diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,10169 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "global_step": 1688,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.352941176470589e-07,
+      "loss": 1.5544,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4705882352941177e-06,
+      "loss": 1.3787,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2058823529411767e-06,
+      "loss": 0.9109,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9411764705882355e-06,
+      "loss": 0.7882,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6764705882352942e-06,
+      "loss": 0.6814,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.411764705882353e-06,
+      "loss": 0.7527,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.147058823529412e-06,
+      "loss": 0.7286,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.882352941176471e-06,
+      "loss": 0.5626,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.61764705882353e-06,
+      "loss": 0.3991,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.3529411764705884e-06,
+      "loss": 0.469,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.088235294117648e-06,
+      "loss": 0.4538,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.823529411764707e-06,
+      "loss": 0.3775,
+      "step": 12
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.558823529411764e-06,
+      "loss": 0.4343,
+      "step": 13
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0294117647058824e-05,
+      "loss": 0.3084,
+      "step": 14
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1029411764705883e-05,
+      "loss": 0.3595,
+      "step": 15
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1764705882352942e-05,
+      "loss": 0.4142,
+      "step": 16
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.25e-05,
+      "loss": 0.4158,
+      "step": 17
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.323529411764706e-05,
+      "loss": 0.3162,
+      "step": 18
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3970588235294118e-05,
+      "loss": 0.4296,
+      "step": 19
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4705882352941177e-05,
+      "loss": 0.2731,
+      "step": 20
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5441176470588237e-05,
+      "loss": 0.2992,
+      "step": 21
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.6176470588235296e-05,
+      "loss": 0.3625,
+      "step": 22
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.6911764705882355e-05,
+      "loss": 0.2786,
+      "step": 23
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.7647058823529414e-05,
+      "loss": 0.2467,
+      "step": 24
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.8382352941176472e-05,
+      "loss": 0.2544,
+      "step": 25
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.9117647058823528e-05,
+      "loss": 0.2681,
+      "step": 26
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.9852941176470586e-05,
+      "loss": 0.4685,
+      "step": 27
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.058823529411765e-05,
+      "loss": 0.3739,
+      "step": 28
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.1323529411764707e-05,
+      "loss": 0.3166,
+      "step": 29
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.2058823529411766e-05,
+      "loss": 0.3276,
+      "step": 30
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.2794117647058825e-05,
+      "loss": 0.2656,
+      "step": 31
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.3529411764705884e-05,
+      "loss": 0.383,
+      "step": 32
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.4264705882352942e-05,
+      "loss": 0.2887,
+      "step": 33
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.5e-05,
+      "loss": 0.363,
+      "step": 34
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.5735294117647057e-05,
+      "loss": 0.3501,
+      "step": 35
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.647058823529412e-05,
+      "loss": 0.2481,
+      "step": 36
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.7205882352941174e-05,
+      "loss": 0.3818,
+      "step": 37
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.7941176470588236e-05,
+      "loss": 0.3355,
+      "step": 38
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.8676470588235295e-05,
+      "loss": 0.302,
+      "step": 39
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.9411764705882354e-05,
+      "loss": 0.3458,
+      "step": 40
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 3.0147058823529413e-05,
+      "loss": 0.2199,
+      "step": 41
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 3.0882352941176475e-05,
+      "loss": 0.306,
+      "step": 42
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 3.161764705882353e-05,
+      "loss": 0.2886,
+      "step": 43
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 3.235294117647059e-05,
+      "loss": 0.3439,
+      "step": 44
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 3.308823529411765e-05,
+      "loss": 0.3267,
+      "step": 45
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 3.382352941176471e-05,
+      "loss": 0.2609,
+      "step": 46
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.455882352941177e-05,
+      "loss": 0.3037,
+      "step": 47
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.529411764705883e-05,
+      "loss": 0.3316,
+      "step": 48
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.6029411764705886e-05,
+      "loss": 0.2266,
+      "step": 49
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.6764705882352945e-05,
+      "loss": 0.1846,
+      "step": 50
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.4235,
+      "step": 51
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.8235294117647055e-05,
+      "loss": 0.1886,
+      "step": 52
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.897058823529412e-05,
+      "loss": 0.3208,
+      "step": 53
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.970588235294117e-05,
+      "loss": 0.3269,
+      "step": 54
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.044117647058824e-05,
+      "loss": 0.3745,
+      "step": 55
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.11764705882353e-05,
+      "loss": 0.3509,
+      "step": 56
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.1911764705882356e-05,
+      "loss": 0.3891,
+      "step": 57
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.2647058823529415e-05,
+      "loss": 0.1774,
+      "step": 58
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.3382352941176474e-05,
+      "loss": 0.3896,
+      "step": 59
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.411764705882353e-05,
+      "loss": 0.2753,
+      "step": 60
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.485294117647059e-05,
+      "loss": 0.2608,
+      "step": 61
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.558823529411765e-05,
+      "loss": 0.3358,
+      "step": 62
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.632352941176471e-05,
+      "loss": 0.3424,
+      "step": 63
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.705882352941177e-05,
+      "loss": 0.4011,
+      "step": 64
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.7794117647058826e-05,
+      "loss": 0.3513,
+      "step": 65
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8529411764705885e-05,
+      "loss": 0.5082,
+      "step": 66
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.9264705882352944e-05,
+      "loss": 0.4238,
+      "step": 67
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 0.3546,
+      "step": 68
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.9999952991153656e-05,
+      "loss": 0.435,
+      "step": 69
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.99998119647914e-05,
+      "loss": 0.2671,
+      "step": 70
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.999957692144361e-05,
+      "loss": 0.2809,
+      "step": 71
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.9999247861994194e-05,
+      "loss": 0.2987,
+      "step": 72
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.9998824787680656e-05,
+      "loss": 0.4445,
+      "step": 73
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.999830770009406e-05,
+      "loss": 0.2814,
+      "step": 74
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.999769660117901e-05,
+      "loss": 0.5113,
+      "step": 75
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.9996991493233693e-05,
+      "loss": 0.3154,
+      "step": 76
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.9996192378909786e-05,
+      "loss": 0.2822,
+      "step": 77
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.9995299261212536e-05,
+      "loss": 0.4062,
+      "step": 78
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.9994312143500706e-05,
+      "loss": 0.4488,
+      "step": 79
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.9993231029486544e-05,
+      "loss": 0.4507,
+      "step": 80
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9992055923235815e-05,
+      "loss": 0.1764,
+      "step": 81
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.999078682916774e-05,
+      "loss": 0.3376,
+      "step": 82
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.998942375205502e-05,
+      "loss": 0.3232,
+      "step": 83
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.998796669702378e-05,
+      "loss": 0.3122,
+      "step": 84
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9986415669553586e-05,
+      "loss": 0.3315,
+      "step": 85
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.99847706754774e-05,
+      "loss": 0.3443,
+      "step": 86
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.998303172098155e-05,
+      "loss": 0.4327,
+      "step": 87
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.998119881260576e-05,
+      "loss": 0.3069,
+      "step": 88
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.997927195724303e-05,
+      "loss": 0.4294,
+      "step": 89
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.997725116213973e-05,
+      "loss": 0.2244,
+      "step": 90
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.997513643489547e-05,
+      "loss": 0.3895,
+      "step": 91
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.997292778346312e-05,
+      "loss": 0.316,
+      "step": 92
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.997062521614876e-05,
+      "loss": 0.4678,
+      "step": 93
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.996822874161169e-05,
+      "loss": 0.4502,
+      "step": 94
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.996573836886435e-05,
+      "loss": 0.2164,
+      "step": 95
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.9963154107272295e-05,
+      "loss": 0.5023,
+      "step": 96
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.996047596655418e-05,
+      "loss": 0.2756,
+      "step": 97
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.995770395678171e-05,
+      "loss": 0.4203,
+      "step": 98
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.99548380883796e-05,
+      "loss": 0.2576,
+      "step": 99
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9951878372125547e-05,
+      "loss": 0.3604,
+      "step": 100
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9948824819150185e-05,
+      "loss": 0.3917,
+      "step": 101
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.994567744093703e-05,
+      "loss": 0.4184,
+      "step": 102
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9942436249322444e-05,
+      "loss": 0.2433,
+      "step": 103
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.993910125649561e-05,
+      "loss": 0.2559,
+      "step": 104
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.993567247499845e-05,
+      "loss": 0.35,
+      "step": 105
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.993214991772563e-05,
+      "loss": 0.3242,
+      "step": 106
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.992853359792444e-05,
+      "loss": 0.4184,
+      "step": 107
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.99248235291948e-05,
+      "loss": 0.2682,
+      "step": 108
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.992101972548921e-05,
+      "loss": 0.2615,
+      "step": 109
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.9917122201112656e-05,
+      "loss": 0.3701,
+      "step": 110
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.9913130970722584e-05,
+      "loss": 0.4397,
+      "step": 111
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.9909046049328846e-05,
+      "loss": 0.2802,
+      "step": 112
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.990486745229364e-05,
+      "loss": 0.3457,
+      "step": 113
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.990059519533144e-05,
+      "loss": 0.4439,
+      "step": 114
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9896229294508976e-05,
+      "loss": 0.3117,
+      "step": 115
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.989176976624511e-05,
+      "loss": 0.2405,
+      "step": 116
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.988721662731083e-05,
+      "loss": 0.3129,
+      "step": 117
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9882569894829144e-05,
+      "loss": 0.3435,
+      "step": 118
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.987782958627508e-05,
+      "loss": 0.3055,
+      "step": 119
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.987299571947553e-05,
+      "loss": 0.378,
+      "step": 120
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.986806831260927e-05,
+      "loss": 0.3637,
+      "step": 121
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9863047384206835e-05,
+      "loss": 0.2853,
+      "step": 122
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.9857932953150465e-05,
+      "loss": 0.4996,
+      "step": 123
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.985272503867403e-05,
+      "loss": 0.2734,
+      "step": 124
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.9847423660363e-05,
+      "loss": 0.3786,
+      "step": 125
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.9842028838154285e-05,
+      "loss": 0.2956,
+      "step": 126
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.983654059233626e-05,
+      "loss": 0.357,
+      "step": 127
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.983095894354858e-05,
+      "loss": 0.3909,
+      "step": 128
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.982528391278222e-05,
+      "loss": 0.2797,
+      "step": 129
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.981951552137929e-05,
+      "loss": 0.4412,
+      "step": 130
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9813653791033057e-05,
+      "loss": 0.4442,
+      "step": 131
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9807698743787744e-05,
+      "loss": 0.3597,
+      "step": 132
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9801650402038555e-05,
+      "loss": 0.3196,
+      "step": 133
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.979550878853154e-05,
+      "loss": 0.347,
+      "step": 134
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.978927392636351e-05,
+      "loss": 0.7105,
+      "step": 135
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.978294583898196e-05,
+      "loss": 0.3297,
+      "step": 136
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9776524550184965e-05,
+      "loss": 0.3836,
+      "step": 137
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.977001008412113e-05,
+      "loss": 0.4608,
+      "step": 138
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.976340246528945e-05,
+      "loss": 0.4569,
+      "step": 139
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.975670171853926e-05,
+      "loss": 0.394,
+      "step": 140
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.97499078690701e-05,
+      "loss": 0.3661,
+      "step": 141
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.974302094243164e-05,
+      "loss": 0.2955,
+      "step": 142
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.973604096452361e-05,
+      "loss": 0.3605,
+      "step": 143
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.972896796159568e-05,
+      "loss": 0.3348,
+      "step": 144
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.972180196024733e-05,
+      "loss": 0.3177,
+      "step": 145
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.971454298742779e-05,
+      "loss": 0.2532,
+      "step": 146
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.970719107043595e-05,
+      "loss": 0.3529,
+      "step": 147
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.969974623692023e-05,
+      "loss": 0.2531,
+      "step": 148
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.9692208514878444e-05,
+      "loss": 0.2248,
+      "step": 149
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.9684577932657786e-05,
+      "loss": 0.3158,
+      "step": 150
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.967685451895463e-05,
+      "loss": 0.3665,
+      "step": 151
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.966903830281449e-05,
+      "loss": 0.3679,
+      "step": 152
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.966112931363185e-05,
+      "loss": 0.1797,
+      "step": 153
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.9653127581150126e-05,
+      "loss": 0.3741,
+      "step": 154
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.9645033135461494e-05,
+      "loss": 0.3707,
+      "step": 155
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.963684600700679e-05,
+      "loss": 0.3889,
+      "step": 156
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.962856622657541e-05,
+      "loss": 0.3145,
+      "step": 157
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.962019382530521e-05,
+      "loss": 0.4141,
+      "step": 158
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.961172883468231e-05,
+      "loss": 0.3809,
+      "step": 159
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.960317128654108e-05,
+      "loss": 0.4337,
+      "step": 160
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.9594521213063974e-05,
+      "loss": 0.4036,
+      "step": 161
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.9585778646781364e-05,
+      "loss": 0.3361,
+      "step": 162
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.9576943620571507e-05,
+      "loss": 0.3469,
+      "step": 163
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.9568016167660334e-05,
+      "loss": 0.2622,
+      "step": 164
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.9558996321621405e-05,
+      "loss": 0.3236,
+      "step": 165
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.9549884116375714e-05,
+      "loss": 0.3198,
+      "step": 166
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.9540679586191605e-05,
+      "loss": 0.2858,
+      "step": 167
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.953138276568462e-05,
+      "loss": 0.3696,
+      "step": 168
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.952199368981739e-05,
+      "loss": 0.2745,
+      "step": 169
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.951251239389948e-05,
+      "loss": 0.4968,
+      "step": 170
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.950293891358728e-05,
+      "loss": 0.4382,
+      "step": 171
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.9493273284883854e-05,
+      "loss": 0.3721,
+      "step": 172
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.948351554413879e-05,
+      "loss": 0.5201,
+      "step": 173
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.947366572804811e-05,
+      "loss": 0.374,
+      "step": 174
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.946372387365409e-05,
+      "loss": 0.4549,
+      "step": 175
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.9453690018345144e-05,
+      "loss": 0.3247,
+      "step": 176
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.9443564199855666e-05,
+      "loss": 0.5224,
+      "step": 177
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.94333464562659e-05,
+      "loss": 0.3707,
+      "step": 178
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.942303682600178e-05,
+      "loss": 0.5609,
+      "step": 179
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.941263534783482e-05,
+      "loss": 0.295,
+      "step": 180
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.9402142060881964e-05,
+      "loss": 0.3231,
+      "step": 181
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.939155700460536e-05,
+      "loss": 0.5371,
+      "step": 182
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.938088021881233e-05,
+      "loss": 0.3053,
+      "step": 183
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.9370111743655145e-05,
+      "loss": 0.3319,
+      "step": 184
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.9359251619630886e-05,
+      "loss": 0.4106,
+      "step": 185
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.934829988758131e-05,
+      "loss": 0.3124,
+      "step": 186
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.933725658869267e-05,
+      "loss": 0.2441,
+      "step": 187
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.9326121764495596e-05,
+      "loss": 0.356,
+      "step": 188
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.9314895456864904e-05,
+      "loss": 0.3997,
+      "step": 189
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.930357770801947e-05,
+      "loss": 0.3406,
+      "step": 190
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.9292168560522014e-05,
+      "loss": 0.3038,
+      "step": 191
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.9280668057279014e-05,
+      "loss": 0.2946,
+      "step": 192
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.9269076241540505e-05,
+      "loss": 0.2809,
+      "step": 193
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.925739315689991e-05,
+      "loss": 0.4038,
+      "step": 194
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.924561884729391e-05,
+      "loss": 0.3231,
+      "step": 195
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.923375335700223e-05,
+      "loss": 0.4288,
+      "step": 196
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.9221796730647516e-05,
+      "loss": 0.3591,
+      "step": 197
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.920974901319515e-05,
+      "loss": 0.4282,
+      "step": 198
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.9197610249953065e-05,
+      "loss": 0.3657,
+      "step": 199
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.9185380486571595e-05,
+      "loss": 0.3192,
+      "step": 200
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.917305976904333e-05,
+      "loss": 0.4179,
+      "step": 201
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.916064814370287e-05,
+      "loss": 0.3324,
+      "step": 202
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.914814565722671e-05,
+      "loss": 0.4862,
+      "step": 203
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.913555235663305e-05,
+      "loss": 0.5322,
+      "step": 204
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.912286828928162e-05,
+      "loss": 0.3895,
+      "step": 205
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.9110093502873476e-05,
+      "loss": 0.2983,
+      "step": 206
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.9097228045450864e-05,
+      "loss": 0.5021,
+      "step": 207
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.9084271965397014e-05,
+      "loss": 0.2682,
+      "step": 208
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.907122531143594e-05,
+      "loss": 0.333,
+      "step": 209
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.9058088132632306e-05,
+      "loss": 0.3926,
+      "step": 210
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.904486047839121e-05,
+      "loss": 0.2588,
+      "step": 211
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.9031542398457974e-05,
+      "loss": 0.3177,
+      "step": 212
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.901813394291801e-05,
+      "loss": 0.2823,
+      "step": 213
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.900463516219661e-05,
+      "loss": 0.2356,
+      "step": 214
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.8991046107058735e-05,
+      "loss": 0.5614,
+      "step": 215
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.897736682860885e-05,
+      "loss": 0.2671,
+      "step": 216
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.896359737829071e-05,
+      "loss": 0.2985,
+      "step": 217
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.894973780788722e-05,
+      "loss": 0.3246,
+      "step": 218
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.893578816952016e-05,
+      "loss": 0.3455,
+      "step": 219
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.892174851565004e-05,
+      "loss": 0.318,
+      "step": 220
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.890761889907589e-05,
+      "loss": 0.3999,
+      "step": 221
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.889339937293508e-05,
+      "loss": 0.4288,
+      "step": 222
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.887908999070308e-05,
+      "loss": 0.2957,
+      "step": 223
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.88646908061933e-05,
+      "loss": 0.2453,
+      "step": 224
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.885020187355687e-05,
+      "loss": 0.4468,
+      "step": 225
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.883562324728241e-05,
+      "loss": 0.3517,
+      "step": 226
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.8820954982195905e-05,
+      "loss": 0.4454,
+      "step": 227
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.880619713346039e-05,
+      "loss": 0.2959,
+      "step": 228
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.8791349756575814e-05,
+      "loss": 0.3715,
+      "step": 229
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.877641290737884e-05,
+      "loss": 0.2904,
+      "step": 230
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.8761386642042584e-05,
+      "loss": 0.4513,
+      "step": 231
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.874627101707644e-05,
+      "loss": 0.3151,
+      "step": 232
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.873106608932585e-05,
+      "loss": 0.3398,
+      "step": 233
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.871577191597211e-05,
+      "loss": 0.3502,
+      "step": 234
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.870038855453213e-05,
+      "loss": 0.2512,
+      "step": 235
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.868491606285823e-05,
+      "loss": 0.3682,
+      "step": 236
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.8669354499137955e-05,
+      "loss": 0.415,
+      "step": 237
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.8653703921893766e-05,
+      "loss": 0.4233,
+      "step": 238
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.8637964389982926e-05,
+      "loss": 0.253,
+      "step": 239
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.86221359625972e-05,
+      "loss": 0.302,
+      "step": 240
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.860621869926271e-05,
+      "loss": 0.3677,
+      "step": 241
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.859021265983959e-05,
+      "loss": 0.3096,
+      "step": 242
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.85741179045219e-05,
+      "loss": 0.388,
+      "step": 243
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.855793449383731e-05,
+      "loss": 0.3175,
+      "step": 244
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.854166248864689e-05,
+      "loss": 0.3613,
+      "step": 245
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.8525301950144894e-05,
+      "loss": 0.434,
+      "step": 246
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.850885293985853e-05,
+      "loss": 0.3275,
+      "step": 247
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.849231551964771e-05,
+      "loss": 0.4166,
+      "step": 248
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.8475689751704855e-05,
+      "loss": 0.2651,
+      "step": 249
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.84589756985546e-05,
+      "loss": 0.3465,
+      "step": 250
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.844217342305363e-05,
+      "loss": 0.3547,
+      "step": 251
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.8425282988390376e-05,
+      "loss": 0.3044,
+      "step": 252
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.840830445808483e-05,
+      "loss": 0.3513,
+      "step": 253
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.839123789598829e-05,
+      "loss": 0.2609,
+      "step": 254
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.8374083366283096e-05,
+      "loss": 0.4136,
+      "step": 255
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.835684093348244e-05,
+      "loss": 0.3119,
+      "step": 256
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.8339510662430046e-05,
+      "loss": 0.2062,
+      "step": 257
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.832209261830002e-05,
+      "loss": 0.3218,
+      "step": 258
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.8304586866596515e-05,
+      "loss": 0.3071,
+      "step": 259
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.828699347315356e-05,
+      "loss": 0.3227,
+      "step": 260
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.826931250413477e-05,
+      "loss": 0.28,
+      "step": 261
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.825154402603308e-05,
+      "loss": 0.2957,
+      "step": 262
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.823368810567056e-05,
+      "loss": 0.3776,
+      "step": 263
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.821574481019811e-05,
+      "loss": 0.3162,
+      "step": 264
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.8197714207095205e-05,
+      "loss": 0.3972,
+      "step": 265
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.817959636416969e-05,
+      "loss": 0.2932,
+      "step": 266
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.816139134955746e-05,
+      "loss": 0.399,
+      "step": 267
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.814309923172227e-05,
+      "loss": 0.3041,
+      "step": 268
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.81247200794554e-05,
+      "loss": 0.3107,
+      "step": 269
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.8106253961875506e-05,
+      "loss": 0.2649,
+      "step": 270
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.8087700948428236e-05,
+      "loss": 0.3994,
+      "step": 271
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.806906110888606e-05,
+      "loss": 0.3061,
+      "step": 272
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.805033451334796e-05,
+      "loss": 0.3092,
+      "step": 273
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.8031521232239195e-05,
+      "loss": 0.272,
+      "step": 274
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.8012621336311016e-05,
+      "loss": 0.3009,
+      "step": 275
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.7993634896640394e-05,
+      "loss": 0.4287,
+      "step": 276
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.797456198462979e-05,
+      "loss": 0.2823,
+      "step": 277
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.7955402672006854e-05,
+      "loss": 0.2985,
+      "step": 278
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.793615703082415e-05,
+      "loss": 0.3632,
+      "step": 279
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.791682513345892e-05,
+      "loss": 0.3735,
+      "step": 280
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.789740705261278e-05,
+      "loss": 0.382,
+      "step": 281
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.7877902861311446e-05,
+      "loss": 0.3346,
+      "step": 282
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.785831263290449e-05,
+      "loss": 0.432,
+      "step": 283
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.783863644106502e-05,
+      "loss": 0.2894,
+      "step": 284
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.781887435978947e-05,
+      "loss": 0.2604,
+      "step": 285
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.779902646339722e-05,
+      "loss": 0.3985,
+      "step": 286
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.777909282653042e-05,
+      "loss": 0.2818,
+      "step": 287
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.775907352415367e-05,
+      "loss": 0.1794,
+      "step": 288
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.7738968631553685e-05,
+      "loss": 0.4045,
+      "step": 289
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.771877822433911e-05,
+      "loss": 0.3694,
+      "step": 290
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.7698502378440155e-05,
+      "loss": 0.3761,
+      "step": 291
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.7678141170108345e-05,
+      "loss": 0.2488,
+      "step": 292
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.765769467591625e-05,
+      "loss": 0.2353,
+      "step": 293
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.763716297275715e-05,
+      "loss": 0.2735,
+      "step": 294
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.761654613784477e-05,
+      "loss": 0.3059,
+      "step": 295
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.759584424871302e-05,
+      "loss": 0.309,
+      "step": 296
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.757505738321563e-05,
+      "loss": 0.35,
+      "step": 297
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.755418561952595e-05,
+      "loss": 0.2475,
+      "step": 298
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.7533229036136553e-05,
+      "loss": 0.3117,
+      "step": 299
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.751218771185906e-05,
+      "loss": 0.2364,
+      "step": 300
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.749106172582371e-05,
+      "loss": 0.3342,
+      "step": 301
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.7469851157479177e-05,
+      "loss": 0.3991,
+      "step": 302
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.74485560865922e-05,
+      "loss": 0.3362,
+      "step": 303
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.7427176593247334e-05,
+      "loss": 0.3082,
+      "step": 304
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.740571275784659e-05,
+      "loss": 0.3617,
+      "step": 305
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.7384164661109176e-05,
+      "loss": 0.2207,
+      "step": 306
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.736253238407119e-05,
+      "loss": 0.3742,
+      "step": 307
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.734081600808531e-05,
+      "loss": 0.4731,
+      "step": 308
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.7319015614820464e-05,
+      "loss": 0.2491,
+      "step": 309
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.729713128626158e-05,
+      "loss": 0.3909,
+      "step": 310
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.72751631047092e-05,
+      "loss": 0.544,
+      "step": 311
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.725311115277924e-05,
+      "loss": 0.2997,
+      "step": 312
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.7230975513402655e-05,
+      "loss": 0.4305,
+      "step": 313
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.7208756269825104e-05,
+      "loss": 0.2567,
+      "step": 314
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.7186453505606676e-05,
+      "loss": 0.3025,
+      "step": 315
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.7164067304621536e-05,
+      "loss": 0.2402,
+      "step": 316
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.714159775105765e-05,
+      "loss": 0.273,
+      "step": 317
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.7119044929416443e-05,
+      "loss": 0.3799,
+      "step": 318
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.7096408924512495e-05,
+      "loss": 0.2189,
+      "step": 319
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.707368982147318e-05,
+      "loss": 0.2667,
+      "step": 320
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.705088770573841e-05,
+      "loss": 0.3559,
+      "step": 321
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.70280026630603e-05,
+      "loss": 0.5352,
+      "step": 322
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.700503477950278e-05,
+      "loss": 0.3618,
+      "step": 323
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.698198414144136e-05,
+      "loss": 0.3828,
+      "step": 324
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.695885083556275e-05,
+      "loss": 0.3886,
+      "step": 325
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.693563494886455e-05,
+      "loss": 0.3056,
+      "step": 326
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.6912336568654925e-05,
+      "loss": 0.3461,
+      "step": 327
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.6888955782552274e-05,
+      "loss": 0.3666,
+      "step": 328
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.6865492678484895e-05,
+      "loss": 0.2631,
+      "step": 329
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.684194734469067e-05,
+      "loss": 0.3742,
+      "step": 330
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.681831986971672e-05,
+      "loss": 0.3199,
+      "step": 331
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.679461034241906e-05,
+      "loss": 0.2793,
+      "step": 332
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.6770818851962305e-05,
+      "loss": 0.5359,
+      "step": 333
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.6746945487819286e-05,
+      "loss": 0.4189,
+      "step": 334
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.672299033977076e-05,
+      "loss": 0.3239,
+      "step": 335
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.669895349790502e-05,
+      "loss": 0.3618,
+      "step": 336
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.667483505261762e-05,
+      "loss": 0.4336,
+      "step": 337
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.665063509461097e-05,
+      "loss": 0.3322,
+      "step": 338
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.662635371489404e-05,
+      "loss": 0.3514,
+      "step": 339
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.660199100478202e-05,
+      "loss": 0.4099,
+      "step": 340
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.657754705589591e-05,
+      "loss": 0.364,
+      "step": 341
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.655302196016228e-05,
+      "loss": 0.3298,
+      "step": 342
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.652841580981284e-05,
+      "loss": 0.2952,
+      "step": 343
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.650372869738414e-05,
+      "loss": 0.2466,
+      "step": 344
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.6478960715717176e-05,
+      "loss": 0.3615,
+      "step": 345
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.645411195795709e-05,
+      "loss": 0.3667,
+      "step": 346
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.642918251755281e-05,
+      "loss": 0.2582,
+      "step": 347
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.640417248825667e-05,
+      "loss": 0.2597,
+      "step": 348
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.6379081964124064e-05,
+      "loss": 0.3325,
+      "step": 349
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.6353911039513145e-05,
+      "loss": 0.4136,
+      "step": 350
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.6328659809084384e-05,
+      "loss": 0.2952,
+      "step": 351
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.6303328367800284e-05,
+      "loss": 0.448,
+      "step": 352
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.627791681092499e-05,
+      "loss": 0.3601,
+      "step": 353
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.6252425234023954e-05,
+      "loss": 0.4233,
+      "step": 354
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.622685373296353e-05,
+      "loss": 0.3501,
+      "step": 355
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.620120240391065e-05,
+      "loss": 0.3049,
+      "step": 356
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.6175471343332485e-05,
+      "loss": 0.242,
+      "step": 357
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.614966064799603e-05,
+      "loss": 0.384,
+      "step": 358
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.612377041496776e-05,
+      "loss": 0.2767,
+      "step": 359
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.609780074161327e-05,
+      "loss": 0.3445,
+      "step": 360
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.607175172559692e-05,
+      "loss": 0.3682,
+      "step": 361
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.604562346488144e-05,
+      "loss": 0.2524,
+      "step": 362
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.6019416057727585e-05,
+      "loss": 0.2311,
+      "step": 363
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.599312960269375e-05,
+      "loss": 0.3197,
+      "step": 364
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.5966764198635606e-05,
+      "loss": 0.2251,
+      "step": 365
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.5940319944705736e-05,
+      "loss": 0.3971,
+      "step": 366
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.591379694035325e-05,
+      "loss": 0.3965,
+      "step": 367
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.588719528532342e-05,
+      "loss": 0.303,
+      "step": 368
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.5860515079657276e-05,
+      "loss": 0.447,
+      "step": 369
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.583375642369129e-05,
+      "loss": 0.4007,
+      "step": 370
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.580691941805695e-05,
+      "loss": 0.3005,
+      "step": 371
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.5780004163680365e-05,
+      "loss": 0.3122,
+      "step": 372
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.5753010761781976e-05,
+      "loss": 0.33,
+      "step": 373
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.572593931387604e-05,
+      "loss": 0.2651,
+      "step": 374
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.569878992177039e-05,
+      "loss": 0.1777,
+      "step": 375
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.567156268756594e-05,
+      "loss": 0.2792,
+      "step": 376
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.5644257713656356e-05,
+      "loss": 0.4003,
+      "step": 377
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.561687510272767e-05,
+      "loss": 0.3523,
+      "step": 378
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.558941495775788e-05,
+      "loss": 0.3388,
+      "step": 379
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.556187738201656e-05,
+      "loss": 0.296,
+      "step": 380
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.553426247906448e-05,
+      "loss": 0.3271,
+      "step": 381
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.550657035275323e-05,
+      "loss": 0.4371,
+      "step": 382
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.54788011072248e-05,
+      "loss": 0.2513,
+      "step": 383
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.5450954846911195e-05,
+      "loss": 0.3197,
+      "step": 384
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.5423031676534065e-05,
+      "loss": 0.2382,
+      "step": 385
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.539503170110431e-05,
+      "loss": 0.3806,
+      "step": 386
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.536695502592162e-05,
+      "loss": 0.3329,
+      "step": 387
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.533880175657419e-05,
+      "loss": 0.3906,
+      "step": 388
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.531057199893824e-05,
+      "loss": 0.3553,
+      "step": 389
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.528226585917761e-05,
+      "loss": 0.3219,
+      "step": 390
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.525388344374344e-05,
+      "loss": 0.3331,
+      "step": 391
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.522542485937369e-05,
+      "loss": 0.2032,
+      "step": 392
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.519689021309278e-05,
+      "loss": 0.3207,
+      "step": 393
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.516827961221118e-05,
+      "loss": 0.374,
+      "step": 394
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.5139593164324986e-05,
+      "loss": 0.1848,
+      "step": 395
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.5110830977315556e-05,
+      "loss": 0.4075,
+      "step": 396
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.5081993159349056e-05,
+      "loss": 0.2669,
+      "step": 397
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.50530798188761e-05,
+      "loss": 0.3688,
+      "step": 398
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.50240910646313e-05,
+      "loss": 0.3771,
+      "step": 399
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.49950270056329e-05,
+      "loss": 0.2826,
+      "step": 400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.496588775118232e-05,
+      "loss": 0.3984,
+      "step": 401
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.493667341086379e-05,
+      "loss": 0.4486,
+      "step": 402
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.490738409454389e-05,
+      "loss": 0.2194,
+      "step": 403
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.48780199123712e-05,
+      "loss": 0.3468,
+      "step": 404
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.48485809747758e-05,
+      "loss": 0.4001,
+      "step": 405
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.4819067392468944e-05,
+      "loss": 0.2892,
+      "step": 406
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.478947927644258e-05,
+      "loss": 0.3491,
+      "step": 407
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.475981673796899e-05,
+      "loss": 0.2766,
+      "step": 408
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.473007988860027e-05,
+      "loss": 0.2798,
+      "step": 409
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.4700268840168045e-05,
+      "loss": 0.3565,
+      "step": 410
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.467038370478296e-05,
+      "loss": 0.2931,
+      "step": 411
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.464042459483425e-05,
+      "loss": 0.3839,
+      "step": 412
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.4610391622989396e-05,
+      "loss": 0.2549,
+      "step": 413
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.458028490219361e-05,
+      "loss": 0.3829,
+      "step": 414
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.455010454566947e-05,
+      "loss": 0.3571,
+      "step": 415
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.4519850666916484e-05,
+      "loss": 0.2678,
+      "step": 416
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.448952337971064e-05,
+      "loss": 0.2837,
+      "step": 417
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.4459122798104004e-05,
+      "loss": 0.7083,
+      "step": 418
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.442864903642428e-05,
+      "loss": 0.2896,
+      "step": 419
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.439810220927436e-05,
+      "loss": 0.3597,
+      "step": 420
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.436748243153196e-05,
+      "loss": 0.3126,
+      "step": 421
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.43367898183491e-05,
+      "loss": 0.3274,
+      "step": 422
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.430602448515173e-05,
+      "loss": 0.33,
+      "step": 423
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.4275186547639266e-05,
+      "loss": 0.2752,
+      "step": 424
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.4244276121784195e-05,
+      "loss": 0.3774,
+      "step": 425
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.4213293323831585e-05,
+      "loss": 0.4436,
+      "step": 426
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.418223827029867e-05,
+      "loss": 0.2347,
+      "step": 427
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.415111107797445e-05,
+      "loss": 0.3051,
+      "step": 428
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.411991186391919e-05,
+      "loss": 0.3971,
+      "step": 429
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.408864074546401e-05,
+      "loss": 0.2568,
+      "step": 430
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.405729784021046e-05,
+      "loss": 0.1899,
+      "step": 431
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.402588326603002e-05,
+      "loss": 0.2829,
+      "step": 432
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.3994397141063734e-05,
+      "loss": 0.4066,
+      "step": 433
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.396283958372173e-05,
+      "loss": 0.3069,
+      "step": 434
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.393121071268274e-05,
+      "loss": 0.2567,
+      "step": 435
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.38995106468937e-05,
+      "loss": 0.2195,
+      "step": 436
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.386773950556931e-05,
+      "loss": 0.3517,
+      "step": 437
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.3835897408191516e-05,
+      "loss": 0.3231,
+      "step": 438
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.380398447450918e-05,
+      "loss": 0.3845,
+      "step": 439
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.377200082453749e-05,
+      "loss": 0.3016,
+      "step": 440
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.37399465785576e-05,
+      "loss": 0.3339,
+      "step": 441
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.3707821857116176e-05,
+      "loss": 0.3208,
+      "step": 442
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.36756267810249e-05,
+      "loss": 0.2914,
+      "step": 443
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.364336147136004e-05,
+      "loss": 0.4541,
+      "step": 444
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.361102604946201e-05,
+      "loss": 0.33,
+      "step": 445
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.357862063693486e-05,
+      "loss": 0.5008,
+      "step": 446
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.354614535564588e-05,
+      "loss": 0.3086,
+      "step": 447
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.351360032772512e-05,
+      "loss": 0.3469,
+      "step": 448
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.34809856755649e-05,
+      "loss": 0.3614,
+      "step": 449
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.344830152181941e-05,
+      "loss": 0.4183,
+      "step": 450
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.34155479894042e-05,
+      "loss": 0.3402,
+      "step": 451
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.3382725201495723e-05,
+      "loss": 0.4674,
+      "step": 452
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.334983328153088e-05,
+      "loss": 0.2362,
+      "step": 453
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.3316872353206595e-05,
+      "loss": 0.4046,
+      "step": 454
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.3283842540479264e-05,
+      "loss": 0.3018,
+      "step": 455
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.325074396756437e-05,
+      "loss": 0.3368,
+      "step": 456
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.321757675893596e-05,
+      "loss": 0.2633,
+      "step": 457
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.318434103932622e-05,
+      "loss": 0.3504,
+      "step": 458
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.315103693372497e-05,
+      "loss": 0.3008,
+      "step": 459
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.3117664567379237e-05,
+      "loss": 0.2016,
+      "step": 460
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.30842240657927e-05,
+      "loss": 0.4831,
+      "step": 461
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.305071555472534e-05,
+      "loss": 0.4645,
+      "step": 462
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.301713916019287e-05,
+      "loss": 0.3625,
+      "step": 463
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.2983495008466276e-05,
+      "loss": 0.2665,
+      "step": 464
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.2949783226071406e-05,
+      "loss": 0.3006,
+      "step": 465
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.2916003939788403e-05,
+      "loss": 0.3767,
+      "step": 466
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.288215727665129e-05,
+      "loss": 0.3873,
+      "step": 467
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.2848243363947484e-05,
+      "loss": 0.3114,
+      "step": 468
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.281426232921728e-05,
+      "loss": 0.512,
+      "step": 469
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.278021430025343e-05,
+      "loss": 0.3207,
+      "step": 470
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.274609940510063e-05,
+      "loss": 0.3105,
+      "step": 471
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.2711917772055e-05,
+      "loss": 0.255,
+      "step": 472
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.267766952966369e-05,
+      "loss": 0.2192,
+      "step": 473
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.2643354806724336e-05,
+      "loss": 0.2394,
+      "step": 474
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.260897373228456e-05,
+      "loss": 0.325,
+      "step": 475
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.257452643564155e-05,
+      "loss": 0.249,
+      "step": 476
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.254001304634151e-05,
+      "loss": 0.2027,
+      "step": 477
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.2505433694179216e-05,
+      "loss": 0.2514,
+      "step": 478
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.24707885091975e-05,
+      "loss": 0.2721,
+      "step": 479
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.2436077621686786e-05,
+      "loss": 0.3148,
+      "step": 480
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.240130116218456e-05,
+      "loss": 0.2723,
+      "step": 481
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.2366459261474933e-05,
+      "loss": 0.2011,
+      "step": 482
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.233155205058811e-05,
+      "loss": 0.2669,
+      "step": 483
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.229657966079989e-05,
+      "loss": 0.3738,
+      "step": 484
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.226154222363124e-05,
+      "loss": 0.392,
+      "step": 485
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.22264398708477e-05,
+      "loss": 0.3212,
+      "step": 486
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.2191272734458955e-05,
+      "loss": 0.3804,
+      "step": 487
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.215604094671835e-05,
+      "loss": 0.3816,
+      "step": 488
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.21207446401223e-05,
+      "loss": 0.8537,
+      "step": 489
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.208538394740993e-05,
+      "loss": 0.1941,
+      "step": 490
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.2049959001562464e-05,
+      "loss": 0.312,
+      "step": 491
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.201446993580276e-05,
+      "loss": 0.401,
+      "step": 492
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.197891688359483e-05,
+      "loss": 0.5724,
+      "step": 493
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.194329997864331e-05,
+      "loss": 0.365,
+      "step": 494
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.1907619354892965e-05,
+      "loss": 0.2877,
+      "step": 495
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.1871875146528195e-05,
+      "loss": 0.4086,
+      "step": 496
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.183606748797251e-05,
+      "loss": 0.2989,
+      "step": 497
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.180019651388806e-05,
+      "loss": 0.2829,
+      "step": 498
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.176426235917509e-05,
+      "loss": 0.415,
+      "step": 499
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.172826515897146e-05,
+      "loss": 0.2476,
+      "step": 500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.16922050486521e-05,
+      "loss": 0.2687,
+      "step": 501
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.1656082163828566e-05,
+      "loss": 0.354,
+      "step": 502
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.1619896640348445e-05,
+      "loss": 0.2655,
+      "step": 503
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.1583648614294935e-05,
+      "loss": 0.313,
+      "step": 504
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.1547338221986266e-05,
+      "loss": 0.3046,
+      "step": 505
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.1510965599975196e-05,
+      "loss": 0.2319,
+      "step": 506
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.147453088504854e-05,
+      "loss": 0.3338,
+      "step": 507
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.14380342142266e-05,
+      "loss": 0.4914,
+      "step": 508
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.140147572476268e-05,
+      "loss": 0.234,
+      "step": 509
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.13648555541426e-05,
+      "loss": 0.4177,
+      "step": 510
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.1328173840084094e-05,
+      "loss": 0.2207,
+      "step": 511
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.129143072053638e-05,
+      "loss": 0.344,
+      "step": 512
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.125462633367959e-05,
+      "loss": 0.3432,
+      "step": 513
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.1217760817924254e-05,
+      "loss": 0.2881,
+      "step": 514
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.118083431191081e-05,
+      "loss": 0.293,
+      "step": 515
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.114384695450906e-05,
+      "loss": 0.1554,
+      "step": 516
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.110679888481763e-05,
+      "loss": 0.281,
+      "step": 517
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.1069690242163484e-05,
+      "loss": 0.3831,
+      "step": 518
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.103252116610139e-05,
+      "loss": 0.2113,
+      "step": 519
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.099529179641337e-05,
+      "loss": 0.2948,
+      "step": 520
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.095800227310821e-05,
+      "loss": 0.2097,
+      "step": 521
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.09206527364209e-05,
+      "loss": 0.3345,
+      "step": 522
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.0883243326812124e-05,
+      "loss": 0.3178,
+      "step": 523
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.0845774184967754e-05,
+      "loss": 0.3077,
+      "step": 524
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.080824545179828e-05,
+      "loss": 0.316,
+      "step": 525
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.077065726843828e-05,
+      "loss": 0.2975,
+      "step": 526
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.073300977624594e-05,
+      "loss": 0.2893,
+      "step": 527
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.069530311680247e-05,
+      "loss": 0.2935,
+      "step": 528
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.065753743191159e-05,
+      "loss": 0.222,
+      "step": 529
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.0619712863599e-05,
+      "loss": 0.3334,
+      "step": 530
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.058182955411187e-05,
+      "loss": 0.2667,
+      "step": 531
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.054388764591822e-05,
+      "loss": 0.3297,
+      "step": 532
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.05058872817065e-05,
+      "loss": 0.1771,
+      "step": 533
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.046782860438497e-05,
+      "loss": 0.2862,
+      "step": 534
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.042971175708118e-05,
+      "loss": 0.2862,
+      "step": 535
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.039153688314145e-05,
+      "loss": 0.316,
+      "step": 536
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.035330412613035e-05,
+      "loss": 0.1798,
+      "step": 537
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.0315013629830076e-05,
+      "loss": 0.2021,
+      "step": 538
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.0276665538239996e-05,
+      "loss": 0.2996,
+      "step": 539
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.0238259995576084e-05,
+      "loss": 0.1563,
+      "step": 540
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.019979714627036e-05,
+      "loss": 0.21,
+      "step": 541
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.0161277134970345e-05,
+      "loss": 0.3481,
+      "step": 542
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.012270010653854e-05,
+      "loss": 0.373,
+      "step": 543
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.0084066206051896e-05,
+      "loss": 0.2331,
+      "step": 544
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.0045375578801214e-05,
+      "loss": 0.2397,
+      "step": 545
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.000662837029062e-05,
+      "loss": 0.3247,
+      "step": 546
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.996782472623705e-05,
+      "loss": 0.2668,
+      "step": 547
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.9928964792569655e-05,
+      "loss": 0.2551,
+      "step": 548
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.989004871542931e-05,
+      "loss": 0.5247,
+      "step": 549
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.985107664116798e-05,
+      "loss": 0.4431,
+      "step": 550
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.981204871634827e-05,
+      "loss": 0.3068,
+      "step": 551
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.977296508774278e-05,
+      "loss": 0.345,
+      "step": 552
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.973382590233362e-05,
+      "loss": 0.3195,
+      "step": 553
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.969463130731183e-05,
+      "loss": 0.2284,
+      "step": 554
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.9655381450076826e-05,
+      "loss": 0.3369,
+      "step": 555
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.961607647823583e-05,
+      "loss": 0.2612,
+      "step": 556
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.957671653960337e-05,
+      "loss": 0.2798,
+      "step": 557
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.953730178220067e-05,
+      "loss": 0.2853,
+      "step": 558
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.9497832354255105e-05,
+      "loss": 0.3595,
+      "step": 559
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.945830840419966e-05,
+      "loss": 0.3869,
+      "step": 560
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.941873008067237e-05,
+      "loss": 0.2644,
+      "step": 561
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.9379097532515725e-05,
+      "loss": 0.2405,
+      "step": 562
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.933941090877615e-05,
+      "loss": 0.2137,
+      "step": 563
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.929967035870346e-05,
+      "loss": 0.3526,
+      "step": 564
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.925987603175023e-05,
+      "loss": 0.2582,
+      "step": 565
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.9220028077571295e-05,
+      "loss": 0.3193,
+      "step": 566
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.918012664602317e-05,
+      "loss": 0.3639,
+      "step": 567
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.914017188716347e-05,
+      "loss": 0.3945,
+      "step": 568
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.910016395125037e-05,
+      "loss": 0.2959,
+      "step": 569
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.9060102988742e-05,
+      "loss": 0.2896,
+      "step": 570
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.901998915029597e-05,
+      "loss": 0.3322,
+      "step": 571
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.897982258676867e-05,
+      "loss": 0.2928,
+      "step": 572
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.893960344921482e-05,
+      "loss": 0.3569,
+      "step": 573
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.889933188888684e-05,
+      "loss": 0.2259,
+      "step": 574
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.885900805723429e-05,
+      "loss": 0.3322,
+      "step": 575
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.881863210590332e-05,
+      "loss": 0.1517,
+      "step": 576
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.8778204186736076e-05,
+      "loss": 0.4635,
+      "step": 577
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.873772445177015e-05,
+      "loss": 0.3193,
+      "step": 578
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.8697193053238e-05,
+      "loss": 0.3252,
+      "step": 579
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.865661014356635e-05,
+      "loss": 0.3649,
+      "step": 580
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.861597587537568e-05,
+      "loss": 0.2961,
+      "step": 581
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.8575290401479586e-05,
+      "loss": 0.22,
+      "step": 582
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.8534553874884244e-05,
+      "loss": 0.2793,
+      "step": 583
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.8493766448787825e-05,
+      "loss": 0.2286,
+      "step": 584
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.8452928276579916e-05,
+      "loss": 0.2361,
+      "step": 585
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.841203951184095e-05,
+      "loss": 0.2495,
+      "step": 586
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.837110030834161e-05,
+      "loss": 0.2306,
+      "step": 587
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.8330110820042285e-05,
+      "loss": 0.3476,
+      "step": 588
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.828907120109245e-05,
+      "loss": 0.2691,
+      "step": 589
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.824798160583012e-05,
+      "loss": 0.7536,
+      "step": 590
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.8206842188781265e-05,
+      "loss": 0.2382,
+      "step": 591
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.8165653104659185e-05,
+      "loss": 0.2825,
+      "step": 592
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.8124414508364e-05,
+      "loss": 0.352,
+      "step": 593
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.8083126554982024e-05,
+      "loss": 0.1925,
+      "step": 594
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.804178939978517e-05,
+      "loss": 0.238,
+      "step": 595
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.8000403198230387e-05,
+      "loss": 0.2963,
+      "step": 596
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.79589681059591e-05,
+      "loss": 0.2556,
+      "step": 597
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.791748427879658e-05,
+      "loss": 0.2802,
+      "step": 598
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.787595187275136e-05,
+      "loss": 0.188,
+      "step": 599
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.783437104401469e-05,
+      "loss": 0.4299,
+      "step": 600
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.779274194895991e-05,
+      "loss": 0.2196,
+      "step": 601
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.775106474414188e-05,
+      "loss": 0.2143,
+      "step": 602
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.770933958629639e-05,
+      "loss": 0.2819,
+      "step": 603
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.766756663233956e-05,
+      "loss": 0.4042,
+      "step": 604
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.762574603936725e-05,
+      "loss": 0.5486,
+      "step": 605
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.75838779646545e-05,
+      "loss": 0.41,
+      "step": 606
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.75419625656549e-05,
+      "loss": 0.2641,
+      "step": 607
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.2676,
+      "step": 608
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.745799042549876e-05,
+      "loss": 0.3497,
+      "step": 609
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.74159340001369e-05,
+      "loss": 0.2294,
+      "step": 610
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.7373830882076354e-05,
+      "loss": 0.3533,
+      "step": 611
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.7331681229654635e-05,
+      "loss": 0.2693,
+      "step": 612
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.728948520138427e-05,
+      "loss": 0.2985,
+      "step": 613
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.7247242955952175e-05,
+      "loss": 0.2668,
+      "step": 614
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.7204954652219104e-05,
+      "loss": 0.246,
+      "step": 615
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.7162620449219e-05,
+      "loss": 0.2588,
+      "step": 616
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.712024050615843e-05,
+      "loss": 0.3915,
+      "step": 617
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.707781498241596e-05,
+      "loss": 0.2716,
+      "step": 618
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.7035344037541606e-05,
+      "loss": 0.3732,
+      "step": 619
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.699282783125616e-05,
+      "loss": 0.1905,
+      "step": 620
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.695026652345065e-05,
+      "loss": 0.4051,
+      "step": 621
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.690766027418573e-05,
+      "loss": 0.2956,
+      "step": 622
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.686500924369101e-05,
+      "loss": 0.1681,
+      "step": 623
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.6822313592364594e-05,
+      "loss": 0.2475,
+      "step": 624
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.6779573480772325e-05,
+      "loss": 0.3389,
+      "step": 625
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.673678906964727e-05,
+      "loss": 0.2716,
+      "step": 626
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.6693960519889106e-05,
+      "loss": 0.3142,
+      "step": 627
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.665108799256348e-05,
+      "loss": 0.3198,
+      "step": 628
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.660817164890143e-05,
+      "loss": 0.3937,
+      "step": 629
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.656521165029879e-05,
+      "loss": 0.2,
+      "step": 630
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.652220815831555e-05,
+      "loss": 0.3395,
+      "step": 631
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.6479161334675296e-05,
+      "loss": 0.2961,
+      "step": 632
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.643607134126452e-05,
+      "loss": 0.3635,
+      "step": 633
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.639293834013211e-05,
+      "loss": 0.2849,
+      "step": 634
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.634976249348867e-05,
+      "loss": 0.3942,
+      "step": 635
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.630654396370594e-05,
+      "loss": 0.1797,
+      "step": 636
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.626328291331618e-05,
+      "loss": 0.3187,
+      "step": 637
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.621997950501156e-05,
+      "loss": 0.2628,
+      "step": 638
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.617663390164353e-05,
+      "loss": 0.2839,
+      "step": 639
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.613324626622224e-05,
+      "loss": 0.2527,
+      "step": 640
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.6089816761915906e-05,
+      "loss": 0.2961,
+      "step": 641
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.60463455520502e-05,
+      "loss": 0.371,
+      "step": 642
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.600283280010762e-05,
+      "loss": 0.2223,
+      "step": 643
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.5959278669726935e-05,
+      "loss": 0.4677,
+      "step": 644
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.591568332470249e-05,
+      "loss": 0.3137,
+      "step": 645
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.5872046928983626e-05,
+      "loss": 0.2421,
+      "step": 646
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.582836964667408e-05,
+      "loss": 0.3713,
+      "step": 647
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.578465164203134e-05,
+      "loss": 0.273,
+      "step": 648
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.5740893079466045e-05,
+      "loss": 0.193,
+      "step": 649
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.569709412354136e-05,
+      "loss": 0.3101,
+      "step": 650
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.565325493897235e-05,
+      "loss": 0.2809,
+      "step": 651
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.560937569062538e-05,
+      "loss": 0.3119,
+      "step": 652
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.556545654351749e-05,
+      "loss": 0.41,
+      "step": 653
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.552149766281573e-05,
+      "loss": 0.2206,
+      "step": 654
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.5477499213836616e-05,
+      "loss": 0.196,
+      "step": 655
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.543346136204545e-05,
+      "loss": 0.2974,
+      "step": 656
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.538938427305573e-05,
+      "loss": 0.1934,
+      "step": 657
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.534526811262848e-05,
+      "loss": 0.2859,
+      "step": 658
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.5301113046671714e-05,
+      "loss": 0.387,
+      "step": 659
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.525691924123971e-05,
+      "loss": 0.2069,
+      "step": 660
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.521268686253246e-05,
+      "loss": 0.3205,
+      "step": 661
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.516841607689501e-05,
+      "loss": 0.1568,
+      "step": 662
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.512410705081684e-05,
+      "loss": 0.2801,
+      "step": 663
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.507975995093126e-05,
+      "loss": 0.3729,
+      "step": 664
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.503537494401473e-05,
+      "loss": 0.2849,
+      "step": 665
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.499095219698631e-05,
+      "loss": 0.2929,
+      "step": 666
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.494649187690695e-05,
+      "loss": 0.1612,
+      "step": 667
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.490199415097892e-05,
+      "loss": 0.1769,
+      "step": 668
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.4857459186545174e-05,
+      "loss": 0.313,
+      "step": 669
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.481288715108868e-05,
+      "loss": 0.27,
+      "step": 670
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.476827821223184e-05,
+      "loss": 0.3673,
+      "step": 671
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.4723632537735846e-05,
+      "loss": 0.3069,
+      "step": 672
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.4678950295500015e-05,
+      "loss": 0.2208,
+      "step": 673
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.463423165356121e-05,
+      "loss": 0.2773,
+      "step": 674
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.4589476780093166e-05,
+      "loss": 0.2973,
+      "step": 675
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.454468584340588e-05,
+      "loss": 0.4149,
+      "step": 676
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.449985901194498e-05,
+      "loss": 0.3398,
+      "step": 677
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.445499645429107e-05,
+      "loss": 0.3351,
+      "step": 678
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.441009833915911e-05,
+      "loss": 0.3106,
+      "step": 679
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.436516483539781e-05,
+      "loss": 0.1937,
+      "step": 680
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.4320196111988916e-05,
+      "loss": 0.3237,
+      "step": 681
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.427519233804667e-05,
+      "loss": 0.3103,
+      "step": 682
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.423015368281711e-05,
+      "loss": 0.2943,
+      "step": 683
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.4185080315677454e-05,
+      "loss": 0.2816,
+      "step": 684
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.4139972406135464e-05,
+      "loss": 0.2863,
+      "step": 685
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.409483012382879e-05,
+      "loss": 0.3009,
+      "step": 686
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.404965363852437e-05,
+      "loss": 0.3009,
+      "step": 687
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.400444312011776e-05,
+      "loss": 0.2604,
+      "step": 688
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.39591987386325e-05,
+      "loss": 0.2639,
+      "step": 689
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.39139206642195e-05,
+      "loss": 0.2918,
+      "step": 690
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.386860906715635e-05,
+      "loss": 0.5156,
+      "step": 691
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.382326411784672e-05,
+      "loss": 0.2907,
+      "step": 692
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.377788598681972e-05,
+      "loss": 0.3794,
+      "step": 693
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.373247484472923e-05,
+      "loss": 0.3009,
+      "step": 694
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.3687030862353286e-05,
+      "loss": 0.2779,
+      "step": 695
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.364155421059342e-05,
+      "loss": 0.3246,
+      "step": 696
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.359604506047403e-05,
+      "loss": 0.2393,
+      "step": 697
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.355050358314172e-05,
+      "loss": 0.5318,
+      "step": 698
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.350492994986468e-05,
+      "loss": 0.3002,
+      "step": 699
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.3459324332032035e-05,
+      "loss": 0.1798,
+      "step": 700
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.3413686901153165e-05,
+      "loss": 0.2082,
+      "step": 701
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.336801782885712e-05,
+      "loss": 0.3749,
+      "step": 702
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.3322317286891913e-05,
+      "loss": 0.2748,
+      "step": 703
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.327658544712395e-05,
+      "loss": 0.3169,
+      "step": 704
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.32308224815373e-05,
+      "loss": 0.2393,
+      "step": 705
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.318502856223311e-05,
+      "loss": 0.1476,
+      "step": 706
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.313920386142892e-05,
+      "loss": 0.1642,
+      "step": 707
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.309334855145803e-05,
+      "loss": 0.2829,
+      "step": 708
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.304746280476886e-05,
+      "loss": 0.2958,
+      "step": 709
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.3001546793924285e-05,
+      "loss": 0.2265,
+      "step": 710
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.295560069160102e-05,
+      "loss": 0.2271,
+      "step": 711
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.290962467058891e-05,
+      "loss": 0.259,
+      "step": 712
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.2863618903790346e-05,
+      "loss": 0.3264,
+      "step": 713
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.281758356421955e-05,
+      "loss": 0.3091,
+      "step": 714
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.277151882500199e-05,
+      "loss": 0.745,
+      "step": 715
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.272542485937369e-05,
+      "loss": 0.2028,
+      "step": 716
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.267930184068057e-05,
+      "loss": 0.351,
+      "step": 717
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.2633149942377834e-05,
+      "loss": 0.231,
+      "step": 718
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.2586969338029274e-05,
+      "loss": 0.3303,
+      "step": 719
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.254076020130664e-05,
+      "loss": 0.2304,
+      "step": 720
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.2494522705989e-05,
+      "loss": 0.3012,
+      "step": 721
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.244825702596205e-05,
+      "loss": 0.1952,
+      "step": 722
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.24019633352175e-05,
+      "loss": 0.265,
+      "step": 723
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.235564180785238e-05,
+      "loss": 0.2814,
+      "step": 724
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.230929261806842e-05,
+      "loss": 0.2583,
+      "step": 725
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.2262915940171376e-05,
+      "loss": 0.2556,
+      "step": 726
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.2216511948570374e-05,
+      "loss": 0.2894,
+      "step": 727
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.217008081777726e-05,
+      "loss": 0.3761,
+      "step": 728
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.212362272240595e-05,
+      "loss": 0.36,
+      "step": 729
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.207713783717176e-05,
+      "loss": 0.3217,
+      "step": 730
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.203062633689077e-05,
+      "loss": 0.3445,
+      "step": 731
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.198408839647911e-05,
+      "loss": 0.3089,
+      "step": 732
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.193752419095239e-05,
+      "loss": 0.2837,
+      "step": 733
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.1890933895424976e-05,
+      "loss": 0.3138,
+      "step": 734
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.1844317685109354e-05,
+      "loss": 0.3639,
+      "step": 735
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.1797675735315455e-05,
+      "loss": 0.2605,
+      "step": 736
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.1751008221450025e-05,
+      "loss": 0.2831,
+      "step": 737
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.170431531901594e-05,
+      "loss": 0.1588,
+      "step": 738
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.1657597203611554e-05,
+      "loss": 0.2759,
+      "step": 739
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.161085405093006e-05,
+      "loss": 0.2487,
+      "step": 740
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.156408603675878e-05,
+      "loss": 0.205,
+      "step": 741
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.151729333697854e-05,
+      "loss": 0.4166,
+      "step": 742
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.147047612756302e-05,
+      "loss": 0.2017,
+      "step": 743
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.1423634584578055e-05,
+      "loss": 0.3748,
+      "step": 744
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.137676888418099e-05,
+      "loss": 0.1318,
+      "step": 745
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.132987920262005e-05,
+      "loss": 0.2815,
+      "step": 746
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.1282965716233594e-05,
+      "loss": 0.4187,
+      "step": 747
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.1236028601449534e-05,
+      "loss": 0.2007,
+      "step": 748
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.118906803478465e-05,
+      "loss": 0.2955,
+      "step": 749
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.114208419284391e-05,
+      "loss": 0.3559,
+      "step": 750
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.10950772523198e-05,
+      "loss": 0.322,
+      "step": 751
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.104804738999169e-05,
+      "loss": 0.2595,
+      "step": 752
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.100099478272515e-05,
+      "loss": 0.3493,
+      "step": 753
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.0953919607471275e-05,
+      "loss": 0.2842,
+      "step": 754
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.090682204126604e-05,
+      "loss": 0.2658,
+      "step": 755
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.085970226122962e-05,
+      "loss": 0.1989,
+      "step": 756
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.0812560444565745e-05,
+      "loss": 0.1445,
+      "step": 757
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.076539676856101e-05,
+      "loss": 0.3248,
+      "step": 758
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.07182114105842e-05,
+      "loss": 0.2059,
+      "step": 759
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.0671004548085675e-05,
+      "loss": 0.2601,
+      "step": 760
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.062377635859663e-05,
+      "loss": 0.2539,
+      "step": 761
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.057652701972848e-05,
+      "loss": 0.281,
+      "step": 762
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.0529256709172195e-05,
+      "loss": 0.1776,
+      "step": 763
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.048196560469758e-05,
+      "loss": 0.3513,
+      "step": 764
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.043465388415267e-05,
+      "loss": 0.339,
+      "step": 765
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.0387321725463e-05,
+      "loss": 0.1377,
+      "step": 766
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.0339969306631005e-05,
+      "loss": 0.1687,
+      "step": 767
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.0292596805735274e-05,
+      "loss": 0.1796,
+      "step": 768
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.0245204400929934e-05,
+      "loss": 0.2331,
+      "step": 769
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.0197792270443982e-05,
+      "loss": 0.2069,
+      "step": 770
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.015036059258058e-05,
+      "loss": 0.2889,
+      "step": 771
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.0102909545716396e-05,
+      "loss": 0.3174,
+      "step": 772
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.0055439308300952e-05,
+      "loss": 0.4017,
+      "step": 773
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.0007950058855937e-05,
+      "loss": 0.25,
+      "step": 774
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.9960441975974534e-05,
+      "loss": 0.222,
+      "step": 775
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.9912915238320754e-05,
+      "loss": 0.2477,
+      "step": 776
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.9865370024628775e-05,
+      "loss": 0.3002,
+      "step": 777
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.9817806513702244e-05,
+      "loss": 0.3163,
+      "step": 778
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.9770224884413623e-05,
+      "loss": 0.3735,
+      "step": 779
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.9722625315703512e-05,
+      "loss": 0.2001,
+      "step": 780
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.9675007986579977e-05,
+      "loss": 0.4169,
+      "step": 781
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.9627373076117863e-05,
+      "loss": 0.2539,
+      "step": 782
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.957972076345817e-05,
+      "loss": 0.3454,
+      "step": 783
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.9532051227807296e-05,
+      "loss": 0.2849,
+      "step": 784
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.9484364648436437e-05,
+      "loss": 0.3067,
+      "step": 785
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.9436661204680882e-05,
+      "loss": 0.2424,
+      "step": 786
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.9388941075939334e-05,
+      "loss": 0.2334,
+      "step": 787
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.9341204441673266e-05,
+      "loss": 0.3158,
+      "step": 788
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.9293451481406174e-05,
+      "loss": 0.2411,
+      "step": 789
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.9245682374723016e-05,
+      "loss": 0.4071,
+      "step": 790
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.9197897301269435e-05,
+      "loss": 0.2498,
+      "step": 791
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.9150096440751107e-05,
+      "loss": 0.2204,
+      "step": 792
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.9102279972933106e-05,
+      "loss": 0.2728,
+      "step": 793
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.905444807763919e-05,
+      "loss": 0.2423,
+      "step": 794
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.9006600934751145e-05,
+      "loss": 0.3622,
+      "step": 795
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.8958738724208072e-05,
+      "loss": 0.2328,
+      "step": 796
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.8910861626005776e-05,
+      "loss": 0.2866,
+      "step": 797
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.8862969820196016e-05,
+      "loss": 0.2753,
+      "step": 798
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.8815063486885878e-05,
+      "loss": 0.2671,
+      "step": 799
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.876714280623708e-05,
+      "loss": 0.2786,
+      "step": 800
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.87192079584653e-05,
+      "loss": 0.3364,
+      "step": 801
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.8671259123839472e-05,
+      "loss": 0.273,
+      "step": 802
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.8623296482681166e-05,
+      "loss": 0.3908,
+      "step": 803
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.8575320215363855e-05,
+      "loss": 0.4106,
+      "step": 804
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.8527330502312248e-05,
+      "loss": 0.2356,
+      "step": 805
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.8479327524001636e-05,
+      "loss": 0.2324,
+      "step": 806
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.843131146095719e-05,
+      "loss": 0.215,
+      "step": 807
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.8383282493753283e-05,
+      "loss": 0.2047,
+      "step": 808
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.833524080301282e-05,
+      "loss": 0.2984,
+      "step": 809
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.8287186569406566e-05,
+      "loss": 0.2273,
+      "step": 810
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.8239119973652444e-05,
+      "loss": 0.1912,
+      "step": 811
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.8191041196514873e-05,
+      "loss": 0.2427,
+      "step": 812
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.814295041880407e-05,
+      "loss": 0.287,
+      "step": 813
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.8094847821375414e-05,
+      "loss": 0.2624,
+      "step": 814
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.8046733585128687e-05,
+      "loss": 0.415,
+      "step": 815
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.7998607891007495e-05,
+      "loss": 0.2809,
+      "step": 816
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.795047091999849e-05,
+      "loss": 0.2281,
+      "step": 817
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.7902322853130757e-05,
+      "loss": 0.2843,
+      "step": 818
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.7854163871475103e-05,
+      "loss": 0.2133,
+      "step": 819
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.7805994156143376e-05,
+      "loss": 0.2125,
+      "step": 820
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.7757813888287798e-05,
+      "loss": 0.247,
+      "step": 821
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.770962324910027e-05,
+      "loss": 0.2505,
+      "step": 822
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.7661422419811707e-05,
+      "loss": 0.2746,
+      "step": 823
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.761321158169134e-05,
+      "loss": 0.3182,
+      "step": 824
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.756499091604603e-05,
+      "loss": 0.1577,
+      "step": 825
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.7516760604219617e-05,
+      "loss": 0.4288,
+      "step": 826
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.7468520827592197e-05,
+      "loss": 0.3321,
+      "step": 827
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.742027176757948e-05,
+      "loss": 0.3154,
+      "step": 828
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.7372013605632068e-05,
+      "loss": 0.2171,
+      "step": 829
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.732374652323481e-05,
+      "loss": 0.3084,
+      "step": 830
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.7275470701906088e-05,
+      "loss": 0.3476,
+      "step": 831
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.7227186323197162e-05,
+      "loss": 0.2265,
+      "step": 832
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.717889356869146e-05,
+      "loss": 0.2544,
+      "step": 833
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.7130592620003918e-05,
+      "loss": 0.2601,
+      "step": 834
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.7082283658780288e-05,
+      "loss": 0.2604,
+      "step": 835
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.7033966866696457e-05,
+      "loss": 0.2447,
+      "step": 836
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.6985642425457757e-05,
+      "loss": 0.3529,
+      "step": 837
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.6937310516798275e-05,
+      "loss": 0.3681,
+      "step": 838
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.6888971322480205e-05,
+      "loss": 0.2899,
+      "step": 839
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.684062502429312e-05,
+      "loss": 0.1836,
+      "step": 840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6792271804053325e-05,
+      "loss": 0.2551,
+      "step": 841
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.674391184360313e-05,
+      "loss": 0.2387,
+      "step": 842
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6695545324810235e-05,
+      "loss": 0.2874,
+      "step": 843
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6647172429566963e-05,
+      "loss": 0.2223,
+      "step": 844
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.2612282335758209,
+      "eval_runtime": 59.6019,
+      "eval_samples_per_second": 5.033,
+      "eval_steps_per_second": 0.638,
+      "step": 844
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.659879333978964e-05,
+      "loss": 0.1381,
+      "step": 845
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6550408237417885e-05,
+      "loss": 0.1698,
+      "step": 846
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.650201730441392e-05,
+      "loss": 0.171,
+      "step": 847
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6453620722761896e-05,
+      "loss": 0.2113,
+      "step": 848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6405218674467225e-05,
+      "loss": 0.0948,
+      "step": 849
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.635681134155585e-05,
+      "loss": 0.1014,
+      "step": 850
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.63083989060736e-05,
+      "loss": 0.1089,
+      "step": 851
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6259981550085504e-05,
+      "loss": 0.284,
+      "step": 852
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.621155945567508e-05,
+      "loss": 0.0688,
+      "step": 853
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6163132804943675e-05,
+      "loss": 0.1787,
+      "step": 854
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6114701780009753e-05,
+      "loss": 0.146,
+      "step": 855
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6066266563008267e-05,
+      "loss": 0.1663,
+      "step": 856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.60178273360899e-05,
+      "loss": 0.1115,
+      "step": 857
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5969384281420424e-05,
+      "loss": 0.17,
+      "step": 858
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5920937581180004e-05,
+      "loss": 0.14,
+      "step": 859
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.587248741756253e-05,
+      "loss": 0.0993,
+      "step": 860
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5824033972774896e-05,
+      "loss": 0.1213,
+      "step": 861
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5775577429036345e-05,
+      "loss": 0.2115,
+      "step": 862
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5727117968577784e-05,
+      "loss": 0.1085,
+      "step": 863
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5678655773641075e-05,
+      "loss": 0.1225,
+      "step": 864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5630191026478368e-05,
+      "loss": 0.2955,
+      "step": 865
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.5581723909351406e-05,
+      "loss": 0.1452,
+      "step": 866
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.553325460453086e-05,
+      "loss": 0.1186,
+      "step": 867
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.548478329429561e-05,
+      "loss": 0.1242,
+      "step": 868
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.5436310160932092e-05,
+      "loss": 0.1396,
+      "step": 869
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.5387835386733584e-05,
+      "loss": 0.1392,
+      "step": 870
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.5339359153999553e-05,
+      "loss": 0.1301,
+      "step": 871
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.5290881645034932e-05,
+      "loss": 0.1539,
+      "step": 872
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.5242403042149466e-05,
+      "loss": 0.1007,
+      "step": 873
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.519392352765701e-05,
+      "loss": 0.1181,
+      "step": 874
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.5145443283874848e-05,
+      "loss": 0.1618,
+      "step": 875
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.5096962493123012e-05,
+      "loss": 0.1233,
+      "step": 876
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.504848133772358e-05,
+      "loss": 0.1469,
+      "step": 877
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.5e-05,
+      "loss": 0.1318,
+      "step": 878
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.4951518662276422e-05,
+      "loss": 0.1114,
+      "step": 879
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.4903037506876997e-05,
+      "loss": 0.1648,
+      "step": 880
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.485455671612515e-05,
+      "loss": 0.1411,
+      "step": 881
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.4806076472342997e-05,
+      "loss": 0.1075,
+      "step": 882
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.475759695785054e-05,
+      "loss": 0.1541,
+      "step": 883
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.470911835496508e-05,
+      "loss": 0.1165,
+      "step": 884
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.4660640846000453e-05,
+      "loss": 0.2047,
+      "step": 885
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.461216461326642e-05,
+      "loss": 0.1234,
+      "step": 886
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.4563689839067913e-05,
+      "loss": 0.163,
+      "step": 887
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.4515216705704395e-05,
+      "loss": 0.1178,
+      "step": 888
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.4466745395469143e-05,
+      "loss": 0.1457,
+      "step": 889
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.4418276090648596e-05,
+      "loss": 0.1167,
+      "step": 890
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.4369808973521638e-05,
+      "loss": 0.186,
+      "step": 891
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.432134422635893e-05,
+      "loss": 0.105,
+      "step": 892
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.4272882031422215e-05,
+      "loss": 0.1301,
+      "step": 893
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.4224422570963658e-05,
+      "loss": 0.1562,
+      "step": 894
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.4175966027225107e-05,
+      "loss": 0.1136,
+      "step": 895
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.4127512582437485e-05,
+      "loss": 0.1492,
+      "step": 896
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.4079062418820002e-05,
+      "loss": 0.1399,
+      "step": 897
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.403061571857958e-05,
+      "loss": 0.1513,
+      "step": 898
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.3982172663910108e-05,
+      "loss": 0.2295,
+      "step": 899
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.3933733436991732e-05,
+      "loss": 0.1692,
+      "step": 900
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.3885298219990253e-05,
+      "loss": 0.1224,
+      "step": 901
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.3836867195056335e-05,
+      "loss": 0.1251,
+      "step": 902
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.378844054432493e-05,
+      "loss": 0.1526,
+      "step": 903
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.3740018449914502e-05,
+      "loss": 0.1625,
+      "step": 904
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.3691601093926404e-05,
+      "loss": 0.1305,
+      "step": 905
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.364318865844416e-05,
+      "loss": 0.1411,
+      "step": 906
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.3594781325532784e-05,
+      "loss": 0.1862,
+      "step": 907
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.3546379277238107e-05,
+      "loss": 0.1878,
+      "step": 908
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.3497982695586087e-05,
+      "loss": 0.1419,
+      "step": 909
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.344959176258212e-05,
+      "loss": 0.169,
+      "step": 910
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.3401206660210363e-05,
+      "loss": 0.1143,
+      "step": 911
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.3352827570433036e-05,
+      "loss": 0.18,
+      "step": 912
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.330445467518977e-05,
+      "loss": 0.2251,
+      "step": 913
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.3256088156396868e-05,
+      "loss": 0.1889,
+      "step": 914
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.3207728195946688e-05,
+      "loss": 0.1149,
+      "step": 915
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.3159374975706884e-05,
+      "loss": 0.1248,
+      "step": 916
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.3111028677519804e-05,
+      "loss": 0.1786,
+      "step": 917
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.306268948320173e-05,
+      "loss": 0.1793,
+      "step": 918
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.3014357574542245e-05,
+      "loss": 0.2057,
+      "step": 919
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.2966033133303545e-05,
+      "loss": 0.1818,
+      "step": 920
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.2917716341219707e-05,
+      "loss": 0.0988,
+      "step": 921
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.2869407379996088e-05,
+      "loss": 0.1705,
+      "step": 922
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.2821106431308544e-05,
+      "loss": 0.2104,
+      "step": 923
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.277281367680285e-05,
+      "loss": 0.1397,
+      "step": 924
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.2724529298093915e-05,
+      "loss": 0.1264,
+      "step": 925
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.2676253476765196e-05,
+      "loss": 0.0956,
+      "step": 926
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.2627986394367938e-05,
+      "loss": 0.1525,
+      "step": 927
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.2579728232420525e-05,
+      "loss": 0.1733,
+      "step": 928
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.2531479172407805e-05,
+      "loss": 0.0823,
+      "step": 929
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.248323939578039e-05,
+      "loss": 0.186,
+      "step": 930
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.2435009083953978e-05,
+      "loss": 0.1517,
+      "step": 931
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.238678841830867e-05,
+      "loss": 0.1051,
+      "step": 932
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.2338577580188296e-05,
+      "loss": 0.1284,
+      "step": 933
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.229037675089974e-05,
+      "loss": 0.1083,
+      "step": 934
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.2242186111712208e-05,
+      "loss": 0.1427,
+      "step": 935
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.2194005843856636e-05,
+      "loss": 0.1045,
+      "step": 936
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.2145836128524902e-05,
+      "loss": 0.2202,
+      "step": 937
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.2097677146869242e-05,
+      "loss": 0.1509,
+      "step": 938
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.2049529080001512e-05,
+      "loss": 0.1039,
+      "step": 939
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.2001392108992504e-05,
+      "loss": 0.1045,
+      "step": 940
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.195326641487132e-05,
+      "loss": 0.1196,
+      "step": 941
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.1905152178624595e-05,
+      "loss": 0.1297,
+      "step": 942
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.1857049581195938e-05,
+      "loss": 0.2113,
+      "step": 943
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.1808958803485136e-05,
+      "loss": 0.1321,
+      "step": 944
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.1760880026347562e-05,
+      "loss": 0.145,
+      "step": 945
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.1712813430593436e-05,
+      "loss": 0.1424,
+      "step": 946
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.1664759196987182e-05,
+      "loss": 0.1185,
+      "step": 947
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.161671750624673e-05,
+      "loss": 0.1664,
+      "step": 948
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.1568688539042816e-05,
+      "loss": 0.132,
+      "step": 949
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.1520672475998373e-05,
+      "loss": 0.1768,
+      "step": 950
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.1472669497687754e-05,
+      "loss": 0.1476,
+      "step": 951
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.1424679784636144e-05,
+      "loss": 0.1619,
+      "step": 952
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.1376703517318837e-05,
+      "loss": 0.1164,
+      "step": 953
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.1328740876160527e-05,
+      "loss": 0.1415,
+      "step": 954
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.1280792041534714e-05,
+      "loss": 0.1162,
+      "step": 955
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.1232857193762924e-05,
+      "loss": 0.1799,
+      "step": 956
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.118493651311413e-05,
+      "loss": 0.2352,
+      "step": 957
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.113703017980399e-05,
+      "loss": 0.1966,
+      "step": 958
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.1089138373994223e-05,
+      "loss": 0.2398,
+      "step": 959
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.1041261275791933e-05,
+      "loss": 0.2376,
+      "step": 960
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.099339906524886e-05,
+      "loss": 0.1184,
+      "step": 961
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.0945551922360818e-05,
+      "loss": 0.1683,
+      "step": 962
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.08977200270669e-05,
+      "loss": 0.0813,
+      "step": 963
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.0849903559248906e-05,
+      "loss": 0.1073,
+      "step": 964
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.0802102698730574e-05,
+      "loss": 0.1239,
+      "step": 965
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.0754317625276983e-05,
+      "loss": 0.1072,
+      "step": 966
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.070654851859383e-05,
+      "loss": 0.1826,
+      "step": 967
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.0658795558326743e-05,
+      "loss": 0.124,
+      "step": 968
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.061105892406067e-05,
+      "loss": 0.1454,
+      "step": 969
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.0563338795319123e-05,
+      "loss": 0.2054,
+      "step": 970
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.0515635351563565e-05,
+      "loss": 0.0783,
+      "step": 971
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.0467948772192713e-05,
+      "loss": 0.124,
+      "step": 972
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.0420279236541832e-05,
+      "loss": 0.1955,
+      "step": 973
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.037262692388214e-05,
+      "loss": 0.1657,
+      "step": 974
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.032499201342003e-05,
+      "loss": 0.1476,
+      "step": 975
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.02773746842965e-05,
+      "loss": 0.157,
+      "step": 976
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.022977511558638e-05,
+      "loss": 0.1258,
+      "step": 977
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.0182193486297755e-05,
+      "loss": 0.1135,
+      "step": 978
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.0134629975371227e-05,
+      "loss": 0.1333,
+      "step": 979
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.0087084761679245e-05,
+      "loss": 0.1336,
+      "step": 980
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.0039558024025475e-05,
+      "loss": 0.1739,
+      "step": 981
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.9992049941144066e-05,
+      "loss": 0.1297,
+      "step": 982
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.9944560691699057e-05,
+      "loss": 0.1465,
+      "step": 983
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.9897090454283613e-05,
+      "loss": 0.1423,
+      "step": 984
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.9849639407419423e-05,
+      "loss": 0.1315,
+      "step": 985
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.980220772955602e-05,
+      "loss": 0.2371,
+      "step": 986
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.9754795599070068e-05,
+      "loss": 0.102,
+      "step": 987
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.970740319426474e-05,
+      "loss": 0.1585,
+      "step": 988
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.9660030693369004e-05,
+      "loss": 0.1208,
+      "step": 989
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.9612678274537005e-05,
+      "loss": 0.1956,
+      "step": 990
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.9565346115847337e-05,
+      "loss": 0.0686,
+      "step": 991
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.9518034395302414e-05,
+      "loss": 0.1391,
+      "step": 992
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.9470743290827808e-05,
+      "loss": 0.1147,
+      "step": 993
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.942347298027152e-05,
+      "loss": 0.1018,
+      "step": 994
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.937622364140338e-05,
+      "loss": 0.1902,
+      "step": 995
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.932899545191433e-05,
+      "loss": 0.1892,
+      "step": 996
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.9281788589415804e-05,
+      "loss": 0.146,
+      "step": 997
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.9234603231438995e-05,
+      "loss": 0.1023,
+      "step": 998
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.9187439555434254e-05,
+      "loss": 0.1287,
+      "step": 999
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.9140297738770385e-05,
+      "loss": 0.1345,
+      "step": 1000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.9093177958733966e-05,
+      "loss": 0.1012,
+      "step": 1001
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.9046080392528735e-05,
+      "loss": 0.1274,
+      "step": 1002
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.8999005217274857e-05,
+      "loss": 0.1094,
+      "step": 1003
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.895195261000831e-05,
+      "loss": 0.0979,
+      "step": 1004
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.8904922747680204e-05,
+      "loss": 0.1045,
+      "step": 1005
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.8857915807156092e-05,
+      "loss": 0.2168,
+      "step": 1006
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.8810931965215356e-05,
+      "loss": 0.114,
+      "step": 1007
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.876397139855047e-05,
+      "loss": 0.1452,
+      "step": 1008
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.8717034283766422e-05,
+      "loss": 0.2213,
+      "step": 1009
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.8670120797379958e-05,
+      "loss": 0.203,
+      "step": 1010
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.8623231115819002e-05,
+      "loss": 0.1415,
+      "step": 1011
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.857636541542195e-05,
+      "loss": 0.0695,
+      "step": 1012
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.852952387243698e-05,
+      "loss": 0.1679,
+      "step": 1013
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.8482706663021465e-05,
+      "loss": 0.0949,
+      "step": 1014
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.8435913963241226e-05,
+      "loss": 0.153,
+      "step": 1015
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.838914594906995e-05,
+      "loss": 0.126,
+      "step": 1016
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.8342402796388445e-05,
+      "loss": 0.1813,
+      "step": 1017
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.8295684680984062e-05,
+      "loss": 0.1117,
+      "step": 1018
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.8248991778549984e-05,
+      "loss": 0.1714,
+      "step": 1019
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.8202324264684544e-05,
+      "loss": 0.0874,
+      "step": 1020
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.815568231489065e-05,
+      "loss": 0.162,
+      "step": 1021
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.8109066104575023e-05,
+      "loss": 0.0952,
+      "step": 1022
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.806247580904761e-05,
+      "loss": 0.1473,
+      "step": 1023
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.8015911603520894e-05,
+      "loss": 0.1005,
+      "step": 1024
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.7969373663109234e-05,
+      "loss": 0.1447,
+      "step": 1025
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.792286216282824e-05,
+      "loss": 0.133,
+      "step": 1026
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.7876377277594053e-05,
+      "loss": 0.1142,
+      "step": 1027
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.7829919182222752e-05,
+      "loss": 0.1852,
+      "step": 1028
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.7783488051429635e-05,
+      "loss": 0.1208,
+      "step": 1029
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.7737084059828637e-05,
+      "loss": 0.1788,
+      "step": 1030
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.7690707381931583e-05,
+      "loss": 0.1228,
+      "step": 1031
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.764435819214762e-05,
+      "loss": 0.2109,
+      "step": 1032
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.7598036664782508e-05,
+      "loss": 0.1537,
+      "step": 1033
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.755174297403795e-05,
+      "loss": 0.1616,
+      "step": 1034
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.750547729401101e-05,
+      "loss": 0.0851,
+      "step": 1035
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.7459239798693364e-05,
+      "loss": 0.0597,
+      "step": 1036
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.7413030661970742e-05,
+      "loss": 0.2018,
+      "step": 1037
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.7366850057622175e-05,
+      "loss": 0.212,
+      "step": 1038
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.7320698159319433e-05,
+      "loss": 0.1599,
+      "step": 1039
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.7274575140626318e-05,
+      "loss": 0.1063,
+      "step": 1040
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.7228481174998014e-05,
+      "loss": 0.1141,
+      "step": 1041
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.7182416435780454e-05,
+      "loss": 0.1232,
+      "step": 1042
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.7136381096209664e-05,
+      "loss": 0.1827,
+      "step": 1043
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.7090375329411086e-05,
+      "loss": 0.1334,
+      "step": 1044
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.7044399308398983e-05,
+      "loss": 0.1759,
+      "step": 1045
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.699845320607571e-05,
+      "loss": 0.1718,
+      "step": 1046
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.695253719523115e-05,
+      "loss": 0.1376,
+      "step": 1047
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.690665144854198e-05,
+      "loss": 0.1144,
+      "step": 1048
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.686079613857109e-05,
+      "loss": 0.1582,
+      "step": 1049
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.681497143776689e-05,
+      "loss": 0.17,
+      "step": 1050
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.6769177518462693e-05,
+      "loss": 0.1704,
+      "step": 1051
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.672341455287605e-05,
+      "loss": 0.1584,
+      "step": 1052
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.6677682713108082e-05,
+      "loss": 0.1416,
+      "step": 1053
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.6631982171142895e-05,
+      "loss": 0.2092,
+      "step": 1054
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.658631309884684e-05,
+      "loss": 0.1524,
+      "step": 1055
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.6540675667967974e-05,
+      "loss": 0.1102,
+      "step": 1056
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.649507005013532e-05,
+      "loss": 0.1046,
+      "step": 1057
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.6449496416858284e-05,
+      "loss": 0.1082,
+      "step": 1058
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.640395493952598e-05,
+      "loss": 0.142,
+      "step": 1059
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.6358445789406584e-05,
+      "loss": 0.1374,
+      "step": 1060
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.6312969137646716e-05,
+      "loss": 0.144,
+      "step": 1061
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.6267525155270773e-05,
+      "loss": 0.1183,
+      "step": 1062
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.6222114013180283e-05,
+      "loss": 0.2672,
+      "step": 1063
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.617673588215328e-05,
+      "loss": 0.1272,
+      "step": 1064
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.6131390932843648e-05,
+      "loss": 0.188,
+      "step": 1065
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.60860793357805e-05,
+      "loss": 0.1384,
+      "step": 1066
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.6040801261367493e-05,
+      "loss": 0.1736,
+      "step": 1067
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.5995556879882246e-05,
+      "loss": 0.1125,
+      "step": 1068
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.595034636147563e-05,
+      "loss": 0.1378,
+      "step": 1069
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.5905169876171223e-05,
+      "loss": 0.1647,
+      "step": 1070
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.5860027593864545e-05,
+      "loss": 0.1102,
+      "step": 1071
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.5814919684322545e-05,
+      "loss": 0.1392,
+      "step": 1072
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.5769846317182893e-05,
+      "loss": 0.1424,
+      "step": 1073
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.572480766195333e-05,
+      "loss": 0.1029,
+      "step": 1074
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.567980388801109e-05,
+      "loss": 0.2023,
+      "step": 1075
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.56348351646022e-05,
+      "loss": 0.1216,
+      "step": 1076
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.5589901660840896e-05,
+      "loss": 0.1846,
+      "step": 1077
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.554500354570894e-05,
+      "loss": 0.1715,
+      "step": 1078
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.550014098805503e-05,
+      "loss": 0.1121,
+      "step": 1079
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.5455314156594124e-05,
+      "loss": 0.144,
+      "step": 1080
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.5410523219906843e-05,
+      "loss": 0.1107,
+      "step": 1081
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.5365768346438797e-05,
+      "loss": 0.1866,
+      "step": 1082
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.532104970449999e-05,
+      "loss": 0.1242,
+      "step": 1083
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.5276367462264153e-05,
+      "loss": 0.1247,
+      "step": 1084
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.523172178776816e-05,
+      "loss": 0.166,
+      "step": 1085
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.5187112848911323e-05,
+      "loss": 0.1867,
+      "step": 1086
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.5142540813454836e-05,
+      "loss": 0.1703,
+      "step": 1087
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.509800584902108e-05,
+      "loss": 0.1807,
+      "step": 1088
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.505350812309306e-05,
+      "loss": 0.1151,
+      "step": 1089
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.5009047803013699e-05,
+      "loss": 0.1389,
+      "step": 1090
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.4964625055985265e-05,
+      "loss": 0.1208,
+      "step": 1091
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.4920240049068748e-05,
+      "loss": 0.1556,
+      "step": 1092
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.487589294918316e-05,
+      "loss": 0.1727,
+      "step": 1093
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.4831583923104999e-05,
+      "loss": 0.1575,
+      "step": 1094
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.4787313137467546e-05,
+      "loss": 0.1114,
+      "step": 1095
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.47430807587603e-05,
+      "loss": 0.1171,
+      "step": 1096
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.4698886953328292e-05,
+      "loss": 0.0774,
+      "step": 1097
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.4654731887371523e-05,
+      "loss": 0.172,
+      "step": 1098
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.4610615726944282e-05,
+      "loss": 0.1049,
+      "step": 1099
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.4566538637954554e-05,
+      "loss": 0.1877,
+      "step": 1100
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.452250078616339e-05,
+      "loss": 0.1768,
+      "step": 1101
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.4478502337184274e-05,
+      "loss": 0.1829,
+      "step": 1102
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.443454345648252e-05,
+      "loss": 0.098,
+      "step": 1103
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.4390624309374618e-05,
+      "loss": 0.1336,
+      "step": 1104
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.4346745061027644e-05,
+      "loss": 0.1088,
+      "step": 1105
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.430290587645865e-05,
+      "loss": 0.087,
+      "step": 1106
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.4259106920533955e-05,
+      "loss": 0.1292,
+      "step": 1107
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.4215348357968669e-05,
+      "loss": 0.08,
+      "step": 1108
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.4171630353325932e-05,
+      "loss": 0.1044,
+      "step": 1109
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.4127953071016383e-05,
+      "loss": 0.1357,
+      "step": 1110
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.408431667529752e-05,
+      "loss": 0.1367,
+      "step": 1111
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.4040721330273062e-05,
+      "loss": 0.1749,
+      "step": 1112
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.3997167199892386e-05,
+      "loss": 0.2081,
+      "step": 1113
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.395365444794981e-05,
+      "loss": 0.1421,
+      "step": 1114
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.3910183238084112e-05,
+      "loss": 0.1239,
+      "step": 1115
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.3866753733777765e-05,
+      "loss": 0.1411,
+      "step": 1116
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.3823366098356487e-05,
+      "loss": 0.1506,
+      "step": 1117
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.3780020494988446e-05,
+      "loss": 0.1369,
+      "step": 1118
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.373671708668382e-05,
+      "loss": 0.1449,
+      "step": 1119
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.369345603629406e-05,
+      "loss": 0.2007,
+      "step": 1120
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.3650237506511331e-05,
+      "loss": 0.1206,
+      "step": 1121
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.3607061659867892e-05,
+      "loss": 0.0914,
+      "step": 1122
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.3563928658735484e-05,
+      "loss": 0.0825,
+      "step": 1123
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.3520838665324703e-05,
+      "loss": 0.1501,
+      "step": 1124
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.3477791841684451e-05,
+      "loss": 0.1,
+      "step": 1125
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.343478834970121e-05,
+      "loss": 0.1697,
+      "step": 1126
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.3391828351098578e-05,
+      "loss": 0.0744,
+      "step": 1127
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.3348912007436537e-05,
+      "loss": 0.0993,
+      "step": 1128
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.3306039480110905e-05,
+      "loss": 0.2353,
+      "step": 1129
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.3263210930352737e-05,
+      "loss": 0.161,
+      "step": 1130
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.3220426519227677e-05,
+      "loss": 0.1815,
+      "step": 1131
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.3177686407635417e-05,
+      "loss": 0.1748,
+      "step": 1132
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.313499075630899e-05,
+      "loss": 0.1728,
+      "step": 1133
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.3092339725814293e-05,
+      "loss": 0.2176,
+      "step": 1134
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.3049733476549352e-05,
+      "loss": 0.1322,
+      "step": 1135
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.3007172168743854e-05,
+      "loss": 0.1324,
+      "step": 1136
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.29646559624584e-05,
+      "loss": 0.1108,
+      "step": 1137
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2922185017584037e-05,
+      "loss": 0.1789,
+      "step": 1138
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2879759493841575e-05,
+      "loss": 0.1135,
+      "step": 1139
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2837379550781003e-05,
+      "loss": 0.2114,
+      "step": 1140
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2795045347780898e-05,
+      "loss": 0.0791,
+      "step": 1141
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2752757044047827e-05,
+      "loss": 0.0991,
+      "step": 1142
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2710514798615739e-05,
+      "loss": 0.1599,
+      "step": 1143
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.2668318770345369e-05,
+      "loss": 0.1356,
+      "step": 1144
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.262616911792365e-05,
+      "loss": 0.191,
+      "step": 1145
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.2584065999863102e-05,
+      "loss": 0.108,
+      "step": 1146
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.2542009574501246e-05,
+      "loss": 0.1626,
+      "step": 1147
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.2500000000000006e-05,
+      "loss": 0.1085,
+      "step": 1148
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.245803743434511e-05,
+      "loss": 0.1339,
+      "step": 1149
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.2416122035345507e-05,
+      "loss": 0.1227,
+      "step": 1150
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.2374253960632757e-05,
+      "loss": 0.1423,
+      "step": 1151
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.2332433367660442e-05,
+      "loss": 0.1217,
+      "step": 1152
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.229066041370362e-05,
+      "loss": 0.1852,
+      "step": 1153
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.2248935255858117e-05,
+      "loss": 0.1111,
+      "step": 1154
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.2207258051040099e-05,
+      "loss": 0.1887,
+      "step": 1155
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.2165628955985314e-05,
+      "loss": 0.1269,
+      "step": 1156
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.2124048127248644e-05,
+      "loss": 0.1176,
+      "step": 1157
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.2082515721203427e-05,
+      "loss": 0.1005,
+      "step": 1158
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.2041031894040903e-05,
+      "loss": 0.1029,
+      "step": 1159
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.1999596801769616e-05,
+      "loss": 0.2219,
+      "step": 1160
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1958210600214839e-05,
+      "loss": 0.0807,
+      "step": 1161
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1916873445017982e-05,
+      "loss": 0.1202,
+      "step": 1162
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1875585491636e-05,
+      "loss": 0.1552,
+      "step": 1163
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1834346895340817e-05,
+      "loss": 0.0716,
+      "step": 1164
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.179315781121874e-05,
+      "loss": 0.1303,
+      "step": 1165
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.175201839416988e-05,
+      "loss": 0.1615,
+      "step": 1166
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1710928798907556e-05,
+      "loss": 0.1759,
+      "step": 1167
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1669889179957725e-05,
+      "loss": 0.1452,
+      "step": 1168
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1628899691658399e-05,
+      "loss": 0.1575,
+      "step": 1169
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.158796048815906e-05,
+      "loss": 0.0556,
+      "step": 1170
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1547071723420083e-05,
+      "loss": 0.2003,
+      "step": 1171
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1506233551212186e-05,
+      "loss": 0.182,
+      "step": 1172
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1465446125115758e-05,
+      "loss": 0.1683,
+      "step": 1173
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1424709598520425e-05,
+      "loss": 0.0866,
+      "step": 1174
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1384024124624324e-05,
+      "loss": 0.1715,
+      "step": 1175
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1343389856433658e-05,
+      "loss": 0.1375,
+      "step": 1176
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1302806946762004e-05,
+      "loss": 0.0937,
+      "step": 1177
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.126227554822985e-05,
+      "loss": 0.1105,
+      "step": 1178
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1221795813263925e-05,
+      "loss": 0.1319,
+      "step": 1179
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1181367894096684e-05,
+      "loss": 0.2046,
+      "step": 1180
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1140991942765713e-05,
+      "loss": 0.1987,
+      "step": 1181
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1100668111113166e-05,
+      "loss": 0.127,
+      "step": 1182
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1060396550785182e-05,
+      "loss": 0.136,
+      "step": 1183
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1020177413231334e-05,
+      "loss": 0.1428,
+      "step": 1184
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.0980010849704036e-05,
+      "loss": 0.1863,
+      "step": 1185
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.0939897011258001e-05,
+      "loss": 0.1334,
+      "step": 1186
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.0899836048749645e-05,
+      "loss": 0.0984,
+      "step": 1187
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.085982811283654e-05,
+      "loss": 0.1384,
+      "step": 1188
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.081987335397684e-05,
+      "loss": 0.0985,
+      "step": 1189
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.0779971922428711e-05,
+      "loss": 0.1221,
+      "step": 1190
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.074012396824978e-05,
+      "loss": 0.1366,
+      "step": 1191
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.0700329641296541e-05,
+      "loss": 0.1501,
+      "step": 1192
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.0660589091223855e-05,
+      "loss": 0.1357,
+      "step": 1193
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.0620902467484278e-05,
+      "loss": 0.1224,
+      "step": 1194
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.0581269919327643e-05,
+      "loss": 0.1812,
+      "step": 1195
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.0541691595800337e-05,
+      "loss": 0.1064,
+      "step": 1196
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.0502167645744895e-05,
+      "loss": 0.1371,
+      "step": 1197
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.0462698217799332e-05,
+      "loss": 0.1241,
+      "step": 1198
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.0423283460396633e-05,
+      "loss": 0.1686,
+      "step": 1199
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.0383923521764174e-05,
+      "loss": 0.1061,
+      "step": 1200
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.0344618549923185e-05,
+      "loss": 0.1587,
+      "step": 1201
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.0305368692688174e-05,
+      "loss": 0.1595,
+      "step": 1202
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.026617409766638e-05,
+      "loss": 0.0922,
+      "step": 1203
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.0227034912257222e-05,
+      "loss": 0.1475,
+      "step": 1204
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.0187951283651736e-05,
+      "loss": 0.0772,
+      "step": 1205
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.0148923358832022e-05,
+      "loss": 0.1883,
+      "step": 1206
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.01099512845707e-05,
+      "loss": 0.1626,
+      "step": 1207
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.0071035207430352e-05,
+      "loss": 0.1475,
+      "step": 1208
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.0032175273762965e-05,
+      "loss": 0.167,
+      "step": 1209
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 9.993371629709391e-06,
+      "loss": 0.1389,
+      "step": 1210
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 9.954624421198792e-06,
+      "loss": 0.1156,
+      "step": 1211
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 9.91593379394811e-06,
+      "loss": 0.0989,
+      "step": 1212
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 9.877299893461456e-06,
+      "loss": 0.1161,
+      "step": 1213
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 9.838722865029673e-06,
+      "loss": 0.1509,
+      "step": 1214
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 9.800202853729651e-06,
+      "loss": 0.1284,
+      "step": 1215
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 9.761740004423927e-06,
+      "loss": 0.1241,
+      "step": 1216
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 9.723334461760006e-06,
+      "loss": 0.0987,
+      "step": 1217
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 9.68498637016993e-06,
+      "loss": 0.1256,
+      "step": 1218
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 9.646695873869655e-06,
+      "loss": 0.1226,
+      "step": 1219
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 9.608463116858542e-06,
+      "loss": 0.1326,
+      "step": 1220
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 9.570288242918823e-06,
+      "loss": 0.1136,
+      "step": 1221
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 9.532171395615036e-06,
+      "loss": 0.0855,
+      "step": 1222
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 9.494112718293501e-06,
+      "loss": 0.1491,
+      "step": 1223
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 9.456112354081782e-06,
+      "loss": 0.1145,
+      "step": 1224
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 9.418170445888139e-06,
+      "loss": 0.1138,
+      "step": 1225
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 9.380287136401e-06,
+      "loss": 0.1297,
+      "step": 1226
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 9.342462568088416e-06,
+      "loss": 0.1196,
+      "step": 1227
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 9.304696883197542e-06,
+      "loss": 0.0972,
+      "step": 1228
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 9.266990223754069e-06,
+      "loss": 0.1696,
+      "step": 1229
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 9.22934273156172e-06,
+      "loss": 0.1833,
+      "step": 1230
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 9.19175454820173e-06,
+      "loss": 0.1948,
+      "step": 1231
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 9.154225815032242e-06,
+      "loss": 0.1516,
+      "step": 1232
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 9.116756673187878e-06,
+      "loss": 0.1593,
+      "step": 1233
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 9.079347263579103e-06,
+      "loss": 0.1297,
+      "step": 1234
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 9.0419977268918e-06,
+      "loss": 0.5721,
+      "step": 1235
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 9.00470820358663e-06,
+      "loss": 0.1469,
+      "step": 1236
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 8.967478833898612e-06,
+      "loss": 0.1723,
+      "step": 1237
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 8.930309757836517e-06,
+      "loss": 0.1576,
+      "step": 1238
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 8.893201115182375e-06,
+      "loss": 0.3551,
+      "step": 1239
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 8.856153045490948e-06,
+      "loss": 0.1471,
+      "step": 1240
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 8.819165688089193e-06,
+      "loss": 0.1418,
+      "step": 1241
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 8.78223918207575e-06,
+      "loss": 0.1408,
+      "step": 1242
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 8.745373666320417e-06,
+      "loss": 0.125,
+      "step": 1243
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 8.708569279463622e-06,
+      "loss": 0.1356,
+      "step": 1244
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 8.671826159915907e-06,
+      "loss": 0.0884,
+      "step": 1245
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 8.635144445857406e-06,
+      "loss": 0.1073,
+      "step": 1246
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 8.598524275237322e-06,
+      "loss": 0.13,
+      "step": 1247
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 8.561965785773413e-06,
+      "loss": 0.1253,
+      "step": 1248
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 8.525469114951473e-06,
+      "loss": 0.1184,
+      "step": 1249
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 8.489034400024812e-06,
+      "loss": 0.1617,
+      "step": 1250
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 8.452661778013738e-06,
+      "loss": 0.109,
+      "step": 1251
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 8.41635138570507e-06,
+      "loss": 0.1616,
+      "step": 1252
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 8.380103359651553e-06,
+      "loss": 0.1381,
+      "step": 1253
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 8.343917836171447e-06,
+      "loss": 0.0848,
+      "step": 1254
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 8.3077949513479e-06,
+      "loss": 0.1798,
+      "step": 1255
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 8.271734841028553e-06,
+      "loss": 0.158,
+      "step": 1256
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 8.235737640824908e-06,
+      "loss": 0.1192,
+      "step": 1257
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 8.19980348611194e-06,
+      "loss": 0.1859,
+      "step": 1258
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 8.163932512027492e-06,
+      "loss": 0.115,
+      "step": 1259
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 8.128124853471814e-06,
+      "loss": 0.1743,
+      "step": 1260
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 8.09238064510704e-06,
+      "loss": 0.1313,
+      "step": 1261
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 8.056700021356694e-06,
+      "loss": 0.1416,
+      "step": 1262
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 8.021083116405173e-06,
+      "loss": 0.1047,
+      "step": 1263
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 7.98553006419724e-06,
+      "loss": 0.095,
+      "step": 1264
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 7.950040998437542e-06,
+      "loss": 0.093,
+      "step": 1265
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 7.914616052590071e-06,
+      "loss": 0.1078,
+      "step": 1266
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 7.879255359877705e-06,
+      "loss": 0.1348,
+      "step": 1267
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 7.843959053281663e-06,
+      "loss": 0.1035,
+      "step": 1268
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 7.808727265541047e-06,
+      "loss": 0.1432,
+      "step": 1269
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 7.7735601291523e-06,
+      "loss": 0.1406,
+      "step": 1270
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 7.738457776368766e-06,
+      "loss": 0.0883,
+      "step": 1271
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 7.703420339200101e-06,
+      "loss": 0.1182,
+      "step": 1272
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 7.6684479494119e-06,
+      "loss": 0.1518,
+      "step": 1273
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 7.633540738525066e-06,
+      "loss": 0.1452,
+      "step": 1274
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 7.598698837815449e-06,
+      "loss": 0.1197,
+      "step": 1275
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 7.563922378313218e-06,
+      "loss": 0.1564,
+      "step": 1276
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 7.529211490802498e-06,
+      "loss": 0.0923,
+      "step": 1277
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 7.494566305820788e-06,
+      "loss": 0.1688,
+      "step": 1278
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 7.459986953658493e-06,
+      "loss": 0.2256,
+      "step": 1279
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 7.4254735643584564e-06,
+      "loss": 0.0959,
+      "step": 1280
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 7.391026267715448e-06,
+      "loss": 0.1094,
+      "step": 1281
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 7.3566451932756744e-06,
+      "loss": 0.1359,
+      "step": 1282
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 7.3223304703363135e-06,
+      "loss": 0.1068,
+      "step": 1283
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 7.2880822279450064e-06,
+      "loss": 0.1096,
+      "step": 1284
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 7.2539005948993825e-06,
+      "loss": 0.1558,
+      "step": 1285
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 7.219785699746573e-06,
+      "loss": 0.1908,
+      "step": 1286
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 7.185737670782727e-06,
+      "loss": 0.1039,
+      "step": 1287
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 7.1517566360525284e-06,
+      "loss": 0.1257,
+      "step": 1288
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 7.117842723348717e-06,
+      "loss": 0.2259,
+      "step": 1289
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 7.083996060211607e-06,
+      "loss": 0.1608,
+      "step": 1290
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 7.050216773928594e-06,
+      "loss": 0.1423,
+      "step": 1291
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 7.016504991533726e-06,
+      "loss": 0.0972,
+      "step": 1292
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 6.982860839807135e-06,
+      "loss": 0.1529,
+      "step": 1293
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 6.949284445274667e-06,
+      "loss": 0.1292,
+      "step": 1294
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 6.9157759342072995e-06,
+      "loss": 0.1856,
+      "step": 1295
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 6.882335432620779e-06,
+      "loss": 0.1457,
+      "step": 1296
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 6.848963066275027e-06,
+      "loss": 0.1312,
+      "step": 1297
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 6.815658960673782e-06,
+      "loss": 0.1335,
+      "step": 1298
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 6.782423241064043e-06,
+      "loss": 0.1016,
+      "step": 1299
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 6.7492560324356355e-06,
+      "loss": 0.1059,
+      "step": 1300
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 6.716157459520739e-06,
+      "loss": 0.1383,
+      "step": 1301
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 6.683127646793411e-06,
+      "loss": 0.1066,
+      "step": 1302
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 6.65016671846912e-06,
+      "loss": 0.1186,
+      "step": 1303
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 6.617274798504286e-06,
+      "loss": 0.2051,
+      "step": 1304
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 6.584452010595807e-06,
+      "loss": 0.1092,
+      "step": 1305
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 6.55169847818059e-06,
+      "loss": 0.1126,
+      "step": 1306
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 6.519014324435102e-06,
+      "loss": 0.0893,
+      "step": 1307
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 6.48639967227489e-06,
+      "loss": 0.168,
+      "step": 1308
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 6.453854644354126e-06,
+      "loss": 0.1004,
+      "step": 1309
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 6.421379363065142e-06,
+      "loss": 0.1057,
+      "step": 1310
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 6.388973950537999e-06,
+      "loss": 0.1299,
+      "step": 1311
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 6.356638528639955e-06,
+      "loss": 0.0923,
+      "step": 1312
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 6.324373218975105e-06,
+      "loss": 0.1468,
+      "step": 1313
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 6.292178142883826e-06,
+      "loss": 0.1033,
+      "step": 1314
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 6.26005342144241e-06,
+      "loss": 0.1288,
+      "step": 1315
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 6.22799917546252e-06,
+      "loss": 0.1173,
+      "step": 1316
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 6.196015525490825e-06,
+      "loss": 0.1114,
+      "step": 1317
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 6.1641025918084825e-06,
+      "loss": 0.0979,
+      "step": 1318
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 6.1322604944307e-06,
+      "loss": 0.1537,
+      "step": 1319
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 6.100489353106304e-06,
+      "loss": 0.127,
+      "step": 1320
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 6.06878928731727e-06,
+      "loss": 0.0724,
+      "step": 1321
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 6.037160416278278e-06,
+      "loss": 0.1118,
+      "step": 1322
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 6.0056028589362684e-06,
+      "loss": 0.152,
+      "step": 1323
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5.97411673396999e-06,
+      "loss": 0.1198,
+      "step": 1324
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5.942702159789554e-06,
+      "loss": 0.1027,
+      "step": 1325
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5.9113592545359945e-06,
+      "loss": 0.1532,
+      "step": 1326
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5.880088136080814e-06,
+      "loss": 0.1071,
+      "step": 1327
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5.848888922025553e-06,
+      "loss": 0.1411,
+      "step": 1328
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5.817761729701332e-06,
+      "loss": 0.1667,
+      "step": 1329
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.786706676168424e-06,
+      "loss": 0.1018,
+      "step": 1330
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.755723878215802e-06,
+      "loss": 0.1073,
+      "step": 1331
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.724813452360736e-06,
+      "loss": 0.17,
+      "step": 1332
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.693975514848271e-06,
+      "loss": 0.2136,
+      "step": 1333
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.663210181650905e-06,
+      "loss": 0.1077,
+      "step": 1334
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.6325175684680374e-06,
+      "loss": 0.1464,
+      "step": 1335
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.601897790725643e-06,
+      "loss": 0.2062,
+      "step": 1336
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.571350963575728e-06,
+      "loss": 0.1539,
+      "step": 1337
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5.5408772018959995e-06,
+      "loss": 0.1115,
+      "step": 1338
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5.510476620289362e-06,
+      "loss": 0.1321,
+      "step": 1339
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5.48014933308352e-06,
+      "loss": 0.1233,
+      "step": 1340
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5.449895454330531e-06,
+      "loss": 0.1512,
+      "step": 1341
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5.4197150978063965e-06,
+      "loss": 0.1188,
+      "step": 1342
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5.389608377010608e-06,
+      "loss": 0.1216,
+      "step": 1343
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5.359575405165748e-06,
+      "loss": 0.1131,
+      "step": 1344
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5.329616295217046e-06,
+      "loss": 0.1213,
+      "step": 1345
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5.299731159831953e-06,
+      "loss": 0.1715,
+      "step": 1346
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.269920111399732e-06,
+      "loss": 0.1077,
+      "step": 1347
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.240183262031021e-06,
+      "loss": 0.1157,
+      "step": 1348
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.210520723557419e-06,
+      "loss": 0.1144,
+      "step": 1349
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.180932607531056e-06,
+      "loss": 0.128,
+      "step": 1350
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.151419025224208e-06,
+      "loss": 0.0909,
+      "step": 1351
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.121980087628803e-06,
+      "loss": 0.182,
+      "step": 1352
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.092615905456111e-06,
+      "loss": 0.1613,
+      "step": 1353
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.063326589136211e-06,
+      "loss": 0.129,
+      "step": 1354
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.034112248817685e-06,
+      "loss": 0.1345,
+      "step": 1355
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.004972994367102e-06,
+      "loss": 0.1175,
+      "step": 1356
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.975908935368701e-06,
+      "loss": 0.1396,
+      "step": 1357
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.946920181123904e-06,
+      "loss": 0.217,
+      "step": 1358
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.918006840650949e-06,
+      "loss": 0.0865,
+      "step": 1359
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.88916902268445e-06,
+      "loss": 0.066,
+      "step": 1360
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.860406835675016e-06,
+      "loss": 0.1051,
+      "step": 1361
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.831720387788827e-06,
+      "loss": 0.1201,
+      "step": 1362
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.803109786907223e-06,
+      "loss": 0.1527,
+      "step": 1363
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.7745751406263165e-06,
+      "loss": 0.1342,
+      "step": 1364
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.746116556256569e-06,
+      "loss": 0.0959,
+      "step": 1365
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.7177341408224e-06,
+      "loss": 0.1587,
+      "step": 1366
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.689428001061774e-06,
+      "loss": 0.1189,
+      "step": 1367
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.661198243425813e-06,
+      "loss": 0.1301,
+      "step": 1368
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.633044974078385e-06,
+      "loss": 0.128,
+      "step": 1369
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.604968298895703e-06,
+      "loss": 0.114,
+      "step": 1370
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.57696832346593e-06,
+      "loss": 0.1097,
+      "step": 1371
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.549045153088813e-06,
+      "loss": 0.1107,
+      "step": 1372
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.521198892775203e-06,
+      "loss": 0.098,
+      "step": 1373
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.493429647246774e-06,
+      "loss": 0.1187,
+      "step": 1374
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.465737520935517e-06,
+      "loss": 0.0918,
+      "step": 1375
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.438122617983443e-06,
+      "loss": 0.1041,
+      "step": 1376
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.410585042242124e-06,
+      "loss": 0.1789,
+      "step": 1377
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.383124897272331e-06,
+      "loss": 0.1677,
+      "step": 1378
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.355742286343648e-06,
+      "loss": 0.1484,
+      "step": 1379
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.328437312434067e-06,
+      "loss": 0.1844,
+      "step": 1380
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.301210078229612e-06,
+      "loss": 0.1237,
+      "step": 1381
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.274060686123959e-06,
+      "loss": 0.1276,
+      "step": 1382
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.24698923821803e-06,
+      "loss": 0.1418,
+      "step": 1383
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.219995836319632e-06,
+      "loss": 0.1709,
+      "step": 1384
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.19308058194306e-06,
+      "loss": 0.167,
+      "step": 1385
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.166243576308712e-06,
+      "loss": 0.09,
+      "step": 1386
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.1394849203427284e-06,
+      "loss": 0.1081,
+      "step": 1387
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.112804714676594e-06,
+      "loss": 0.1272,
+      "step": 1388
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.086203059646756e-06,
+      "loss": 0.1187,
+      "step": 1389
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.059680055294266e-06,
+      "loss": 0.1168,
+      "step": 1390
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.0332358013644016e-06,
+      "loss": 0.0824,
+      "step": 1391
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.006870397306256e-06,
+      "loss": 0.1152,
+      "step": 1392
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.980583942272425e-06,
+      "loss": 0.1229,
+      "step": 1393
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.954376535118562e-06,
+      "loss": 0.1299,
+      "step": 1394
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.92824827440309e-06,
+      "loss": 0.1153,
+      "step": 1395
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.9021992583867325e-06,
+      "loss": 0.0906,
+      "step": 1396
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.876229585032245e-06,
+      "loss": 0.1159,
+      "step": 1397
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.850339352003973e-06,
+      "loss": 0.1952,
+      "step": 1398
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.824528656667515e-06,
+      "loss": 0.1423,
+      "step": 1399
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.798797596089351e-06,
+      "loss": 0.1878,
+      "step": 1400
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.7731462670364785e-06,
+      "loss": 0.1646,
+      "step": 1401
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.7475747659760502e-06,
+      "loss": 0.1644,
+      "step": 1402
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.7220831890750067e-06,
+      "loss": 0.0783,
+      "step": 1403
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.696671632199719e-06,
+      "loss": 0.1291,
+      "step": 1404
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.6713401909156204e-06,
+      "loss": 0.0625,
+      "step": 1405
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.6460889604868626e-06,
+      "loss": 0.1561,
+      "step": 1406
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.6209180358759394e-06,
+      "loss": 0.1465,
+      "step": 1407
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.595827511743341e-06,
+      "loss": 0.0885,
+      "step": 1408
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.5708174824471947e-06,
+      "loss": 0.0843,
+      "step": 1409
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.5458880420429135e-06,
+      "loss": 0.1595,
+      "step": 1410
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.521039284282829e-06,
+      "loss": 0.1589,
+      "step": 1411
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.4962713026158694e-06,
+      "loss": 0.161,
+      "step": 1412
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.471584190187155e-06,
+      "loss": 0.1077,
+      "step": 1413
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.4469780398377256e-06,
+      "loss": 0.1186,
+      "step": 1414
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.4224529441040904e-06,
+      "loss": 0.1207,
+      "step": 1415
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.398008995217988e-06,
+      "loss": 0.1496,
+      "step": 1416
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.373646285105958e-06,
+      "loss": 0.1521,
+      "step": 1417
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.3493649053890326e-06,
+      "loss": 0.1364,
+      "step": 1418
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.3251649473823837e-06,
+      "loss": 0.117,
+      "step": 1419
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.3010465020949818e-06,
+      "loss": 0.1636,
+      "step": 1420
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.2770096602292465e-06,
+      "loss": 0.1205,
+      "step": 1421
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.2530545121807145e-06,
+      "loss": 0.1571,
+      "step": 1422
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.2291811480376995e-06,
+      "loss": 0.1278,
+      "step": 1423
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.205389657580943e-06,
+      "loss": 0.0974,
+      "step": 1424
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.1816801302832848e-06,
+      "loss": 0.1071,
+      "step": 1425
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.158052655309332e-06,
+      "loss": 0.1182,
+      "step": 1426
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.1345073215151066e-06,
+      "loss": 0.1257,
+      "step": 1427
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.111044217447731e-06,
+      "loss": 0.1307,
+      "step": 1428
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.087663431345081e-06,
+      "loss": 0.1021,
+      "step": 1429
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.0643650511354484e-06,
+      "loss": 0.1704,
+      "step": 1430
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.041149164437254e-06,
+      "loss": 0.1002,
+      "step": 1431
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.0180158585586397e-06,
+      "loss": 0.1544,
+      "step": 1432
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.9949652204972254e-06,
+      "loss": 0.1201,
+      "step": 1433
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.9719973369397023e-06,
+      "loss": 0.0933,
+      "step": 1434
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.949112294261591e-06,
+      "loss": 0.1031,
+      "step": 1435
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.9263101785268254e-06,
+      "loss": 0.1115,
+      "step": 1436
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.9035910754875136e-06,
+      "loss": 0.0988,
+      "step": 1437
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.8809550705835548e-06,
+      "loss": 0.1198,
+      "step": 1438
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.858402248942349e-06,
+      "loss": 0.1115,
+      "step": 1439
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.8359326953784737e-06,
+      "loss": 0.1025,
+      "step": 1440
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.813546494393329e-06,
+      "loss": 0.0915,
+      "step": 1441
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.7912437301749026e-06,
+      "loss": 0.1202,
+      "step": 1442
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.769024486597349e-06,
+      "loss": 0.1137,
+      "step": 1443
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.7468888472207603e-06,
+      "loss": 0.1175,
+      "step": 1444
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.7248368952908053e-06,
+      "loss": 0.0996,
+      "step": 1445
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.7028687137384267e-06,
+      "loss": 0.105,
+      "step": 1446
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.6809843851795357e-06,
+      "loss": 0.4433,
+      "step": 1447
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.659183991914696e-06,
+      "loss": 0.076,
+      "step": 1448
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.637467615928807e-06,
+      "loss": 0.0989,
+      "step": 1449
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.6158353388908293e-06,
+      "loss": 0.1583,
+      "step": 1450
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.5942872421534146e-06,
+      "loss": 0.093,
+      "step": 1451
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.57282340675267e-06,
+      "loss": 0.1294,
+      "step": 1452
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.5514439134077945e-06,
+      "loss": 0.1141,
+      "step": 1453
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.5301488425208296e-06,
+      "loss": 0.1023,
+      "step": 1454
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.5089382741762925e-06,
+      "loss": 0.11,
+      "step": 1455
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.487812288140945e-06,
+      "loss": 0.2201,
+      "step": 1456
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.4667709638634434e-06,
+      "loss": 0.0637,
+      "step": 1457
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.445814380474057e-06,
+      "loss": 0.12,
+      "step": 1458
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.4249426167843757e-06,
+      "loss": 0.0818,
+      "step": 1459
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.4041557512869878e-06,
+      "loss": 0.1241,
+      "step": 1460
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.3834538621552356e-06,
+      "loss": 0.1573,
+      "step": 1461
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.3628370272428564e-06,
+      "loss": 0.0897,
+      "step": 1462
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.3423053240837515e-06,
+      "loss": 0.1307,
+      "step": 1463
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.3218588298916544e-06,
+      "loss": 0.0943,
+      "step": 1464
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.3014976215598503e-06,
+      "loss": 0.11,
+      "step": 1465
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.281221775660894e-06,
+      "loss": 0.1354,
+      "step": 1466
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.2610313684463177e-06,
+      "loss": 0.1439,
+      "step": 1467
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.2409264758463363e-06,
+      "loss": 0.1397,
+      "step": 1468
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.2209071734695746e-06,
+      "loss": 0.1396,
+      "step": 1469
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.2009735366027795e-06,
+      "loss": 0.1308,
+      "step": 1470
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.181125640210538e-06,
+      "loss": 0.1363,
+      "step": 1471
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.1613635589349756e-06,
+      "loss": 0.1239,
+      "step": 1472
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.1416873670955184e-06,
+      "loss": 0.1017,
+      "step": 1473
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.1220971386885556e-06,
+      "loss": 0.0957,
+      "step": 1474
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.1025929473872274e-06,
+      "loss": 0.1403,
+      "step": 1475
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.0831748665410765e-06,
+      "loss": 0.1079,
+      "step": 1476
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.063842969175847e-06,
+      "loss": 0.1514,
+      "step": 1477
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.044597327993153e-06,
+      "loss": 0.1554,
+      "step": 1478
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.0254380153702123e-06,
+      "loss": 0.1544,
+      "step": 1479
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.0063651033596143e-06,
+      "loss": 0.1627,
+      "step": 1480
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.9873786636889906e-06,
+      "loss": 0.1347,
+      "step": 1481
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.968478767760812e-06,
+      "loss": 0.0995,
+      "step": 1482
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.9496654866520414e-06,
+      "loss": 0.0805,
+      "step": 1483
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.9309388911139426e-06,
+      "loss": 0.1176,
+      "step": 1484
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.912299051571764e-06,
+      "loss": 0.1118,
+      "step": 1485
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.893746038124497e-06,
+      "loss": 0.1289,
+      "step": 1486
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.8752799205445982e-06,
+      "loss": 0.1698,
+      "step": 1487
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.8569007682777417e-06,
+      "loss": 0.1075,
+      "step": 1488
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.8386086504425416e-06,
+      "loss": 0.1274,
+      "step": 1489
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.8204036358303173e-06,
+      "loss": 0.1314,
+      "step": 1490
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.8022857929047926e-06,
+      "loss": 0.1374,
+      "step": 1491
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.784255189801895e-06,
+      "loss": 0.1409,
+      "step": 1492
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.7663118943294366e-06,
+      "loss": 0.1399,
+      "step": 1493
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.7484559739669226e-06,
+      "loss": 0.1213,
+      "step": 1494
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.7306874958652408e-06,
+      "loss": 0.1044,
+      "step": 1495
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.713006526846439e-06,
+      "loss": 0.1098,
+      "step": 1496
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.6954131334034922e-06,
+      "loss": 0.0977,
+      "step": 1497
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.6779073816999864e-06,
+      "loss": 0.1058,
+      "step": 1498
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.6604893375699594e-06,
+      "loss": 0.1495,
+      "step": 1499
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.643159066517566e-06,
+      "loss": 0.1001,
+      "step": 1500
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.6259166337169063e-06,
+      "loss": 0.1207,
+      "step": 1501
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.6087621040117157e-06,
+      "loss": 0.1032,
+      "step": 1502
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.5916955419151725e-06,
+      "loss": 0.1352,
+      "step": 1503
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.5747170116096332e-06,
+      "loss": 0.131,
+      "step": 1504
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.5578265769463806e-06,
+      "loss": 0.1237,
+      "step": 1505
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.541024301445404e-06,
+      "loss": 0.1696,
+      "step": 1506
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.524310248295152e-06,
+      "loss": 0.1359,
+      "step": 1507
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.5076844803522922e-06,
+      "loss": 0.0959,
+      "step": 1508
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.4911470601414789e-06,
+      "loss": 0.1185,
+      "step": 1509
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.4746980498551112e-06,
+      "loss": 0.1085,
+      "step": 1510
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.4583375113531195e-06,
+      "loss": 0.1402,
+      "step": 1511
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.4420655061626932e-06,
+      "loss": 0.1461,
+      "step": 1512
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.4258820954781039e-06,
+      "loss": 0.1584,
+      "step": 1513
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.4097873401604122e-06,
+      "loss": 0.0977,
+      "step": 1514
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.3937813007373013e-06,
+      "loss": 0.1242,
+      "step": 1515
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.3778640374027985e-06,
+      "loss": 0.0998,
+      "step": 1516
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.362035610017079e-06,
+      "loss": 0.1346,
+      "step": 1517
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.3462960781062434e-06,
+      "loss": 0.0888,
+      "step": 1518
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.3306455008620516e-06,
+      "loss": 0.0745,
+      "step": 1519
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.31508393714177e-06,
+      "loss": 0.1318,
+      "step": 1520
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.2996114454678742e-06,
+      "loss": 0.0952,
+      "step": 1521
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.2842280840278997e-06,
+      "loss": 0.0801,
+      "step": 1522
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.2689339106741527e-06,
+      "loss": 0.2555,
+      "step": 1523
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.2537289829235633e-06,
+      "loss": 0.1345,
+      "step": 1524
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.2386133579574189e-06,
+      "loss": 0.1477,
+      "step": 1525
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.2235870926211619e-06,
+      "loss": 0.0972,
+      "step": 1526
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.2086502434241865e-06,
+      "loss": 0.1834,
+      "step": 1527
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.1938028665396173e-06,
+      "loss": 0.0931,
+      "step": 1528
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.1790450178040951e-06,
+      "loss": 0.1288,
+      "step": 1529
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.1643767527175857e-06,
+      "loss": 0.093,
+      "step": 1530
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.1497981264431329e-06,
+      "loss": 0.1419,
+      "step": 1531
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.1353091938067023e-06,
+      "loss": 0.0902,
+      "step": 1532
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.1209100092969244e-06,
+      "loss": 0.0956,
+      "step": 1533
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.1066006270649272e-06,
+      "loss": 0.1606,
+      "step": 1534
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.0923811009241142e-06,
+      "loss": 0.1574,
+      "step": 1535
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.0782514843499653e-06,
+      "loss": 0.0668,
+      "step": 1536
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.0642118304798442e-06,
+      "loss": 0.0941,
+      "step": 1537
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.0502621921127776e-06,
+      "loss": 0.1511,
+      "step": 1538
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.036402621709287e-06,
+      "loss": 0.1605,
+      "step": 1539
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.0226331713911546e-06,
+      "loss": 0.1818,
+      "step": 1540
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.0089538929412724e-06,
+      "loss": 0.097,
+      "step": 1541
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.95364837803392e-07,
+      "loss": 0.1255,
+      "step": 1542
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.818660570819872e-07,
+      "loss": 0.1328,
+      "step": 1543
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.684576015420278e-07,
+      "loss": 0.0945,
+      "step": 1544
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.551395216087944e-07,
+      "loss": 0.1167,
+      "step": 1545
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.419118673676924e-07,
+      "loss": 0.0919,
+      "step": 1546
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.287746885640603e-07,
+      "loss": 0.1006,
+      "step": 1547
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.157280346029918e-07,
+      "loss": 0.143,
+      "step": 1548
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.027719545491392e-07,
+      "loss": 0.1601,
+      "step": 1549
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 8.899064971265276e-07,
+      "loss": 0.1562,
+      "step": 1550
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 8.771317107183874e-07,
+      "loss": 0.1446,
+      "step": 1551
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 8.64447643366953e-07,
+      "loss": 0.1544,
+      "step": 1552
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 8.51854342773295e-07,
+      "loss": 0.1078,
+      "step": 1553
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 8.393518562971353e-07,
+      "loss": 0.1786,
+      "step": 1554
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 8.269402309566743e-07,
+      "loss": 0.112,
+      "step": 1555
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 8.146195134284052e-07,
+      "loss": 0.149,
+      "step": 1556
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 8.023897500469391e-07,
+      "loss": 0.1732,
+      "step": 1557
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 7.902509868048552e-07,
+      "loss": 0.107,
+      "step": 1558
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 7.782032693524816e-07,
+      "loss": 0.1065,
+      "step": 1559
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 7.662466429977699e-07,
+      "loss": 0.126,
+      "step": 1560
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 7.543811527060906e-07,
+      "loss": 0.1592,
+      "step": 1561
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 7.426068431000882e-07,
+      "loss": 0.0956,
+      "step": 1562
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 7.309237584595008e-07,
+      "loss": 0.1711,
+      "step": 1563
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 7.19331942720991e-07,
+      "loss": 0.2258,
+      "step": 1564
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 7.078314394779961e-07,
+      "loss": 0.0735,
+      "step": 1565
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.964222919805391e-07,
+      "loss": 0.1208,
+      "step": 1566
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.851045431350927e-07,
+      "loss": 0.1938,
+      "step": 1567
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.738782355044049e-07,
+      "loss": 0.1112,
+      "step": 1568
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.627434113073344e-07,
+      "loss": 0.1388,
+      "step": 1569
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.517001124186989e-07,
+      "loss": 0.1496,
+      "step": 1570
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.407483803691216e-07,
+      "loss": 0.1151,
+      "step": 1571
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.298882563448599e-07,
+      "loss": 0.2096,
+      "step": 1572
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.191197811876742e-07,
+      "loss": 0.149,
+      "step": 1573
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.084429953946452e-07,
+      "loss": 0.0902,
+      "step": 1574
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5.978579391180461e-07,
+      "loss": 0.1493,
+      "step": 1575
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5.87364652165176e-07,
+      "loss": 0.1295,
+      "step": 1576
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5.769631739982267e-07,
+      "loss": 0.1058,
+      "step": 1577
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5.666535437341108e-07,
+      "loss": 0.1026,
+      "step": 1578
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5.564358001443421e-07,
+      "loss": 0.1621,
+      "step": 1579
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5.463099816548579e-07,
+      "loss": 0.0721,
+      "step": 1580
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5.362761263459137e-07,
+      "loss": 0.1551,
+      "step": 1581
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5.263342719518921e-07,
+      "loss": 0.1553,
+      "step": 1582
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5.164844558612131e-07,
+      "loss": 0.1179,
+      "step": 1583
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5.067267151161515e-07,
+      "loss": 0.1173,
+      "step": 1584
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 4.970610864127173e-07,
+      "loss": 0.0836,
+      "step": 1585
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 4.874876061005173e-07,
+      "loss": 0.1214,
+      "step": 1586
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 4.780063101826132e-07,
+      "loss": 0.0901,
+      "step": 1587
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 4.6861723431538276e-07,
+      "loss": 0.0931,
+      "step": 1588
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 4.5932041380840065e-07,
+      "loss": 0.1349,
+      "step": 1589
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 4.5011588362429134e-07,
+      "loss": 0.0852,
+      "step": 1590
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 4.410036783786015e-07,
+      "loss": 0.1381,
+      "step": 1591
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 4.319838323396691e-07,
+      "loss": 0.1044,
+      "step": 1592
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 4.23056379428502e-07,
+      "loss": 0.1274,
+      "step": 1593
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 4.142213532186384e-07,
+      "loss": 0.1163,
+      "step": 1594
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 4.05478786936031e-07,
+      "loss": 0.1706,
+      "step": 1595
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.9682871345891883e-07,
+      "loss": 0.0837,
+      "step": 1596
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.88271165317694e-07,
+      "loss": 0.1054,
+      "step": 1597
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.7980617469479953e-07,
+      "loss": 0.0719,
+      "step": 1598
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.714337734245871e-07,
+      "loss": 0.1413,
+      "step": 1599
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.6315399299321484e-07,
+      "loss": 0.08,
+      "step": 1600
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.5496686453850846e-07,
+      "loss": 0.2635,
+      "step": 1601
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.468724188498751e-07,
+      "loss": 0.1491,
+      "step": 1602
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.3887068636815346e-07,
+      "loss": 0.1173,
+      "step": 1603
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.309616971855195e-07,
+      "loss": 0.1246,
+      "step": 1604
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.2314548104537545e-07,
+      "loss": 0.124,
+      "step": 1605
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.1542206734221924e-07,
+      "loss": 0.1376,
+      "step": 1606
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.077914851215585e-07,
+      "loss": 0.1324,
+      "step": 1607
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.002537630797747e-07,
+      "loss": 0.1088,
+      "step": 1608
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 2.9280892956404513e-07,
+      "loss": 0.135,
+      "step": 1609
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 2.8545701257221e-07,
+      "loss": 0.1281,
+      "step": 1610
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 2.781980397526779e-07,
+      "loss": 0.1454,
+      "step": 1611
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 2.71032038404323e-07,
+      "loss": 0.144,
+      "step": 1612
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 2.6395903547638825e-07,
+      "loss": 0.1146,
+      "step": 1613
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 2.569790575683656e-07,
+      "loss": 0.1136,
+      "step": 1614
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 2.5009213092991034e-07,
+      "loss": 0.1394,
+      "step": 1615
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 2.4329828146074095e-07,
+      "loss": 0.0707,
+      "step": 1616
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 2.365975347105448e-07,
+      "loss": 0.1057,
+      "step": 1617
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 2.2998991587886709e-07,
+      "loss": 0.1245,
+      "step": 1618
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 2.234754498150332e-07,
+      "loss": 0.0786,
+      "step": 1619
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 2.170541610180432e-07,
+      "loss": 0.0907,
+      "step": 1620
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 2.1072607363649132e-07,
+      "loss": 0.5867,
+      "step": 1621
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 2.0449121146845774e-07,
+      "loss": 0.1584,
+      "step": 1622
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.9834959796144192e-07,
+      "loss": 0.111,
+      "step": 1623
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.9230125621225724e-07,
+      "loss": 0.0891,
+      "step": 1624
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.8634620896695043e-07,
+      "loss": 0.08,
+      "step": 1625
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.8048447862070718e-07,
+      "loss": 0.1836,
+      "step": 1626
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.747160872177883e-07,
+      "loss": 0.0804,
+      "step": 1627
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.6904105645142444e-07,
+      "loss": 0.0849,
+      "step": 1628
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.634594076637519e-07,
+      "loss": 0.1283,
+      "step": 1629
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.5797116184571304e-07,
+      "loss": 0.1915,
+      "step": 1630
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.5257633963700058e-07,
+      "loss": 0.1081,
+      "step": 1631
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.472749613259661e-07,
+      "loss": 0.1229,
+      "step": 1632
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.4206704684953943e-07,
+      "loss": 0.1189,
+      "step": 1633
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.3695261579316777e-07,
+      "loss": 0.4,
+      "step": 1634
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.319316873907267e-07,
+      "loss": 0.1102,
+      "step": 1635
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.2700428052447033e-07,
+      "loss": 0.1282,
+      "step": 1636
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.22170413724923e-07,
+      "loss": 0.1337,
+      "step": 1637
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.1743010517085428e-07,
+      "loss": 0.1529,
+      "step": 1638
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.1278337268917638e-07,
+      "loss": 0.1305,
+      "step": 1639
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.0823023375489127e-07,
+      "loss": 0.1139,
+      "step": 1640
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.0377070549102141e-07,
+      "loss": 0.099,
+      "step": 1641
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 9.940480466855417e-08,
+      "loss": 0.1106,
+      "step": 1642
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 9.513254770636137e-08,
+      "loss": 0.1359,
+      "step": 1643
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 9.095395067115765e-08,
+      "loss": 0.104,
+      "step": 1644
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.686902927741991e-08,
+      "loss": 0.1064,
+      "step": 1645
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.28777988873486e-08,
+      "loss": 0.0852,
+      "step": 1646
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 7.898027451078982e-08,
+      "loss": 0.1196,
+      "step": 1647
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 7.51764708051994e-08,
+      "loss": 0.0874,
+      "step": 1648
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 7.14664020755651e-08,
+      "loss": 0.1,
+      "step": 1649
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 6.785008227437329e-08,
+      "loss": 0.0636,
+      "step": 1650
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 6.432752500154793e-08,
+      "loss": 0.1006,
+      "step": 1651
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 6.089874350439506e-08,
+      "loss": 0.1569,
+      "step": 1652
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.756375067755837e-08,
+      "loss": 0.0804,
+      "step": 1653
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.4322559062974787e-08,
+      "loss": 0.1229,
+      "step": 1654
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.117518084981621e-08,
+      "loss": 0.0802,
+      "step": 1655
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 4.8121627874450625e-08,
+      "loss": 0.089,
+      "step": 1656
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 4.516191162040051e-08,
+      "loss": 0.1708,
+      "step": 1657
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 4.229604321829561e-08,
+      "loss": 0.0969,
+      "step": 1658
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.9524033445825784e-08,
+      "loss": 0.1339,
+      "step": 1659
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.684589272771044e-08,
+      "loss": 0.1114,
+      "step": 1660
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.426163113565417e-08,
+      "loss": 0.1539,
+      "step": 1661
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.177125838830786e-08,
+      "loss": 0.1131,
+      "step": 1662
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 2.9374783851240928e-08,
+      "loss": 0.1232,
+      "step": 1663
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 2.7072216536885853e-08,
+      "loss": 0.1,
+      "step": 1664
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 2.486356510453258e-08,
+      "loss": 0.0949,
+      "step": 1665
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 2.2748837860270267e-08,
+      "loss": 0.1544,
+      "step": 1666
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2.0728042756967824e-08,
+      "loss": 0.1158,
+      "step": 1667
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.8801187394248965e-08,
+      "loss": 0.1208,
+      "step": 1668
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.696827901845055e-08,
+      "loss": 0.1405,
+      "step": 1669
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.522932452260595e-08,
+      "loss": 0.1005,
+      "step": 1670
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.358433044641727e-08,
+      "loss": 0.1776,
+      "step": 1671
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.2033302976222071e-08,
+      "loss": 0.133,
+      "step": 1672
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.0576247944985018e-08,
+      "loss": 0.1656,
+      "step": 1673
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 9.213170832264584e-09,
+      "loss": 0.1064,
+      "step": 1674
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.944076764190845e-09,
+      "loss": 0.1169,
+      "step": 1675
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 6.768970513457151e-09,
+      "loss": 0.1313,
+      "step": 1676
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.687856499297928e-09,
+      "loss": 0.0968,
+      "step": 1677
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 4.700738787466463e-09,
+      "loss": 0.1149,
+      "step": 1678
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.807621090218261e-09,
+      "loss": 0.1426,
+      "step": 1679
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.008506766313812e-09,
+      "loss": 0.0695,
+      "step": 1680
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 2.3033988209880675e-09,
+      "loss": 0.0958,
+      "step": 1681
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.692299905944883e-09,
+      "loss": 0.1056,
+      "step": 1682
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.1752123193459197e-09,
+      "loss": 0.1037,
+      "step": 1683
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 7.521380058106431e-10,
+      "loss": 0.103,
+      "step": 1684
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.2307855639411865e-10,
+      "loss": 0.0912,
+      "step": 1685
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.8803520859811406e-10,
+      "loss": 0.1589,
+      "step": 1686
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.700884634611935e-11,
+      "loss": 0.1214,
+      "step": 1687
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 0.0,
+      "loss": 0.1061,
+      "step": 1688
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.25203171372413635,
+      "eval_runtime": 59.3935,
+      "eval_samples_per_second": 5.051,
+      "eval_steps_per_second": 0.64,
+      "step": 1688
+    },
+    {
+      "epoch": 2.0,
+      "step": 1688,
+      "total_flos": 4.471289158828032e+18,
+      "train_loss": 0.23114654361115883,
+      "train_runtime": 50320.1158,
+      "train_samples_per_second": 1.073,
+      "train_steps_per_second": 0.034
+    }
+  ],
+  "max_steps": 1688,
+  "num_train_epochs": 2,
+  "total_flos": 4.471289158828032e+18,
+  "trial_name": null,
+  "trial_params": null
+}