diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4770 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 790,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 11.1556,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 11.1531,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.5e-06,
+      "loss": 11.1614,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 11.1446,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 11.0054,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-06,
+      "loss": 10.6955,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.833333333333334e-06,
+      "loss": 10.4968,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 9.706,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 9.5405,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 9.2296,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.166666666666666e-06,
+      "loss": 9.7789,
+      "step": 11
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1e-05,
+      "loss": 8.9721,
+      "step": 12
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0833333333333334e-05,
+      "loss": 8.9369,
+      "step": 13
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1666666666666668e-05,
+      "loss": 9.1961,
+      "step": 14
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.25e-05,
+      "loss": 9.0081,
+      "step": 15
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 9.1296,
+      "step": 16
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.416666666666667e-05,
+      "loss": 8.9271,
+      "step": 17
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 8.6372,
+      "step": 18
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5833333333333333e-05,
+      "loss": 8.2846,
+      "step": 19
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 8.1478,
+      "step": 20
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.7500000000000002e-05,
+      "loss": 8.2667,
+      "step": 21
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.8333333333333333e-05,
+      "loss": 8.1677,
+      "step": 22
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.916666666666667e-05,
+      "loss": 8.1679,
+      "step": 23
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2e-05,
+      "loss": 8.2426,
+      "step": 24
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.9999915896977905e-05,
+      "loss": 8.207,
+      "step": 25
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.999966358932628e-05,
+      "loss": 8.1476,
+      "step": 26
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.999924308128909e-05,
+      "loss": 7.929,
+      "step": 27
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.9998654379939535e-05,
+      "loss": 7.7032,
+      "step": 28
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.9997897495179932e-05,
+      "loss": 7.4376,
+      "step": 29
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.9996972439741537e-05,
+      "loss": 7.3027,
+      "step": 30
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.9995879229184343e-05,
+      "loss": 7.2694,
+      "step": 31
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.999461788189681e-05,
+      "loss": 7.2901,
+      "step": 32
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.9993188419095562e-05,
+      "loss": 7.2255,
+      "step": 33
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.9991590864825026e-05,
+      "loss": 7.0736,
+      "step": 34
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.9989825245957038e-05,
+      "loss": 7.0355,
+      "step": 35
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.9987891592190367e-05,
+      "loss": 6.9484,
+      "step": 36
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.998578993605024e-05,
+      "loss": 7.0153,
+      "step": 37
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.9983520312887785e-05,
+      "loss": 6.9302,
+      "step": 38
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.9981082760879432e-05,
+      "loss": 6.9273,
+      "step": 39
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.9978477321026282e-05,
+      "loss": 6.8868,
+      "step": 40
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.997570403715341e-05,
+      "loss": 6.8568,
+      "step": 41
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.997276295590912e-05,
+      "loss": 6.8307,
+      "step": 42
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.9969654126764183e-05,
+      "loss": 6.7708,
+      "step": 43
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9966377602010984e-05,
+      "loss": 6.6811,
+      "step": 44
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9962933436762644e-05,
+      "loss": 6.6613,
+      "step": 45
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.995932168895211e-05,
+      "loss": 6.6736,
+      "step": 46
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9955542419331162e-05,
+      "loss": 6.5919,
+      "step": 47
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9951595691469397e-05,
+      "loss": 6.5106,
+      "step": 48
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9947481571753165e-05,
+      "loss": 6.4874,
+      "step": 49
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9943200129384444e-05,
+      "loss": 6.4935,
+      "step": 50
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9938751436379684e-05,
+      "loss": 6.4783,
+      "step": 51
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.9934135567568594e-05,
+      "loss": 6.5052,
+      "step": 52
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.992935260059287e-05,
+      "loss": 6.5075,
+      "step": 53
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.992440261590491e-05,
+      "loss": 6.4666,
+      "step": 54
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.9919285696766453e-05,
+      "loss": 6.406,
+      "step": 55
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.991400192924717e-05,
+      "loss": 6.4111,
+      "step": 56
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.9908551402223218e-05,
+      "loss": 6.4167,
+      "step": 57
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.9902934207375758e-05,
+      "loss": 6.4207,
+      "step": 58
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.989715043918941e-05,
+      "loss": 6.422,
+      "step": 59
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9891200194950644e-05,
+      "loss": 6.3422,
+      "step": 60
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9885083574746167e-05,
+      "loss": 6.2456,
+      "step": 61
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9878800681461222e-05,
+      "loss": 6.2807,
+      "step": 62
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9872351620777883e-05,
+      "loss": 6.3047,
+      "step": 63
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9865736501173237e-05,
+      "loss": 6.2419,
+      "step": 64
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9858955433917602e-05,
+      "loss": 6.2308,
+      "step": 65
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9852008533072627e-05,
+      "loss": 6.2322,
+      "step": 66
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9844895915489378e-05,
+      "loss": 6.2124,
+      "step": 67
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9837617700806385e-05,
+      "loss": 6.1977,
+      "step": 68
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9830174011447617e-05,
+      "loss": 6.1633,
+      "step": 69
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.982256497262043e-05,
+      "loss": 6.1553,
+      "step": 70
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9814790712313456e-05,
+      "loss": 6.1803,
+      "step": 71
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.980685136129445e-05,
+      "loss": 6.1184,
+      "step": 72
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9798747053108098e-05,
+      "loss": 6.155,
+      "step": 73
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9790477924073755e-05,
+      "loss": 6.101,
+      "step": 74
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.978204411328318e-05,
+      "loss": 6.0193,
+      "step": 75
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.977344576259816e-05,
+      "loss": 6.03,
+      "step": 76
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9764683016648156e-05,
+      "loss": 5.9886,
+      "step": 77
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9755756022827847e-05,
+      "loss": 6.0071,
+      "step": 78
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9746664931294667e-05,
+      "loss": 5.9421,
+      "step": 79
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9737409894966267e-05,
+      "loss": 5.9954,
+      "step": 80
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.972799106951796e-05,
+      "loss": 5.9408,
+      "step": 81
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9718408613380077e-05,
+      "loss": 5.9167,
+      "step": 82
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9708662687735316e-05,
+      "loss": 5.9676,
+      "step": 83
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9698753456516047e-05,
+      "loss": 5.8603,
+      "step": 84
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9688681086401526e-05,
+      "loss": 5.827,
+      "step": 85
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9678445746815107e-05,
+      "loss": 5.8725,
+      "step": 86
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9668047609921385e-05,
+      "loss": 5.8081,
+      "step": 87
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9657486850623308e-05,
+      "loss": 5.728,
+      "step": 88
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9646763646559234e-05,
+      "loss": 5.8512,
+      "step": 89
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.963587817809993e-05,
+      "loss": 5.7728,
+      "step": 90
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9624830628345562e-05,
+      "loss": 5.7605,
+      "step": 91
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.961362118312259e-05,
+      "loss": 5.7939,
+      "step": 92
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9602250030980657e-05,
+      "loss": 5.7981,
+      "step": 93
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9590717363189424e-05,
+      "loss": 5.6852,
+      "step": 94
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.957902337373532e-05,
+      "loss": 5.8011,
+      "step": 95
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9567168259318324e-05,
+      "loss": 5.649,
+      "step": 96
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.955515221934863e-05,
+      "loss": 5.6449,
+      "step": 97
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9542975455943284e-05,
+      "loss": 5.5907,
+      "step": 98
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.953063817392281e-05,
+      "loss": 5.6107,
+      "step": 99
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9518140580807746e-05,
+      "loss": 5.5694,
+      "step": 100
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9505482886815167e-05,
+      "loss": 5.6611,
+      "step": 101
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.949266530485513e-05,
+      "loss": 5.5533,
+      "step": 102
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.947968805052712e-05,
+      "loss": 5.4755,
+      "step": 103
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.946655134211639e-05,
+      "loss": 5.6145,
+      "step": 104
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.945325540059032e-05,
+      "loss": 5.6456,
+      "step": 105
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.943980044959468e-05,
+      "loss": 5.4775,
+      "step": 106
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.942618671544988e-05,
+      "loss": 5.7434,
+      "step": 107
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.941241442714716e-05,
+      "loss": 5.6461,
+      "step": 108
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9398483816344728e-05,
+      "loss": 5.8886,
+      "step": 109
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.938439511736388e-05,
+      "loss": 6.0387,
+      "step": 110
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9370148567185043e-05,
+      "loss": 5.8731,
+      "step": 111
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.935574440544381e-05,
+      "loss": 5.7988,
+      "step": 112
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.934118287442689e-05,
+      "loss": 5.5043,
+      "step": 113
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9326464219068023e-05,
+      "loss": 5.5951,
+      "step": 114
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9311588686943897e-05,
+      "loss": 5.7133,
+      "step": 115
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9296556528269954e-05,
+      "loss": 5.4865,
+      "step": 116
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9281367995896187e-05,
+      "loss": 5.5316,
+      "step": 117
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9266023345302886e-05,
+      "loss": 5.5545,
+      "step": 118
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9250522834596357e-05,
+      "loss": 5.5339,
+      "step": 119
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9234866724504554e-05,
+      "loss": 5.4314,
+      "step": 120
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9219055278372713e-05,
+      "loss": 5.4064,
+      "step": 121
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9203088762158914e-05,
+      "loss": 5.3135,
+      "step": 122
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9186967444429613e-05,
+      "loss": 5.3226,
+      "step": 123
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9170691596355113e-05,
+      "loss": 5.3193,
+      "step": 124
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.915426149170502e-05,
+      "loss": 5.1942,
+      "step": 125
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.913767740684362e-05,
+      "loss": 5.3009,
+      "step": 126
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.912093962072525e-05,
+      "loss": 5.1603,
+      "step": 127
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.910404841488959e-05,
+      "loss": 5.2039,
+      "step": 128
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9087004073456926e-05,
+      "loss": 5.1635,
+      "step": 129
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9069806883123386e-05,
+      "loss": 5.2517,
+      "step": 130
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9052457133156103e-05,
+      "loss": 5.2288,
+      "step": 131
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9034955115388364e-05,
+      "loss": 5.1153,
+      "step": 132
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.901730112421468e-05,
+      "loss": 5.0183,
+      "step": 133
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.8999495456585856e-05,
+      "loss": 5.1861,
+      "step": 134
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.898153841200398e-05,
+      "loss": 5.0127,
+      "step": 135
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.8963430292517398e-05,
+      "loss": 5.0361,
+      "step": 136
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.8945171402715628e-05,
+      "loss": 4.976,
+      "step": 137
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.892676204972423e-05,
+      "loss": 4.9992,
+      "step": 138
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.8908202543199646e-05,
+      "loss": 4.9535,
+      "step": 139
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.8889493195324e-05,
+      "loss": 5.0195,
+      "step": 140
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.8870634320799822e-05,
+      "loss": 4.9628,
+      "step": 141
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.8851626236844787e-05,
+      "loss": 4.9592,
+      "step": 142
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.8832469263186352e-05,
+      "loss": 4.8603,
+      "step": 143
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.8813163722056397e-05,
+      "loss": 4.9138,
+      "step": 144
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.879370993818579e-05,
+      "loss": 5.0571,
+      "step": 145
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.8774108238798932e-05,
+      "loss": 4.9104,
+      "step": 146
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.875435895360826e-05,
+      "loss": 4.9345,
+      "step": 147
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.873446241480868e-05,
+      "loss": 4.8746,
+      "step": 148
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.8714418957072008e-05,
+      "loss": 4.8118,
+      "step": 149
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.8694228917541313e-05,
+      "loss": 4.9833,
+      "step": 150
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.8673892635825264e-05,
+      "loss": 4.777,
+      "step": 151
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.8653410453992415e-05,
+      "loss": 4.8645,
+      "step": 152
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.8632782716565438e-05,
+      "loss": 4.7069,
+      "step": 153
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.861200977051535e-05,
+      "loss": 4.7141,
+      "step": 154
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8591091965255654e-05,
+      "loss": 4.7068,
+      "step": 155
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.857002965263648e-05,
+      "loss": 4.8115,
+      "step": 156
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.854882318693866e-05,
+      "loss": 4.7032,
+      "step": 157
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.852747292486776e-05,
+      "loss": 4.7396,
+      "step": 158
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.850597922554809e-05,
+      "loss": 4.577,
+      "step": 159
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8484342450516672e-05,
+      "loss": 4.6603,
+      "step": 160
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8462562963717134e-05,
+      "loss": 4.6679,
+      "step": 161
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.844064113149361e-05,
+      "loss": 4.734,
+      "step": 162
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.841857732258457e-05,
+      "loss": 4.6387,
+      "step": 163
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.839637190811661e-05,
+      "loss": 4.7254,
+      "step": 164
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8374025261598224e-05,
+      "loss": 4.5621,
+      "step": 165
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8351537758913518e-05,
+      "loss": 4.5779,
+      "step": 166
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8328909778315876e-05,
+      "loss": 4.5663,
+      "step": 167
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8306141700421606e-05,
+      "loss": 4.444,
+      "step": 168
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8283233908203547e-05,
+      "loss": 4.4928,
+      "step": 169
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.8260186786984603e-05,
+      "loss": 4.6391,
+      "step": 170
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.8237000724431283e-05,
+      "loss": 4.5151,
+      "step": 171
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.8213676110547177e-05,
+      "loss": 4.5845,
+      "step": 172
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.8190213337666384e-05,
+      "loss": 4.6617,
+      "step": 173
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.816661280044693e-05,
+      "loss": 4.4839,
+      "step": 174
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.8142874895864113e-05,
+      "loss": 4.674,
+      "step": 175
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.8119000023203838e-05,
+      "loss": 4.3447,
+      "step": 176
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.809498858405589e-05,
+      "loss": 4.5803,
+      "step": 177
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.807084098230719e-05,
+      "loss": 4.5345,
+      "step": 178
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8046557624134997e-05,
+      "loss": 4.4603,
+      "step": 179
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.802213891800007e-05,
+      "loss": 4.4549,
+      "step": 180
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.799758527463981e-05,
+      "loss": 4.5114,
+      "step": 181
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.797289710706133e-05,
+      "loss": 4.3743,
+      "step": 182
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.7948074830534535e-05,
+      "loss": 4.441,
+      "step": 183
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.7923118862585123e-05,
+      "loss": 4.5071,
+      "step": 184
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.7898029622987555e-05,
+      "loss": 4.3857,
+      "step": 185
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.7872807533758007e-05,
+      "loss": 4.4135,
+      "step": 186
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.7847453019147264e-05,
+      "loss": 4.4598,
+      "step": 187
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.7821966505633587e-05,
+      "loss": 4.4086,
+      "step": 188
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.7796348421915536e-05,
+      "loss": 4.3538,
+      "step": 189
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.7770599198904762e-05,
+      "loss": 4.4076,
+      "step": 190
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.774471926971877e-05,
+      "loss": 4.3045,
+      "step": 191
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.7718709069673595e-05,
+      "loss": 4.5118,
+      "step": 192
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.7692569036276533e-05,
+      "loss": 4.4237,
+      "step": 193
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.7666299609218745e-05,
+      "loss": 4.3038,
+      "step": 194
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.763990123036787e-05,
+      "loss": 4.3774,
+      "step": 195
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.7613374343760595e-05,
+      "loss": 4.282,
+      "step": 196
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.7586719395595185e-05,
+      "loss": 4.2489,
+      "step": 197
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.7559936834223982e-05,
+      "loss": 4.2206,
+      "step": 198
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.7533027110145857e-05,
+      "loss": 4.1831,
+      "step": 199
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.7505990675998632e-05,
+      "loss": 4.4274,
+      "step": 200
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.747882798655147e-05,
+      "loss": 4.2656,
+      "step": 201
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7451539498697225e-05,
+      "loss": 4.355,
+      "step": 202
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.742412567144476e-05,
+      "loss": 4.3059,
+      "step": 203
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.739658696591121e-05,
+      "loss": 4.2162,
+      "step": 204
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7368923845314262e-05,
+      "loss": 4.2575,
+      "step": 205
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7341136774964305e-05,
+      "loss": 4.1695,
+      "step": 206
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7313226222256675e-05,
+      "loss": 4.1604,
+      "step": 207
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.728519265666373e-05,
+      "loss": 4.2444,
+      "step": 208
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7257036549726988e-05,
+      "loss": 4.1928,
+      "step": 209
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.7228758375049186e-05,
+      "loss": 4.2985,
+      "step": 210
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.7200358608286314e-05,
+      "loss": 4.3212,
+      "step": 211
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.7171837727139613e-05,
+      "loss": 4.2987,
+      "step": 212
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.714319621134755e-05,
+      "loss": 4.2621,
+      "step": 213
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.711443454267772e-05,
+      "loss": 4.2537,
+      "step": 214
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.708555320491878e-05,
+      "loss": 4.1606,
+      "step": 215
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.705655268387229e-05,
+      "loss": 4.0981,
+      "step": 216
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.702743346734454e-05,
+      "loss": 4.1428,
+      "step": 217
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.6998196045138354e-05,
+      "loss": 4.0398,
+      "step": 218
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.696884090904484e-05,
+      "loss": 4.1158,
+      "step": 219
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.6939368552835137e-05,
+      "loss": 4.2338,
+      "step": 220
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.6909779472252084e-05,
+      "loss": 4.2012,
+      "step": 221
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.6880074165001906e-05,
+      "loss": 4.1079,
+      "step": 222
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.685025313074582e-05,
+      "loss": 4.2466,
+      "step": 223
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.682031687109165e-05,
+      "loss": 4.0848,
+      "step": 224
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.679026588958538e-05,
+      "loss": 4.2162,
+      "step": 225
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.6760100691702676e-05,
+      "loss": 4.1318,
+      "step": 226
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.6729821784840398e-05,
+      "loss": 4.2007,
+      "step": 227
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.669942967830807e-05,
+      "loss": 4.1688,
+      "step": 228
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.6668924883319288e-05,
+      "loss": 4.1256,
+      "step": 229
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.6638307912983135e-05,
+      "loss": 4.0463,
+      "step": 230
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.6607579282295572e-05,
+      "loss": 4.1453,
+      "step": 231
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.6576739508130725e-05,
+      "loss": 4.0776,
+      "step": 232
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.6545789109232247e-05,
+      "loss": 4.2646,
+      "step": 233
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.651472860620455e-05,
+      "loss": 4.0992,
+      "step": 234
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.6483558521504068e-05,
+      "loss": 4.0812,
+      "step": 235
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.6452279379430466e-05,
+      "loss": 4.0722,
+      "step": 236
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.6420891706117818e-05,
+      "loss": 4.0237,
+      "step": 237
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.638939602952576e-05,
+      "loss": 4.1031,
+      "step": 238
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.6357792879430615e-05,
+      "loss": 4.0923,
+      "step": 239
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.632608278741646e-05,
+      "loss": 4.0142,
+      "step": 240
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.629426628686622e-05,
+      "loss": 4.0109,
+      "step": 241
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.6262343912952656e-05,
+      "loss": 3.9415,
+      "step": 242
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.6230316202629393e-05,
+      "loss": 4.0262,
+      "step": 243
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.619818369462188e-05,
+      "loss": 3.951,
+      "step": 244
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.6165946929418322e-05,
+      "loss": 4.0882,
+      "step": 245
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.613360644926059e-05,
+      "loss": 4.054,
+      "step": 246
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.610116279813511e-05,
+      "loss": 4.0486,
+      "step": 247
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.6068616521763708e-05,
+      "loss": 4.0278,
+      "step": 248
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.603596816759442e-05,
+      "loss": 4.0248,
+      "step": 249
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.60032182847923e-05,
+      "loss": 4.0702,
+      "step": 250
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.5970367424230162e-05,
+      "loss": 3.8974,
+      "step": 251
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.5937416138479344e-05,
+      "loss": 4.0375,
+      "step": 252
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.590436498180039e-05,
+      "loss": 4.0119,
+      "step": 253
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.5871214510133734e-05,
+      "loss": 4.0036,
+      "step": 254
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.5837965281090334e-05,
+      "loss": 3.9336,
+      "step": 255
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.580461785394233e-05,
+      "loss": 3.806,
+      "step": 256
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.57711727896136e-05,
+      "loss": 3.9693,
+      "step": 257
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.5737630650670336e-05,
+      "loss": 3.9221,
+      "step": 258
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.57039920013116e-05,
+      "loss": 3.8656,
+      "step": 259
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.567025740735979e-05,
+      "loss": 4.0068,
+      "step": 260
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.5636427436251182e-05,
+      "loss": 3.9293,
+      "step": 261
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.5602502657026327e-05,
+      "loss": 3.9401,
+      "step": 262
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.556848364032052e-05,
+      "loss": 3.9274,
+      "step": 263
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.5534370958354184e-05,
+      "loss": 4.1849,
+      "step": 264
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.550016518492325e-05,
+      "loss": 3.9264,
+      "step": 265
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.5465866895389497e-05,
+      "loss": 3.8007,
+      "step": 266
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.5431476666670885e-05,
+      "loss": 3.8248,
+      "step": 267
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.5396995077231856e-05,
+      "loss": 3.891,
+      "step": 268
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.5362422707073574e-05,
+      "loss": 3.801,
+      "step": 269
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.5327760137724213e-05,
+      "loss": 3.9024,
+      "step": 270
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.5293007952229127e-05,
+      "loss": 3.9204,
+      "step": 271
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.5258166735141094e-05,
+      "loss": 3.9807,
+      "step": 272
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.5223237072510433e-05,
+      "loss": 3.906,
+      "step": 273
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.518821955187519e-05,
+      "loss": 3.8716,
+      "step": 274
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.5153114762251221e-05,
+      "loss": 3.8384,
+      "step": 275
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.5117923294122312e-05,
+      "loss": 3.7808,
+      "step": 276
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.5082645739430224e-05,
+      "loss": 3.9122,
+      "step": 277
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.5047282691564749e-05,
+      "loss": 3.7987,
+      "step": 278
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.5011834745353725e-05,
+      "loss": 3.967,
+      "step": 279
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.4976302497053036e-05,
+      "loss": 3.9335,
+      "step": 280
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.4940686544336571e-05,
+      "loss": 3.8081,
+      "step": 281
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.4904987486286184e-05,
+      "loss": 3.9681,
+      "step": 282
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.4869205923381609e-05,
+      "loss": 3.8072,
+      "step": 283
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.4833342457490363e-05,
+      "loss": 3.809,
+      "step": 284
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.4797397691857614e-05,
+      "loss": 3.8076,
+      "step": 285
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.4761372231096047e-05,
+      "loss": 3.8011,
+      "step": 286
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.472526668117569e-05,
+      "loss": 3.9507,
+      "step": 287
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.468908164941371e-05,
+      "loss": 3.8249,
+      "step": 288
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.4652817744464214e-05,
+      "loss": 3.8342,
+      "step": 289
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.4616475576308005e-05,
+      "loss": 3.7757,
+      "step": 290
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.4580055756242315e-05,
+      "loss": 3.9205,
+      "step": 291
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.454355889687053e-05,
+      "loss": 3.8701,
+      "step": 292
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.4506985612091889e-05,
+      "loss": 3.8211,
+      "step": 293
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.4470336517091139e-05,
+      "loss": 3.9225,
+      "step": 294
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.4433612228328215e-05,
+      "loss": 3.7957,
+      "step": 295
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.439681336352785e-05,
+      "loss": 3.745,
+      "step": 296
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.435994054166919e-05,
+      "loss": 3.8174,
+      "step": 297
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.4322994382975386e-05,
+      "loss": 3.7996,
+      "step": 298
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.428597550890316e-05,
+      "loss": 3.789,
+      "step": 299
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.4248884542132348e-05,
+      "loss": 3.9083,
+      "step": 300
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.421172210655543e-05,
+      "loss": 3.6434,
+      "step": 301
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.4174488827267032e-05,
+      "loss": 3.6945,
+      "step": 302
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.4137185330553416e-05,
+      "loss": 3.8261,
+      "step": 303
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.4099812243881947e-05,
+      "loss": 3.8346,
+      "step": 304
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.406237019589053e-05,
+      "loss": 3.7792,
+      "step": 305
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.4024859816377046e-05,
+      "loss": 3.7122,
+      "step": 306
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.3987281736288743e-05,
+      "loss": 3.7728,
+      "step": 307
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.3949636587711645e-05,
+      "loss": 3.6981,
+      "step": 308
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.3911925003859907e-05,
+      "loss": 3.8742,
+      "step": 309
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.3874147619065161e-05,
+      "loss": 3.7379,
+      "step": 310
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.3836305068765852e-05,
+      "loss": 3.7293,
+      "step": 311
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.3798397989496549e-05,
+      "loss": 3.7363,
+      "step": 312
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.3760427018877236e-05,
+      "loss": 3.6561,
+      "step": 313
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.3722392795602595e-05,
+      "loss": 3.8963,
+      "step": 314
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.3684295959431241e-05,
+      "loss": 3.6974,
+      "step": 315
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.3646137151174992e-05,
+      "loss": 3.9303,
+      "step": 316
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.3607917012688063e-05,
+      "loss": 3.5819,
+      "step": 317
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.3569636186856288e-05,
+      "loss": 3.7086,
+      "step": 318
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.3531295317586291e-05,
+      "loss": 3.6635,
+      "step": 319
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.349289504979467e-05,
+      "loss": 3.7356,
+      "step": 320
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.3454436029397135e-05,
+      "loss": 3.8162,
+      "step": 321
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.341591890329766e-05,
+      "loss": 3.7543,
+      "step": 322
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.3377344319377585e-05,
+      "loss": 3.627,
+      "step": 323
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.3338712926484725e-05,
+      "loss": 3.7216,
+      "step": 324
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.3300025374422459e-05,
+      "loss": 3.6936,
+      "step": 325
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.3261282313938795e-05,
+      "loss": 3.8305,
+      "step": 326
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.322248439671543e-05,
+      "loss": 3.6515,
+      "step": 327
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.3183632275356777e-05,
+      "loss": 3.7233,
+      "step": 328
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.3144726603379008e-05,
+      "loss": 3.6564,
+      "step": 329
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.3105768035199033e-05,
+      "loss": 3.8174,
+      "step": 330
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.3066757226123522e-05,
+      "loss": 3.6796,
+      "step": 331
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.3027694832337858e-05,
+      "loss": 3.6757,
+      "step": 332
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.2988581510895118e-05,
+      "loss": 3.7172,
+      "step": 333
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.2949417919705008e-05,
+      "loss": 3.6418,
+      "step": 334
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.2910204717522805e-05,
+      "loss": 3.6337,
+      "step": 335
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.2870942563938265e-05,
+      "loss": 3.7748,
+      "step": 336
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.283163211936455e-05,
+      "loss": 3.6219,
+      "step": 337
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.279227404502709e-05,
+      "loss": 3.6606,
+      "step": 338
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.2752869002952492e-05,
+      "loss": 3.5741,
+      "step": 339
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.2713417655957377e-05,
+      "loss": 3.6364,
+      "step": 340
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.2673920667637244e-05,
+      "loss": 3.7044,
+      "step": 341
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.2634378702355314e-05,
+      "loss": 3.7507,
+      "step": 342
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.2594792425231339e-05,
+      "loss": 3.695,
+      "step": 343
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.2555162502130434e-05,
+      "loss": 3.6151,
+      "step": 344
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.251548959965185e-05,
+      "loss": 3.6708,
+      "step": 345
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.2475774385117787e-05,
+      "loss": 3.5673,
+      "step": 346
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.243601752656216e-05,
+      "loss": 3.5632,
+      "step": 347
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.2396219692719364e-05,
+      "loss": 3.6272,
+      "step": 348
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.2356381553013014e-05,
+      "loss": 3.6401,
+      "step": 349
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.23165037775447e-05,
+      "loss": 3.6209,
+      "step": 350
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.2276587037082706e-05,
+      "loss": 3.6672,
+      "step": 351
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.2236632003050736e-05,
+      "loss": 3.7134,
+      "step": 352
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.2196639347516613e-05,
+      "loss": 3.7017,
+      "step": 353
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.215660974318097e-05,
+      "loss": 3.7002,
+      "step": 354
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.211654386336595e-05,
+      "loss": 3.7195,
+      "step": 355
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.207644238200387e-05,
+      "loss": 3.6528,
+      "step": 356
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.2036305973625881e-05,
+      "loss": 3.5246,
+      "step": 357
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.1996135313350636e-05,
+      "loss": 3.5674,
+      "step": 358
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.1955931076872916e-05,
+      "loss": 3.6463,
+      "step": 359
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.191569394045228e-05,
+      "loss": 3.543,
+      "step": 360
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.1875424580901684e-05,
+      "loss": 3.6849,
+      "step": 361
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.1835123675576092e-05,
+      "loss": 3.6798,
+      "step": 362
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.1794791902361093e-05,
+      "loss": 3.639,
+      "step": 363
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.1754429939661492e-05,
+      "loss": 3.6447,
+      "step": 364
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.1714038466389892e-05,
+      "loss": 3.6637,
+      "step": 365
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.1673618161955288e-05,
+      "loss": 3.6719,
+      "step": 366
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.1633169706251637e-05,
+      "loss": 3.7565,
+      "step": 367
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.1592693779646405e-05,
+      "loss": 3.645,
+      "step": 368
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.1552191062969147e-05,
+      "loss": 3.5035,
+      "step": 369
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.1511662237500032e-05,
+      "loss": 3.6058,
+      "step": 370
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.1471107984958405e-05,
+      "loss": 3.5748,
+      "step": 371
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.1430528987491305e-05,
+      "loss": 3.4794,
+      "step": 372
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.1389925927661996e-05,
+      "loss": 3.6624,
+      "step": 373
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.1349299488438485e-05,
+      "loss": 3.5667,
+      "step": 374
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.1308650353182036e-05,
+      "loss": 3.5315,
+      "step": 375
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.1267979205635675e-05,
+      "loss": 3.5055,
+      "step": 376
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.1227286729912684e-05,
+      "loss": 3.5781,
+      "step": 377
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.1186573610485099e-05,
+      "loss": 3.4713,
+      "step": 378
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.1145840532172197e-05,
+      "loss": 3.5313,
+      "step": 379
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.1105088180128975e-05,
+      "loss": 3.5475,
+      "step": 380
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.1064317239834628e-05,
+      "loss": 3.5527,
+      "step": 381
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.1023528397081011e-05,
+      "loss": 3.5274,
+      "step": 382
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.0982722337961116e-05,
+      "loss": 3.5427,
+      "step": 383
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.094189974885752e-05,
+      "loss": 3.4859,
+      "step": 384
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.0901061316430848e-05,
+      "loss": 3.5526,
+      "step": 385
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.0860207727608214e-05,
+      "loss": 3.544,
+      "step": 386
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.0819339669571674e-05,
+      "loss": 3.525,
+      "step": 387
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.0778457829746668e-05,
+      "loss": 3.6301,
+      "step": 388
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.0737562895790447e-05,
+      "loss": 3.607,
+      "step": 389
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.0696655555580524e-05,
+      "loss": 3.5375,
+      "step": 390
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.0655736497203084e-05,
+      "loss": 3.4769,
+      "step": 391
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.0614806408941422e-05,
+      "loss": 3.6209,
+      "step": 392
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.0573865979264362e-05,
+      "loss": 3.5285,
+      "step": 393
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.0532915896814673e-05,
+      "loss": 3.4459,
+      "step": 394
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.0491956850397496e-05,
+      "loss": 3.5381,
+      "step": 395
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.0450989528968747e-05,
+      "loss": 3.6046,
+      "step": 396
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.0410014621623531e-05,
+      "loss": 3.5079,
+      "step": 397
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.036903281758456e-05,
+      "loss": 3.4709,
+      "step": 398
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.0328044806190549e-05,
+      "loss": 3.5259,
+      "step": 399
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.028705127688462e-05,
+      "loss": 3.615,
+      "step": 400
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.0246052919202713e-05,
+      "loss": 3.6054,
+      "step": 401
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.0205050422761989e-05,
+      "loss": 3.5232,
+      "step": 402
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.0164044477249215e-05,
+      "loss": 3.5266,
+      "step": 403
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.0123035772409184e-05,
+      "loss": 3.4681,
+      "step": 404
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.0082024998033092e-05,
+      "loss": 3.4198,
+      "step": 405
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.004101284394696e-05,
+      "loss": 3.4836,
+      "step": 406
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 3.3732,
+      "step": 407
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.958987156053046e-06,
+      "loss": 3.631,
+      "step": 408
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.91797500196691e-06,
+      "loss": 3.3575,
+      "step": 409
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.876964227590821e-06,
+      "loss": 3.5228,
+      "step": 410
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.835955522750789e-06,
+      "loss": 3.4126,
+      "step": 411
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.794949577238014e-06,
+      "loss": 3.5752,
+      "step": 412
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.753947080797289e-06,
+      "loss": 3.6233,
+      "step": 413
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.712948723115384e-06,
+      "loss": 3.5602,
+      "step": 414
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.671955193809453e-06,
+      "loss": 3.3987,
+      "step": 415
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.630967182415441e-06,
+      "loss": 3.5806,
+      "step": 416
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.589985378376474e-06,
+      "loss": 3.5712,
+      "step": 417
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.549010471031256e-06,
+      "loss": 3.5768,
+      "step": 418
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.508043149602509e-06,
+      "loss": 3.5503,
+      "step": 419
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.46708410318533e-06,
+      "loss": 3.4928,
+      "step": 420
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.426134020735642e-06,
+      "loss": 3.5639,
+      "step": 421
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.38519359105858e-06,
+      "loss": 3.4968,
+      "step": 422
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.344263502796918e-06,
+      "loss": 3.5206,
+      "step": 423
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.303344444419476e-06,
+      "loss": 3.5223,
+      "step": 424
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.262437104209555e-06,
+      "loss": 3.5115,
+      "step": 425
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.221542170253339e-06,
+      "loss": 3.5129,
+      "step": 426
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.18066033042833e-06,
+      "loss": 3.4762,
+      "step": 427
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.139792272391791e-06,
+      "loss": 3.4295,
+      "step": 428
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.098938683569155e-06,
+      "loss": 3.402,
+      "step": 429
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.058100251142483e-06,
+      "loss": 3.4283,
+      "step": 430
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.017277662038884e-06,
+      "loss": 3.5585,
+      "step": 431
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.97647160291899e-06,
+      "loss": 3.4686,
+      "step": 432
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.935682760165377e-06,
+      "loss": 3.3924,
+      "step": 433
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.894911819871027e-06,
+      "loss": 3.5207,
+      "step": 434
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.854159467827808e-06,
+      "loss": 3.4214,
+      "step": 435
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.813426389514903e-06,
+      "loss": 3.5747,
+      "step": 436
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.77271327008732e-06,
+      "loss": 3.4337,
+      "step": 437
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.732020794364327e-06,
+      "loss": 3.4204,
+      "step": 438
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.691349646817965e-06,
+      "loss": 3.4279,
+      "step": 439
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.650700511561515e-06,
+      "loss": 3.4796,
+      "step": 440
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.610074072338006e-06,
+      "loss": 3.3757,
+      "step": 441
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.5694710125087e-06,
+      "loss": 3.4833,
+      "step": 442
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.528892015041598e-06,
+      "loss": 3.6173,
+      "step": 443
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.488337762499971e-06,
+      "loss": 3.3047,
+      "step": 444
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.447808937030856e-06,
+      "loss": 3.5363,
+      "step": 445
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.407306220353597e-06,
+      "loss": 3.4376,
+      "step": 446
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.366830293748364e-06,
+      "loss": 3.4486,
+      "step": 447
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.326381838044713e-06,
+      "loss": 3.399,
+      "step": 448
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.28596153361011e-06,
+      "loss": 3.4425,
+      "step": 449
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.245570060338511e-06,
+      "loss": 3.4965,
+      "step": 450
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.20520809763891e-06,
+      "loss": 3.4606,
+      "step": 451
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.16487632442391e-06,
+      "loss": 3.3802,
+      "step": 452
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.124575419098321e-06,
+      "loss": 3.4718,
+      "step": 453
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.084306059547722e-06,
+      "loss": 3.4686,
+      "step": 454
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.044068923127088e-06,
+      "loss": 3.4968,
+      "step": 455
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.003864686649366e-06,
+      "loss": 3.3585,
+      "step": 456
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 7.96369402637412e-06,
+      "loss": 3.4358,
+      "step": 457
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 7.923557617996132e-06,
+      "loss": 3.3583,
+      "step": 458
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 7.883456136634053e-06,
+      "loss": 3.4183,
+      "step": 459
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 7.843390256819034e-06,
+      "loss": 3.4337,
+      "step": 460
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 7.803360652483392e-06,
+      "loss": 3.4407,
+      "step": 461
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 7.763367996949267e-06,
+      "loss": 3.2601,
+      "step": 462
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.723412962917294e-06,
+      "loss": 3.2422,
+      "step": 463
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.683496222455303e-06,
+      "loss": 3.4645,
+      "step": 464
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.643618446986988e-06,
+      "loss": 3.5709,
+      "step": 465
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.6037803072806396e-06,
+      "loss": 3.2944,
+      "step": 466
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.563982473437843e-06,
+      "loss": 3.4792,
+      "step": 467
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.524225614882216e-06,
+      "loss": 3.3571,
+      "step": 468
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.4845104003481565e-06,
+      "loss": 3.3765,
+      "step": 469
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.444837497869569e-06,
+      "loss": 3.4416,
+      "step": 470
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.4052075747686625e-06,
+      "loss": 3.357,
+      "step": 471
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.365621297644686e-06,
+      "loss": 3.5674,
+      "step": 472
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.326079332362756e-06,
+      "loss": 3.2106,
+      "step": 473
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.286582344042625e-06,
+      "loss": 3.3728,
+      "step": 474
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.247130997047511e-06,
+      "loss": 3.4861,
+      "step": 475
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.207725954972913e-06,
+      "loss": 3.4695,
+      "step": 476
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.168367880635454e-06,
+      "loss": 3.3745,
+      "step": 477
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.12905743606174e-06,
+      "loss": 3.366,
+      "step": 478
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.089795282477199e-06,
+      "loss": 3.4958,
+      "step": 479
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.050582080294996e-06,
+      "loss": 3.2569,
+      "step": 480
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.011418489104883e-06,
+      "loss": 3.4334,
+      "step": 481
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 6.972305167662144e-06,
+      "loss": 3.491,
+      "step": 482
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 6.933242773876481e-06,
+      "loss": 3.3694,
+      "step": 483
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 6.89423196480097e-06,
+      "loss": 3.5006,
+      "step": 484
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 6.855273396620999e-06,
+      "loss": 3.3757,
+      "step": 485
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.816367724643225e-06,
+      "loss": 3.4249,
+      "step": 486
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.777515603284575e-06,
+      "loss": 3.3526,
+      "step": 487
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.738717686061207e-06,
+      "loss": 3.4817,
+      "step": 488
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.699974625577545e-06,
+      "loss": 3.455,
+      "step": 489
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.661287073515276e-06,
+      "loss": 3.3259,
+      "step": 490
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.622655680622416e-06,
+      "loss": 3.2455,
+      "step": 491
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.5840810967023405e-06,
+      "loss": 3.2447,
+      "step": 492
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.545563970602867e-06,
+      "loss": 3.2857,
+      "step": 493
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.507104950205336e-06,
+      "loss": 3.3836,
+      "step": 494
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.4687046824137115e-06,
+      "loss": 3.4351,
+      "step": 495
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.430363813143716e-06,
+      "loss": 3.3191,
+      "step": 496
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.392082987311938e-06,
+      "loss": 3.496,
+      "step": 497
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.353862848825011e-06,
+      "loss": 3.2581,
+      "step": 498
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.315704040568761e-06,
+      "loss": 3.3718,
+      "step": 499
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.277607204397409e-06,
+      "loss": 3.445,
+      "step": 500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.239572981122766e-06,
+      "loss": 3.2989,
+      "step": 501
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 6.201602010503454e-06,
+      "loss": 3.2606,
+      "step": 502
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 6.163694931234153e-06,
+      "loss": 3.3261,
+      "step": 503
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 6.125852380934841e-06,
+      "loss": 3.4656,
+      "step": 504
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 6.088074996140096e-06,
+      "loss": 3.3822,
+      "step": 505
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 6.050363412288356e-06,
+      "loss": 3.3442,
+      "step": 506
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 6.012718263711261e-06,
+      "loss": 3.3625,
+      "step": 507
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.975140183622958e-06,
+      "loss": 3.3852,
+      "step": 508
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.93762980410947e-06,
+      "loss": 3.4399,
+      "step": 509
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.900187756118055e-06,
+      "loss": 3.1638,
+      "step": 510
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.862814669446586e-06,
+      "loss": 3.2364,
+      "step": 511
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.825511172732971e-06,
+      "loss": 3.2059,
+      "step": 512
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.788277893444574e-06,
+      "loss": 3.239,
+      "step": 513
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.7511154578676535e-06,
+      "loss": 3.3294,
+      "step": 514
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.714024491096841e-06,
+      "loss": 3.2771,
+      "step": 515
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.677005617024618e-06,
+      "loss": 3.4053,
+      "step": 516
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.640059458330811e-06,
+      "loss": 3.25,
+      "step": 517
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.603186636472156e-06,
+      "loss": 3.2526,
+      "step": 518
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.566387771671788e-06,
+      "loss": 3.4599,
+      "step": 519
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.529663482908864e-06,
+      "loss": 3.3457,
+      "step": 520
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.4930143879081146e-06,
+      "loss": 3.2894,
+      "step": 521
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.4564411031294695e-06,
+      "loss": 3.3749,
+      "step": 522
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.419944243757685e-06,
+      "loss": 3.3198,
+      "step": 523
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.383524423691995e-06,
+      "loss": 3.3991,
+      "step": 524
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.34718225553579e-06,
+      "loss": 3.429,
+      "step": 525
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.310918350586292e-06,
+      "loss": 3.3088,
+      "step": 526
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.274733318824317e-06,
+      "loss": 3.2755,
+      "step": 527
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.238627768903957e-06,
+      "loss": 3.1154,
+      "step": 528
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.202602308142391e-06,
+      "loss": 3.1803,
+      "step": 529
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.16665754250964e-06,
+      "loss": 3.2609,
+      "step": 530
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.130794076618391e-06,
+      "loss": 3.1746,
+      "step": 531
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.095012513713815e-06,
+      "loss": 3.2586,
+      "step": 532
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.059313455663429e-06,
+      "loss": 3.3339,
+      "step": 533
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.02369750294697e-06,
+      "loss": 3.178,
+      "step": 534
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.988165254646278e-06,
+      "loss": 3.3395,
+      "step": 535
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.952717308435254e-06,
+      "loss": 3.3983,
+      "step": 536
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.917354260569779e-06,
+      "loss": 3.1509,
+      "step": 537
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.8820767058776895e-06,
+      "loss": 3.3076,
+      "step": 538
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.846885237748779e-06,
+      "loss": 3.2966,
+      "step": 539
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.811780448124812e-06,
+      "loss": 3.4145,
+      "step": 540
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.776762927489568e-06,
+      "loss": 3.2694,
+      "step": 541
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.74183326485891e-06,
+      "loss": 3.2737,
+      "step": 542
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.706992047770877e-06,
+      "loss": 3.2447,
+      "step": 543
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.672239862275794e-06,
+      "loss": 3.416,
+      "step": 544
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.637577292926428e-06,
+      "loss": 3.3194,
+      "step": 545
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.6030049227681484e-06,
+      "loss": 3.3098,
+      "step": 546
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.568523333329116e-06,
+      "loss": 3.2507,
+      "step": 547
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.534133104610507e-06,
+      "loss": 3.2184,
+      "step": 548
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.4998348150767525e-06,
+      "loss": 3.3042,
+      "step": 549
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.465629041645819e-06,
+      "loss": 3.2976,
+      "step": 550
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.43151635967948e-06,
+      "loss": 3.2555,
+      "step": 551
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.397497342973677e-06,
+      "loss": 3.2512,
+      "step": 552
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.363572563748823e-06,
+      "loss": 3.2136,
+      "step": 553
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.329742592640212e-06,
+      "loss": 3.4037,
+      "step": 554
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.296007998688405e-06,
+      "loss": 3.2952,
+      "step": 555
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.262369349329665e-06,
+      "loss": 3.3537,
+      "step": 556
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.228827210386404e-06,
+      "loss": 3.1924,
+      "step": 557
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.195382146057672e-06,
+      "loss": 3.2783,
+      "step": 558
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.162034718909671e-06,
+      "loss": 3.263,
+      "step": 559
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.12878548986627e-06,
+      "loss": 3.3612,
+      "step": 560
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.095635018199612e-06,
+      "loss": 3.2958,
+      "step": 561
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.062583861520657e-06,
+      "loss": 3.0953,
+      "step": 562
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.029632575769841e-06,
+      "loss": 3.1859,
+      "step": 563
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.996781715207706e-06,
+      "loss": 3.1945,
+      "step": 564
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.964031832405581e-06,
+      "loss": 3.1413,
+      "step": 565
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.931383478236292e-06,
+      "loss": 3.3722,
+      "step": 566
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.8988372018648905e-06,
+      "loss": 3.2463,
+      "step": 567
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.866393550739415e-06,
+      "loss": 3.221,
+      "step": 568
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.834053070581685e-06,
+      "loss": 3.3415,
+      "step": 569
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.8018163053781243e-06,
+      "loss": 3.2295,
+      "step": 570
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.769683797370609e-06,
+      "loss": 3.2582,
+      "step": 571
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.7376560870473465e-06,
+      "loss": 3.2662,
+      "step": 572
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.7057337131337822e-06,
+      "loss": 3.2425,
+      "step": 573
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.6739172125835386e-06,
+      "loss": 3.3492,
+      "step": 574
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.6422071205693866e-06,
+      "loss": 3.2789,
+      "step": 575
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.610603970474239e-06,
+      "loss": 3.1523,
+      "step": 576
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.5791082938821874e-06,
+      "loss": 3.3764,
+      "step": 577
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.5477206205695392e-06,
+      "loss": 3.2249,
+      "step": 578
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.5164414784959368e-06,
+      "loss": 3.1767,
+      "step": 579
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.485271393795453e-06,
+      "loss": 3.1767,
+      "step": 580
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.454210890767755e-06,
+      "loss": 3.296,
+      "step": 581
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.423260491869276e-06,
+      "loss": 3.3156,
+      "step": 582
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.392420717704431e-06,
+      "loss": 3.3446,
+      "step": 583
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.3616920870168633e-06,
+      "loss": 3.3076,
+      "step": 584
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.331075116680715e-06,
+      "loss": 3.3178,
+      "step": 585
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.3005703216919336e-06,
+      "loss": 3.158,
+      "step": 586
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.2701782151596038e-06,
+      "loss": 3.2794,
+      "step": 587
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.2398993082973294e-06,
+      "loss": 3.2356,
+      "step": 588
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.209734110414625e-06,
+      "loss": 3.2024,
+      "step": 589
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.179683128908352e-06,
+      "loss": 3.3495,
+      "step": 590
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.1497468692541812e-06,
+      "loss": 3.2733,
+      "step": 591
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.1199258349980965e-06,
+      "loss": 3.2758,
+      "step": 592
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.090220527747916e-06,
+      "loss": 3.1778,
+      "step": 593
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.0606314471648646e-06,
+      "loss": 3.2262,
+      "step": 594
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.0311590909551626e-06,
+      "loss": 3.3652,
+      "step": 595
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.0018039548616497e-06,
+      "loss": 3.341,
+      "step": 596
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.972566532655462e-06,
+      "loss": 3.1817,
+      "step": 597
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.943447316127712e-06,
+      "loss": 3.386,
+      "step": 598
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.9144467950812203e-06,
+      "loss": 3.138,
+      "step": 599
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.8855654573222824e-06,
+      "loss": 3.2193,
+      "step": 600
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.8568037886524548e-06,
+      "loss": 3.3578,
+      "step": 601
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.8281622728603862e-06,
+      "loss": 3.2952,
+      "step": 602
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.79964139171369e-06,
+      "loss": 3.2715,
+      "step": 603
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.771241624950818e-06,
+      "loss": 3.2274,
+      "step": 604
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.742963450273016e-06,
+      "loss": 3.1811,
+      "step": 605
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.7148073433362732e-06,
+      "loss": 3.3272,
+      "step": 606
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.6867737777433276e-06,
+      "loss": 3.2594,
+      "step": 607
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.658863225035695e-06,
+      "loss": 3.3312,
+      "step": 608
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.6310761546857433e-06,
+      "loss": 3.3856,
+      "step": 609
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.60341303408879e-06,
+      "loss": 3.4417,
+      "step": 610
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.5758743285552435e-06,
+      "loss": 3.4408,
+      "step": 611
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.5484605013027787e-06,
+      "loss": 3.147,
+      "step": 612
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.521172013448534e-06,
+      "loss": 3.1479,
+      "step": 613
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.494009324001372e-06,
+      "loss": 3.1448,
+      "step": 614
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.4669728898541456e-06,
+      "loss": 3.0974,
+      "step": 615
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.440063165776019e-06,
+      "loss": 3.1905,
+      "step": 616
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.4132806044048163e-06,
+      "loss": 3.4658,
+      "step": 617
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.3866256562394084e-06,
+      "loss": 3.2968,
+      "step": 618
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.3600987696321332e-06,
+      "loss": 3.3054,
+      "step": 619
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.333700390781256e-06,
+      "loss": 3.0319,
+      "step": 620
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.3074309637234702e-06,
+      "loss": 3.2635,
+      "step": 621
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.2812909303264084e-06,
+      "loss": 3.2026,
+      "step": 622
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.2552807302812353e-06,
+      "loss": 3.2623,
+      "step": 623
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.2294008010952383e-06,
+      "loss": 3.2525,
+      "step": 624
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.203651578084467e-06,
+      "loss": 3.2043,
+      "step": 625
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.178033494366416e-06,
+      "loss": 3.1327,
+      "step": 626
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.1525469808527376e-06,
+      "loss": 3.1885,
+      "step": 627
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.127192466241994e-06,
+      "loss": 3.2633,
+      "step": 628
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.1019703770124454e-06,
+      "loss": 3.124,
+      "step": 629
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.07688113741488e-06,
+      "loss": 3.3028,
+      "step": 630
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.0519251694654674e-06,
+      "loss": 3.1655,
+      "step": 631
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.027102892938674e-06,
+      "loss": 3.4117,
+      "step": 632
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.0024147253601957e-06,
+      "loss": 3.1954,
+      "step": 633
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.977861081999931e-06,
+      "loss": 3.1659,
+      "step": 634
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.9534423758650043e-06,
+      "loss": 3.1071,
+      "step": 635
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.92915901769281e-06,
+      "loss": 3.2877,
+      "step": 636
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.9050114159441135e-06,
+      "loss": 3.3129,
+      "step": 637
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.880999976796164e-06,
+      "loss": 3.3743,
+      "step": 638
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.8571251041358895e-06,
+      "loss": 3.129,
+      "step": 639
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.8333871995530728e-06,
+      "loss": 3.2712,
+      "step": 640
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.809786662333619e-06,
+      "loss": 3.1361,
+      "step": 641
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.786323889452828e-06,
+      "loss": 3.2079,
+      "step": 642
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.762999275568721e-06,
+      "loss": 3.2124,
+      "step": 643
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.739813213015401e-06,
+      "loss": 3.328,
+      "step": 644
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7167660917964557e-06,
+      "loss": 3.1001,
+      "step": 645
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.6938582995783958e-06,
+      "loss": 3.3397,
+      "step": 646
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.6710902216841241e-06,
+      "loss": 3.2774,
+      "step": 647
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.6484622410864837e-06,
+      "loss": 3.3643,
+      "step": 648
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.6259747384017766e-06,
+      "loss": 3.2594,
+      "step": 649
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.6036280918833924e-06,
+      "loss": 3.1878,
+      "step": 650
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.5814226774154328e-06,
+      "loss": 3.2801,
+      "step": 651
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.5593588685063899e-06,
+      "loss": 3.1553,
+      "step": 652
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.5374370362828662e-06,
+      "loss": 3.1934,
+      "step": 653
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.515657549483328e-06,
+      "loss": 3.1606,
+      "step": 654
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.494020774451912e-06,
+      "loss": 3.2433,
+      "step": 655
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.472527075132245e-06,
+      "loss": 3.2234,
+      "step": 656
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4511768130613434e-06,
+      "loss": 3.1697,
+      "step": 657
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4299703473635217e-06,
+      "loss": 3.3123,
+      "step": 658
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4089080347443485e-06,
+      "loss": 3.2725,
+      "step": 659
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.3879902294846537e-06,
+      "loss": 3.2027,
+      "step": 660
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.3672172834345632e-06,
+      "loss": 3.2765,
+      "step": 661
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.3465895460075872e-06,
+      "loss": 3.1554,
+      "step": 662
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.3261073641747358e-06,
+      "loss": 3.2031,
+      "step": 663
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.30577108245869e-06,
+      "loss": 3.2662,
+      "step": 664
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.2855810429279958e-06,
+      "loss": 3.2323,
+      "step": 665
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.2655375851913232e-06,
+      "loss": 3.2345,
+      "step": 666
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.2456410463917445e-06,
+      "loss": 3.1215,
+      "step": 667
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.22589176120107e-06,
+      "loss": 3.3559,
+      "step": 668
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.2062900618142136e-06,
+      "loss": 3.3381,
+      "step": 669
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.186836277943606e-06,
+      "loss": 3.1843,
+      "step": 670
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1675307368136513e-06,
+      "loss": 3.1898,
+      "step": 671
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1483737631552161e-06,
+      "loss": 3.1972,
+      "step": 672
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1293656792001817e-06,
+      "loss": 3.2011,
+      "step": 673
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1105068046760048e-06,
+      "loss": 3.2577,
+      "step": 674
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.0917974568003552e-06,
+      "loss": 3.2048,
+      "step": 675
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.0732379502757717e-06,
+      "loss": 3.1821,
+      "step": 676
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.054828597284372e-06,
+      "loss": 3.2322,
+      "step": 677
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.036569707482602e-06,
+      "loss": 3.3726,
+      "step": 678
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.0184615879960224e-06,
+      "loss": 3.1546,
+      "step": 679
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.0005045434141503e-06,
+      "loss": 3.2868,
+      "step": 680
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 9.826988757853228e-07,
+      "loss": 3.2473,
+      "step": 681
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 9.6504488461164e-07,
+      "loss": 3.2338,
+      "step": 682
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 9.47542866843899e-07,
+      "loss": 3.3132,
+      "step": 683
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 9.301931168766165e-07,
+      "loss": 3.266,
+      "step": 684
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 9.129959265430766e-07,
+      "loss": 3.2878,
+      "step": 685
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.959515851104117e-07,
+      "loss": 3.1873,
+      "step": 686
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.790603792747499e-07,
+      "loss": 3.0949,
+      "step": 687
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.623225931563806e-07,
+      "loss": 3.1953,
+      "step": 688
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.457385082949842e-07,
+      "loss": 3.1766,
+      "step": 689
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.293084036448895e-07,
+      "loss": 3.3156,
+      "step": 690
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.130325555703911e-07,
+      "loss": 3.1429,
+      "step": 691
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.969112378410882e-07,
+      "loss": 3.2375,
+      "step": 692
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.809447216272892e-07,
+      "loss": 3.2285,
+      "step": 693
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.651332754954477e-07,
+      "loss": 3.2257,
+      "step": 694
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.494771654036448e-07,
+      "loss": 3.1945,
+      "step": 695
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.33976654697115e-07,
+      "loss": 3.1229,
+      "step": 696
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.186320041038175e-07,
+      "loss": 3.2002,
+      "step": 697
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.034434717300509e-07,
+      "loss": 3.1972,
+      "step": 698
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.884113130561043e-07,
+      "loss": 3.0296,
+      "step": 699
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.735357809319809e-07,
+      "loss": 3.3183,
+      "step": 700
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.588171255731157e-07,
+      "loss": 3.2142,
+      "step": 701
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.442555945561901e-07,
+      "loss": 3.1247,
+      "step": 702
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.298514328149574e-07,
+      "loss": 3.2135,
+      "step": 703
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.156048826361239e-07,
+      "loss": 3.2563,
+      "step": 704
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.015161836552764e-07,
+      "loss": 3.1705,
+      "step": 705
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.875855728528412e-07,
+      "loss": 3.3331,
+      "step": 706
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.738132845501199e-07,
+      "loss": 3.2448,
+      "step": 707
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.601995504053193e-07,
+      "loss": 3.2878,
+      "step": 708
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.467445994096821e-07,
+      "loss": 3.2399,
+      "step": 709
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.334486578836118e-07,
+      "loss": 3.3755,
+      "step": 710
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.203119494728826e-07,
+      "loss": 3.2725,
+      "step": 711
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.073346951448699e-07,
+      "loss": 3.3156,
+      "step": 712
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.945171131848358e-07,
+      "loss": 3.2566,
+      "step": 713
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.818594191922577e-07,
+      "loss": 3.2779,
+      "step": 714
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.6936182607719373e-07,
+      "loss": 3.1843,
+      "step": 715
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.5702454405672004e-07,
+      "loss": 3.2624,
+      "step": 716
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.448477806513729e-07,
+      "loss": 3.2075,
+      "step": 717
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.328317406816751e-07,
+      "loss": 3.2446,
+      "step": 718
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.2097662626468085e-07,
+      "loss": 3.1551,
+      "step": 719
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.0928263681057956e-07,
+      "loss": 3.2224,
+      "step": 720
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.97749969019342e-07,
+      "loss": 3.2041,
+      "step": 721
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.863788168774119e-07,
+      "loss": 3.2021,
+      "step": 722
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.7516937165444136e-07,
+      "loss": 3.2621,
+      "step": 723
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.6412182190007086e-07,
+      "loss": 3.3452,
+      "step": 724
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.5323635344077123e-07,
+      "loss": 3.1605,
+      "step": 725
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.4251314937669313e-07,
+      "loss": 3.2814,
+      "step": 726
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.3195239007861815e-07,
+      "loss": 3.2987,
+      "step": 727
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.2155425318489584e-07,
+      "loss": 3.1901,
+      "step": 728
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.1131891359847397e-07,
+      "loss": 3.3101,
+      "step": 729
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.012465434839529e-07,
+      "loss": 3.0684,
+      "step": 730
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.913373122646845e-07,
+      "loss": 3.1588,
+      "step": 731
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.8159138661992824e-07,
+      "loss": 3.2064,
+      "step": 732
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.720089304820417e-07,
+      "loss": 3.1514,
+      "step": 733
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.6259010503373206e-07,
+      "loss": 3.3542,
+      "step": 734
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.533350687053338e-07,
+      "loss": 3.1581,
+      "step": 735
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.442439771721539e-07,
+      "loss": 3.217,
+      "step": 736
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.353169833518454e-07,
+      "loss": 3.1335,
+      "step": 737
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.2655423740183925e-07,
+      "loss": 3.122,
+      "step": 738
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.179558867168219e-07,
+      "loss": 3.0938,
+      "step": 739
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.0952207592624508e-07,
+      "loss": 3.1366,
+      "step": 740
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.0125294689190666e-07,
+      "loss": 3.1353,
+      "step": 741
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.9314863870555257e-07,
+      "loss": 3.208,
+      "step": 742
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.8520928768654745e-07,
+      "loss": 3.1476,
+      "step": 743
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.7743502737957107e-07,
+      "loss": 3.0601,
+      "step": 744
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6982598855238564e-07,
+      "loss": 3.238,
+      "step": 745
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6238229919361858e-07,
+      "loss": 3.1862,
+      "step": 746
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.5510408451062552e-07,
+      "loss": 3.1664,
+      "step": 747
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.4799146692737742e-07,
+      "loss": 3.2697,
+      "step": 748
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.410445660823989e-07,
+      "loss": 3.149,
+      "step": 749
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.3426349882676326e-07,
+      "loss": 3.2801,
+      "step": 750
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.2764837922211947e-07,
+      "loss": 3.2374,
+      "step": 751
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.211993185387772e-07,
+      "loss": 3.2966,
+      "step": 752
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.1491642525383595e-07,
+      "loss": 3.2834,
+      "step": 753
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0879980504935772e-07,
+      "loss": 3.1645,
+      "step": 754
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.0284956081059171e-07,
+      "loss": 3.1979,
+      "step": 755
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 9.706579262424132e-08,
+      "loss": 3.2138,
+      "step": 756
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 9.144859777678539e-08,
+      "loss": 3.284,
+      "step": 757
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.599807075283406e-08,
+      "loss": 3.1706,
+      "step": 758
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.071430323354778e-08,
+      "loss": 3.354,
+      "step": 759
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.559738409508854e-08,
+      "loss": 3.2734,
+      "step": 760
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.064739940713217e-08,
+      "loss": 3.1273,
+      "step": 761
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 6.586443243140839e-08,
+      "loss": 3.2327,
+      "step": 762
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.12485636203164e-08,
+      "loss": 3.2064,
+      "step": 763
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.679987061555703e-08,
+      "loss": 3.2127,
+      "step": 764
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.251842824683717e-08,
+      "loss": 3.1752,
+      "step": 765
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.840430853060518e-08,
+      "loss": 3.198,
+      "step": 766
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.4457580668839653e-08,
+      "loss": 3.1497,
+      "step": 767
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.067831104789033e-08,
+      "loss": 3.2571,
+      "step": 768
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.706656323735569e-08,
+      "loss": 3.1735,
+      "step": 769
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.362239798901712e-08,
+      "loss": 3.1884,
+      "step": 770
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.034587323581639e-08,
+      "loss": 3.1669,
+      "step": 771
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.7237044090879795e-08,
+      "loss": 3.2778,
+      "step": 772
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.429596284659219e-08,
+      "loss": 3.2229,
+      "step": 773
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.1522678973718848e-08,
+      "loss": 3.1956,
+      "step": 774
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.891723912056942e-08,
+      "loss": 3.074,
+      "step": 775
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.6479687112217478e-08,
+      "loss": 3.1022,
+      "step": 776
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.421006394976221e-08,
+      "loss": 3.268,
+      "step": 777
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.2108407809635624e-08,
+      "loss": 3.2414,
+      "step": 778
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.0174754042964197e-08,
+      "loss": 3.1635,
+      "step": 779
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 8.40913517497377e-09,
+      "loss": 3.3233,
+      "step": 780
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 6.8115809044411174e-09,
+      "loss": 3.2068,
+      "step": 781
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.3821181031932235e-09,
+      "loss": 3.1507,
+      "step": 782
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.120770815659869e-09,
+      "loss": 3.2037,
+      "step": 783
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.0275602584650677e-09,
+      "loss": 3.1867,
+      "step": 784
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.102504820069573e-09,
+      "loss": 3.2221,
+      "step": 785
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.345620060465569e-09,
+      "loss": 3.1024,
+      "step": 786
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.569187109124354e-10,
+      "loss": 3.1997,
+      "step": 787
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3641067372358617e-10,
+      "loss": 3.1689,
+      "step": 788
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.410302209660437e-11,
+      "loss": 3.1614,
+      "step": 789
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0,
+      "loss": 3.105,
+      "step": 790
+    },
+    {
+      "epoch": 1.0,
+      "step": 790,
+      "total_flos": 261024117719040.0,
+      "train_loss": 4.121155046209505,
+      "train_runtime": 3165.0402,
+      "train_samples_per_second": 31.949,
+      "train_steps_per_second": 0.25
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 790,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 50000,
+  "total_flos": 261024117719040.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}