9051,
+      "step": 741
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.804320203303685e-06,
+      "loss": 0.9325,
+      "step": 742
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.778907242693774e-06,
+      "loss": 0.8874,
+      "step": 743
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.7534942820838629e-06,
+      "loss": 0.9066,
+      "step": 744
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.728081321473952e-06,
+      "loss": 0.9178,
+      "step": 745
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.7026683608640408e-06,
+      "loss": 0.9581,
+      "step": 746
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.6772554002541296e-06,
+      "loss": 0.9246,
+      "step": 747
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.6518424396442186e-06,
+      "loss": 0.936,
+      "step": 748
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.6264294790343077e-06,
+      "loss": 0.9105,
+      "step": 749
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.6010165184243967e-06,
+      "loss": 0.9148,
+      "step": 750
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.5756035578144854e-06,
+      "loss": 0.8642,
+      "step": 751
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.5501905972045744e-06,
+      "loss": 0.9786,
+      "step": 752
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.5247776365946635e-06,
+      "loss": 0.9654,
+      "step": 753
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.4993646759847525e-06,
+      "loss": 0.9961,
+      "step": 754
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.4739517153748412e-06,
+      "loss": 0.9268,
+      "step": 755
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.4485387547649302e-06,
+      "loss": 0.9869,
+      "step": 756
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.4231257941550193e-06,
+      "loss": 0.9039,
+      "step": 757
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.397712833545108e-06,
+      "loss": 0.9573,
+      "step": 758
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.3722998729351972e-06,
+      "loss": 0.9207,
+      "step": 759
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.346886912325286e-06,
+      "loss": 0.9206,
+      "step": 760
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.3214739517153748e-06,
+      "loss": 0.8987,
+      "step": 761
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.2960609911054639e-06,
+      "loss": 0.9844,
+      "step": 762
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.270648030495553e-06,
+      "loss": 0.9643,
+      "step": 763
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.2452350698856418e-06,
+      "loss": 0.9204,
+      "step": 764
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.2198221092757308e-06,
+      "loss": 0.9537,
+      "step": 765
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.1944091486658197e-06,
+      "loss": 0.9551,
+      "step": 766
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.1689961880559087e-06,
+      "loss": 0.9593,
+      "step": 767
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.1435832274459976e-06,
+      "loss": 0.936,
+      "step": 768
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.1181702668360866e-06,
+      "loss": 0.8839,
+      "step": 769
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.0927573062261754e-06,
+      "loss": 0.9327,
+      "step": 770
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.0673443456162643e-06,
+      "loss": 0.9724,
+      "step": 771
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.0419313850063533e-06,
+      "loss": 0.8627,
+      "step": 772
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.0165184243964422e-06,
+      "loss": 0.9291,
+      "step": 773
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 9.911054637865312e-07,
+      "loss": 0.9767,
+      "step": 774
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 9.6569250317662e-07,
+      "loss": 0.9107,
+      "step": 775
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 9.402795425667091e-07,
+      "loss": 0.9238,
+      "step": 776
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 9.14866581956798e-07,
+      "loss": 0.8996,
+      "step": 777
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.89453621346887e-07,
+      "loss": 0.885,
+      "step": 778
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.64040660736976e-07,
+      "loss": 0.9291,
+      "step": 779
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.386277001270648e-07,
+      "loss": 0.9325,
+      "step": 780
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.132147395171538e-07,
+      "loss": 0.9458,
+      "step": 781
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 7.878017789072427e-07,
+      "loss": 0.8818,
+      "step": 782
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 7.623888182973317e-07,
+      "loss": 0.9199,
+      "step": 783
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 7.369758576874206e-07,
+      "loss": 0.9355,
+      "step": 784
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 7.115628970775096e-07,
+      "loss": 0.9028,
+      "step": 785
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 6.861499364675986e-07,
+      "loss": 0.9365,
+      "step": 786
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 6.607369758576874e-07,
+      "loss": 0.8661,
+      "step": 787
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 6.353240152477765e-07,
+      "loss": 0.95,
+      "step": 788
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 6.099110546378654e-07,
+      "loss": 0.8995,
+      "step": 789
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.844980940279544e-07,
+      "loss": 0.9334,
+      "step": 790
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.590851334180433e-07,
+      "loss": 0.9023,
+      "step": 791
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.336721728081321e-07,
+      "loss": 0.9009,
+      "step": 792
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.082592121982211e-07,
+      "loss": 0.9751,
+      "step": 793
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 4.8284625158831e-07,
+      "loss": 0.9643,
+      "step": 794
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 4.57433290978399e-07,
+      "loss": 0.9558,
+      "step": 795
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 4.32020330368488e-07,
+      "loss": 0.9375,
+      "step": 796
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 4.066073697585769e-07,
+      "loss": 0.9363,
+      "step": 797
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.8119440914866587e-07,
+      "loss": 0.9716,
+      "step": 798
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.557814485387548e-07,
+      "loss": 0.9595,
+      "step": 799
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.303684879288437e-07,
+      "loss": 0.92,
+      "step": 800
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.049555273189327e-07,
+      "loss": 0.942,
+      "step": 801
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 2.7954256670902165e-07,
+      "loss": 0.9242,
+      "step": 802
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2.5412960609911054e-07,
+      "loss": 0.9447,
+      "step": 803
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2.287166454891995e-07,
+      "loss": 0.9344,
+      "step": 804
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2.0330368487928846e-07,
+      "loss": 0.9115,
+      "step": 805
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.778907242693774e-07,
+      "loss": 0.9294,
+      "step": 806
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.5247776365946635e-07,
+      "loss": 0.9203,
+      "step": 807
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.2706480304955527e-07,
+      "loss": 0.9871,
+      "step": 808
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.0165184243964423e-07,
+      "loss": 0.9066,
+      "step": 809
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.623888182973318e-08,
+      "loss": 0.9794,
+      "step": 810
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.0825921219822115e-08,
+      "loss": 0.9405,
+      "step": 811
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.5412960609911058e-08,
+      "loss": 0.9164,
+      "step": 812
+    },
+    {
+      "epoch": 2.0,
+      "step": 812,
+      "total_flos": 2.1351127441756652e+18,
+      "train_loss": 1.0052507809670688,
+      "train_runtime": 9475.3051,
+      "train_samples_per_second": 10.976,
+      "train_steps_per_second": 0.086
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 812,
+  "num_train_epochs": 2,
+  "save_steps": 4000,
+  "total_flos": 2.1351127441756652e+18,
+  "trial_name": null,
+  "trial_params": null
+}