diff --git "a/checkpoint-1524/trainer_state.json" "b/checkpoint-1524/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1524/trainer_state.json"
@@ -0,0 +1,9169 @@
+{
+  "best_metric": 0.03057965449988842,
+  "best_model_checkpoint": "./phishing-email_sender-detection/checkpoint-1524",
+  "epoch": 1.0,
+  "global_step": 1524,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.7203,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.6799,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3e-06,
+      "loss": 0.6674,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.7183,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-06,
+      "loss": 0.6762,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6e-06,
+      "loss": 0.6905,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 0.7062,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.667,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9e-06,
+      "loss": 0.678,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1e-05,
+      "loss": 0.6763,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 0.6819,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2e-05,
+      "loss": 0.6763,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 0.6806,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 0.6705,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5e-05,
+      "loss": 0.652,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.6493,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 0.6686,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8e-05,
+      "loss": 0.6655,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9e-05,
+      "loss": 0.6223,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2e-05,
+      "loss": 0.6469,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1e-05,
+      "loss": 0.6166,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 0.5773,
+      "step": 22
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 0.5797,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4e-05,
+      "loss": 0.5119,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5e-05,
+      "loss": 0.4348,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 0.4558,
+      "step": 26
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 0.4145,
+      "step": 27
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 0.3593,
+      "step": 28
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9e-05,
+      "loss": 0.1727,
+      "step": 29
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3e-05,
+      "loss": 0.2,
+      "step": 30
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1e-05,
+      "loss": 0.3584,
+      "step": 31
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 0.1854,
+      "step": 32
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3e-05,
+      "loss": 0.1301,
+      "step": 33
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 0.1258,
+      "step": 34
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5e-05,
+      "loss": 0.2131,
+      "step": 35
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6e-05,
+      "loss": 0.0449,
+      "step": 36
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7e-05,
+      "loss": 0.1951,
+      "step": 37
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8e-05,
+      "loss": 0.078,
+      "step": 38
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 0.021,
+      "step": 39
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4e-05,
+      "loss": 0.3168,
+      "step": 40
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.1e-05,
+      "loss": 0.8099,
+      "step": 41
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.2e-05,
+      "loss": 0.2544,
+      "step": 42
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.3e-05,
+      "loss": 0.0511,
+      "step": 43
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.0087,
+      "step": 44
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.5e-05,
+      "loss": 0.0393,
+      "step": 45
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.1003,
+      "step": 46
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.7e-05,
+      "loss": 0.3249,
+      "step": 47
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8e-05,
+      "loss": 0.0075,
+      "step": 48
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9e-05,
+      "loss": 0.0109,
+      "step": 49
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 0.0183,
+      "step": 50
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.998332221480987e-05,
+      "loss": 0.0036,
+      "step": 51
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.996664442961975e-05,
+      "loss": 0.3416,
+      "step": 52
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.994996664442963e-05,
+      "loss": 0.6667,
+      "step": 53
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.99332888592395e-05,
+      "loss": 0.035,
+      "step": 54
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.991661107404937e-05,
+      "loss": 0.0066,
+      "step": 55
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.989993328885924e-05,
+      "loss": 0.5418,
+      "step": 56
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9883255503669115e-05,
+      "loss": 0.6135,
+      "step": 57
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.986657771847899e-05,
+      "loss": 1.7134,
+      "step": 58
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.984989993328886e-05,
+      "loss": 0.3707,
+      "step": 59
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.983322214809873e-05,
+      "loss": 0.0196,
+      "step": 60
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.981654436290861e-05,
+      "loss": 0.2132,
+      "step": 61
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.979986657771848e-05,
+      "loss": 0.506,
+      "step": 62
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.978318879252836e-05,
+      "loss": 0.0718,
+      "step": 63
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.976651100733823e-05,
+      "loss": 0.0936,
+      "step": 64
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.97498332221481e-05,
+      "loss": 0.3546,
+      "step": 65
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9733155436957975e-05,
+      "loss": 0.0692,
+      "step": 66
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9716477651767846e-05,
+      "loss": 0.0637,
+      "step": 67
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.969979986657772e-05,
+      "loss": 0.3302,
+      "step": 68
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.968312208138759e-05,
+      "loss": 0.0704,
+      "step": 69
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.966644429619747e-05,
+      "loss": 0.0124,
+      "step": 70
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.964976651100734e-05,
+      "loss": 0.2998,
+      "step": 71
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.963308872581721e-05,
+      "loss": 0.0216,
+      "step": 72
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.961641094062709e-05,
+      "loss": 0.5102,
+      "step": 73
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.959973315543696e-05,
+      "loss": 0.6388,
+      "step": 74
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9583055370246836e-05,
+      "loss": 0.1079,
+      "step": 75
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9566377585056706e-05,
+      "loss": 0.0136,
+      "step": 76
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9549699799866576e-05,
+      "loss": 0.0061,
+      "step": 77
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9533022014676454e-05,
+      "loss": 0.0071,
+      "step": 78
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.951634422948633e-05,
+      "loss": 0.4773,
+      "step": 79
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.94996664442962e-05,
+      "loss": 0.412,
+      "step": 80
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.948298865910607e-05,
+      "loss": 0.0046,
+      "step": 81
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.946631087391594e-05,
+      "loss": 0.0038,
+      "step": 82
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.944963308872582e-05,
+      "loss": 0.003,
+      "step": 83
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9432955303535696e-05,
+      "loss": 0.2076,
+      "step": 84
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9416277518345566e-05,
+      "loss": 0.0028,
+      "step": 85
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.939959973315544e-05,
+      "loss": 0.0024,
+      "step": 86
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.938292194796531e-05,
+      "loss": 0.0458,
+      "step": 87
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.936624416277519e-05,
+      "loss": 0.002,
+      "step": 88
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.934956637758506e-05,
+      "loss": 0.3422,
+      "step": 89
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.933288859239493e-05,
+      "loss": 0.0084,
+      "step": 90
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.93162108072048e-05,
+      "loss": 0.5222,
+      "step": 91
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.929953302201468e-05,
+      "loss": 0.2925,
+      "step": 92
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9282855236824556e-05,
+      "loss": 0.1514,
+      "step": 93
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.926617745163443e-05,
+      "loss": 0.0019,
+      "step": 94
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.92494996664443e-05,
+      "loss": 0.0017,
+      "step": 95
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9232821881254174e-05,
+      "loss": 0.0011,
+      "step": 96
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9216144096064044e-05,
+      "loss": 0.0019,
+      "step": 97
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.919946631087392e-05,
+      "loss": 0.013,
+      "step": 98
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.918278852568379e-05,
+      "loss": 0.4217,
+      "step": 99
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.916611074049366e-05,
+      "loss": 0.0014,
+      "step": 100
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.914943295530354e-05,
+      "loss": 0.0357,
+      "step": 101
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.913275517011341e-05,
+      "loss": 0.0094,
+      "step": 102
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.911607738492329e-05,
+      "loss": 0.002,
+      "step": 103
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.909939959973316e-05,
+      "loss": 0.0018,
+      "step": 104
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.9082721814543034e-05,
+      "loss": 0.0563,
+      "step": 105
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.9066044029352905e-05,
+      "loss": 0.0685,
+      "step": 106
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.9049366244162775e-05,
+      "loss": 0.002,
+      "step": 107
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.903268845897265e-05,
+      "loss": 0.1449,
+      "step": 108
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.901601067378252e-05,
+      "loss": 0.3448,
+      "step": 109
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.89993328885924e-05,
+      "loss": 0.2631,
+      "step": 110
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.898265510340227e-05,
+      "loss": 0.0241,
+      "step": 111
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.896597731821214e-05,
+      "loss": 0.421,
+      "step": 112
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.894929953302202e-05,
+      "loss": 0.4811,
+      "step": 113
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.8932621747831895e-05,
+      "loss": 0.0087,
+      "step": 114
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8915943962641765e-05,
+      "loss": 0.003,
+      "step": 115
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8899266177451635e-05,
+      "loss": 0.0043,
+      "step": 116
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8882588392261506e-05,
+      "loss": 0.2751,
+      "step": 117
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.886591060707138e-05,
+      "loss": 0.0169,
+      "step": 118
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.884923282188126e-05,
+      "loss": 0.002,
+      "step": 119
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.883255503669113e-05,
+      "loss": 0.0018,
+      "step": 120
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8815877251501e-05,
+      "loss": 0.002,
+      "step": 121
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.879919946631087e-05,
+      "loss": 0.1121,
+      "step": 122
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8782521681120755e-05,
+      "loss": 0.0014,
+      "step": 123
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8765843895930625e-05,
+      "loss": 0.093,
+      "step": 124
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8749166110740496e-05,
+      "loss": 0.0013,
+      "step": 125
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8732488325550366e-05,
+      "loss": 0.7647,
+      "step": 126
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.871581054036024e-05,
+      "loss": 1.2515,
+      "step": 127
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.869913275517012e-05,
+      "loss": 0.0031,
+      "step": 128
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.868245496997999e-05,
+      "loss": 0.3657,
+      "step": 129
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.866577718478986e-05,
+      "loss": 0.002,
+      "step": 130
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.864909939959973e-05,
+      "loss": 0.0031,
+      "step": 131
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.863242161440961e-05,
+      "loss": 0.0024,
+      "step": 132
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8615743829219486e-05,
+      "loss": 0.2674,
+      "step": 133
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8599066044029356e-05,
+      "loss": 0.0016,
+      "step": 134
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8582388258839226e-05,
+      "loss": 0.002,
+      "step": 135
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8565710473649103e-05,
+      "loss": 0.0368,
+      "step": 136
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8549032688458974e-05,
+      "loss": 0.0017,
+      "step": 137
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.853235490326885e-05,
+      "loss": 0.1991,
+      "step": 138
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.851567711807872e-05,
+      "loss": 0.0021,
+      "step": 139
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.849899933288859e-05,
+      "loss": 0.0036,
+      "step": 140
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.848232154769847e-05,
+      "loss": 0.0019,
+      "step": 141
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.846564376250834e-05,
+      "loss": 0.002,
+      "step": 142
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8448965977318216e-05,
+      "loss": 0.4211,
+      "step": 143
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8432288192128087e-05,
+      "loss": 0.0073,
+      "step": 144
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8415610406937964e-05,
+      "loss": 0.1416,
+      "step": 145
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8398932621747834e-05,
+      "loss": 0.0017,
+      "step": 146
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8382254836557704e-05,
+      "loss": 0.5239,
+      "step": 147
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.836557705136758e-05,
+      "loss": 0.6471,
+      "step": 148
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.834889926617746e-05,
+      "loss": 0.2049,
+      "step": 149
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.833222148098733e-05,
+      "loss": 0.0104,
+      "step": 150
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.83155436957972e-05,
+      "loss": 0.0021,
+      "step": 151
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.829886591060707e-05,
+      "loss": 0.0024,
+      "step": 152
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.828218812541695e-05,
+      "loss": 0.0015,
+      "step": 153
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8265510340226824e-05,
+      "loss": 0.0019,
+      "step": 154
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8248832555036694e-05,
+      "loss": 0.0476,
+      "step": 155
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8232154769846565e-05,
+      "loss": 0.0024,
+      "step": 156
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8215476984656435e-05,
+      "loss": 0.3098,
+      "step": 157
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.819879919946632e-05,
+      "loss": 0.2202,
+      "step": 158
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.818212141427619e-05,
+      "loss": 0.0182,
+      "step": 159
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.816544362908606e-05,
+      "loss": 0.0052,
+      "step": 160
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.814876584389593e-05,
+      "loss": 0.0177,
+      "step": 161
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.813208805870581e-05,
+      "loss": 0.7179,
+      "step": 162
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8115410273515684e-05,
+      "loss": 0.2988,
+      "step": 163
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8098732488325555e-05,
+      "loss": 0.2633,
+      "step": 164
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8082054703135425e-05,
+      "loss": 0.1607,
+      "step": 165
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8065376917945295e-05,
+      "loss": 0.0518,
+      "step": 166
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.804869913275517e-05,
+      "loss": 0.2721,
+      "step": 167
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.803202134756505e-05,
+      "loss": 0.0075,
+      "step": 168
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.801534356237492e-05,
+      "loss": 0.0078,
+      "step": 169
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.799866577718479e-05,
+      "loss": 0.115,
+      "step": 170
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.798198799199467e-05,
+      "loss": 0.3379,
+      "step": 171
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.796531020680454e-05,
+      "loss": 0.3996,
+      "step": 172
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.7948632421614415e-05,
+      "loss": 0.0749,
+      "step": 173
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.7931954636424285e-05,
+      "loss": 0.2114,
+      "step": 174
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.7915276851234156e-05,
+      "loss": 0.125,
+      "step": 175
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.789859906604403e-05,
+      "loss": 0.1506,
+      "step": 176
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.78819212808539e-05,
+      "loss": 0.0318,
+      "step": 177
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.786524349566378e-05,
+      "loss": 0.0569,
+      "step": 178
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.784856571047365e-05,
+      "loss": 0.0163,
+      "step": 179
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.783188792528353e-05,
+      "loss": 0.0328,
+      "step": 180
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.78152101400934e-05,
+      "loss": 0.2458,
+      "step": 181
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.779853235490327e-05,
+      "loss": 0.0094,
+      "step": 182
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.7781854569713146e-05,
+      "loss": 0.1288,
+      "step": 183
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.7765176784523016e-05,
+      "loss": 0.593,
+      "step": 184
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.774849899933289e-05,
+      "loss": 0.0335,
+      "step": 185
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.773182121414276e-05,
+      "loss": 0.0048,
+      "step": 186
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.7715143428952634e-05,
+      "loss": 0.0053,
+      "step": 187
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.769846564376251e-05,
+      "loss": 0.337,
+      "step": 188
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.768178785857239e-05,
+      "loss": 0.004,
+      "step": 189
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.766511007338226e-05,
+      "loss": 0.0022,
+      "step": 190
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.764843228819213e-05,
+      "loss": 0.0022,
+      "step": 191
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7631754503002e-05,
+      "loss": 0.0015,
+      "step": 192
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7615076717811876e-05,
+      "loss": 0.0032,
+      "step": 193
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.759839893262175e-05,
+      "loss": 0.2427,
+      "step": 194
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7581721147431624e-05,
+      "loss": 0.0014,
+      "step": 195
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7565043362241494e-05,
+      "loss": 0.1157,
+      "step": 196
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.754836557705137e-05,
+      "loss": 0.0014,
+      "step": 197
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.753168779186125e-05,
+      "loss": 0.0672,
+      "step": 198
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.751501000667112e-05,
+      "loss": 0.1869,
+      "step": 199
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.749833222148099e-05,
+      "loss": 0.6225,
+      "step": 200
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.748165443629086e-05,
+      "loss": 0.2323,
+      "step": 201
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7464976651100736e-05,
+      "loss": 0.4761,
+      "step": 202
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7448298865910614e-05,
+      "loss": 0.0235,
+      "step": 203
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7431621080720484e-05,
+      "loss": 0.0009,
+      "step": 204
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7414943295530354e-05,
+      "loss": 0.4716,
+      "step": 205
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.739826551034023e-05,
+      "loss": 1.6486,
+      "step": 206
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.73815877251501e-05,
+      "loss": 0.3304,
+      "step": 207
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.736490993995998e-05,
+      "loss": 0.3394,
+      "step": 208
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.734823215476985e-05,
+      "loss": 0.1649,
+      "step": 209
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.733155436957972e-05,
+      "loss": 0.0401,
+      "step": 210
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.73148765843896e-05,
+      "loss": 0.0541,
+      "step": 211
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.729819879919947e-05,
+      "loss": 0.0255,
+      "step": 212
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7281521014009344e-05,
+      "loss": 0.0047,
+      "step": 213
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7264843228819215e-05,
+      "loss": 0.015,
+      "step": 214
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.724816544362909e-05,
+      "loss": 0.4996,
+      "step": 215
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.723148765843896e-05,
+      "loss": 0.0009,
+      "step": 216
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.721480987324883e-05,
+      "loss": 0.4411,
+      "step": 217
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.719813208805871e-05,
+      "loss": 0.1461,
+      "step": 218
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.718145430286858e-05,
+      "loss": 0.001,
+      "step": 219
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.716477651767846e-05,
+      "loss": 0.0067,
+      "step": 220
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.714809873248833e-05,
+      "loss": 0.5534,
+      "step": 221
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.71314209472982e-05,
+      "loss": 0.0034,
+      "step": 222
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7114743162108075e-05,
+      "loss": 0.0172,
+      "step": 223
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.709806537691795e-05,
+      "loss": 0.0132,
+      "step": 224
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.708138759172782e-05,
+      "loss": 0.0033,
+      "step": 225
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.706470980653769e-05,
+      "loss": 0.0016,
+      "step": 226
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.704803202134756e-05,
+      "loss": 0.0021,
+      "step": 227
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.703135423615744e-05,
+      "loss": 0.0025,
+      "step": 228
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.701467645096732e-05,
+      "loss": 0.0022,
+      "step": 229
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.699799866577719e-05,
+      "loss": 0.0014,
+      "step": 230
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.698132088058706e-05,
+      "loss": 0.0013,
+      "step": 231
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.6964643095396935e-05,
+      "loss": 0.0009,
+      "step": 232
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.694796531020681e-05,
+      "loss": 0.004,
+      "step": 233
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.693128752501668e-05,
+      "loss": 0.0011,
+      "step": 234
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.691460973982655e-05,
+      "loss": 0.0011,
+      "step": 235
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.689793195463642e-05,
+      "loss": 0.0009,
+      "step": 236
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.68812541694463e-05,
+      "loss": 0.0015,
+      "step": 237
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.686457638425617e-05,
+      "loss": 0.1971,
+      "step": 238
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.684789859906605e-05,
+      "loss": 0.4847,
+      "step": 239
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.683122081387592e-05,
+      "loss": 0.0008,
+      "step": 240
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.6814543028685795e-05,
+      "loss": 0.0882,
+      "step": 241
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.6797865243495666e-05,
+      "loss": 0.0008,
+      "step": 242
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.6781187458305536e-05,
+      "loss": 0.0008,
+      "step": 243
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.676450967311541e-05,
+      "loss": 0.6945,
+      "step": 244
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.6747831887925284e-05,
+      "loss": 0.001,
+      "step": 245
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.673115410273516e-05,
+      "loss": 0.0163,
+      "step": 246
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.671447631754503e-05,
+      "loss": 0.3834,
+      "step": 247
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.66977985323549e-05,
+      "loss": 0.0014,
+      "step": 248
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.668112074716478e-05,
+      "loss": 0.2053,
+      "step": 249
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.6664442961974656e-05,
+      "loss": 0.0035,
+      "step": 250
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.6647765176784526e-05,
+      "loss": 0.0102,
+      "step": 251
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.6631087391594396e-05,
+      "loss": 0.0016,
+      "step": 252
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.661440960640427e-05,
+      "loss": 0.0026,
+      "step": 253
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.6597731821214144e-05,
+      "loss": 0.0021,
+      "step": 254
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.658105403602402e-05,
+      "loss": 0.0012,
+      "step": 255
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.656437625083389e-05,
+      "loss": 0.1572,
+      "step": 256
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.654769846564376e-05,
+      "loss": 0.0337,
+      "step": 257
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.653102068045363e-05,
+      "loss": 0.0014,
+      "step": 258
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.6514342895263516e-05,
+      "loss": 0.2481,
+      "step": 259
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.6497665110073386e-05,
+      "loss": 0.2999,
+      "step": 260
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.648098732488326e-05,
+      "loss": 0.001,
+      "step": 261
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.646430953969313e-05,
+      "loss": 0.0117,
+      "step": 262
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.6447631754503004e-05,
+      "loss": 0.0016,
+      "step": 263
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.643095396931288e-05,
+      "loss": 0.0021,
+      "step": 264
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.641427618412275e-05,
+      "loss": 0.0014,
+      "step": 265
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.639759839893262e-05,
+      "loss": 0.1078,
+      "step": 266
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.638092061374249e-05,
+      "loss": 0.0009,
+      "step": 267
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.636424282855237e-05,
+      "loss": 0.0022,
+      "step": 268
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.6347565043362247e-05,
+      "loss": 0.2706,
+      "step": 269
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.633088725817212e-05,
+      "loss": 0.031,
+      "step": 270
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.631420947298199e-05,
+      "loss": 0.0008,
+      "step": 271
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.6297531687791864e-05,
+      "loss": 0.0029,
+      "step": 272
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.6280853902601735e-05,
+      "loss": 0.0007,
+      "step": 273
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.626417611741161e-05,
+      "loss": 0.4306,
+      "step": 274
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.624749833222148e-05,
+      "loss": 0.0017,
+      "step": 275
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.623082054703136e-05,
+      "loss": 0.262,
+      "step": 276
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.621414276184123e-05,
+      "loss": 0.0011,
+      "step": 277
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.61974649766511e-05,
+      "loss": 0.0268,
+      "step": 278
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.618078719146098e-05,
+      "loss": 0.2662,
+      "step": 279
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.616410940627085e-05,
+      "loss": 0.0019,
+      "step": 280
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.6147431621080725e-05,
+      "loss": 0.0084,
+      "step": 281
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.6130753835890595e-05,
+      "loss": 0.1269,
+      "step": 282
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.6114076050700465e-05,
+      "loss": 0.0009,
+      "step": 283
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.609739826551034e-05,
+      "loss": 0.0007,
+      "step": 284
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.608072048032022e-05,
+      "loss": 0.0108,
+      "step": 285
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.606404269513009e-05,
+      "loss": 0.0005,
+      "step": 286
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.604736490993996e-05,
+      "loss": 0.0032,
+      "step": 287
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.603068712474983e-05,
+      "loss": 0.0005,
+      "step": 288
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.601400933955971e-05,
+      "loss": 0.0004,
+      "step": 289
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.5997331554369585e-05,
+      "loss": 0.1748,
+      "step": 290
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.5980653769179455e-05,
+      "loss": 0.4749,
+      "step": 291
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.5963975983989326e-05,
+      "loss": 0.0014,
+      "step": 292
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.5947298198799196e-05,
+      "loss": 0.0004,
+      "step": 293
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.593062041360908e-05,
+      "loss": 0.0004,
+      "step": 294
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.591394262841895e-05,
+      "loss": 0.4376,
+      "step": 295
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.589726484322882e-05,
+      "loss": 0.1717,
+      "step": 296
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.588058705803869e-05,
+      "loss": 0.4225,
+      "step": 297
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.586390927284857e-05,
+      "loss": 0.0007,
+      "step": 298
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.5847231487658445e-05,
+      "loss": 0.5234,
+      "step": 299
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.5830553702468316e-05,
+      "loss": 0.0007,
+      "step": 300
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.5813875917278186e-05,
+      "loss": 0.2654,
+      "step": 301
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.5797198132088056e-05,
+      "loss": 0.0876,
+      "step": 302
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.5780520346897933e-05,
+      "loss": 0.3373,
+      "step": 303
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.576384256170781e-05,
+      "loss": 0.0173,
+      "step": 304
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.574716477651768e-05,
+      "loss": 0.005,
+      "step": 305
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.573048699132755e-05,
+      "loss": 0.0069,
+      "step": 306
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.571380920613743e-05,
+      "loss": 0.0032,
+      "step": 307
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.56971314209473e-05,
+      "loss": 0.4767,
+      "step": 308
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.5680453635757176e-05,
+      "loss": 0.0925,
+      "step": 309
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.5663775850567046e-05,
+      "loss": 0.003,
+      "step": 310
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.5647098065376917e-05,
+      "loss": 0.0436,
+      "step": 311
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.5630420280186794e-05,
+      "loss": 0.0016,
+      "step": 312
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.5613742494996664e-05,
+      "loss": 0.0231,
+      "step": 313
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.559706470980654e-05,
+      "loss": 0.0939,
+      "step": 314
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.558038692461641e-05,
+      "loss": 0.0022,
+      "step": 315
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.556370913942629e-05,
+      "loss": 0.0321,
+      "step": 316
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.554703135423616e-05,
+      "loss": 0.01,
+      "step": 317
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.553035356904603e-05,
+      "loss": 0.0011,
+      "step": 318
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.5513675783855907e-05,
+      "loss": 0.0021,
+      "step": 319
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.549699799866578e-05,
+      "loss": 0.0069,
+      "step": 320
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.5480320213475654e-05,
+      "loss": 0.0144,
+      "step": 321
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.5463642428285524e-05,
+      "loss": 0.0009,
+      "step": 322
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.5446964643095395e-05,
+      "loss": 0.0006,
+      "step": 323
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.543028685790527e-05,
+      "loss": 0.5697,
+      "step": 324
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.541360907271515e-05,
+      "loss": 0.4346,
+      "step": 325
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.539693128752502e-05,
+      "loss": 0.0039,
+      "step": 326
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.538025350233489e-05,
+      "loss": 0.001,
+      "step": 327
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.536357571714476e-05,
+      "loss": 0.3923,
+      "step": 328
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.534689793195464e-05,
+      "loss": 0.0006,
+      "step": 329
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.5330220146764514e-05,
+      "loss": 0.4813,
+      "step": 330
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.5313542361574385e-05,
+      "loss": 0.0009,
+      "step": 331
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.5296864576384255e-05,
+      "loss": 0.1973,
+      "step": 332
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.528018679119413e-05,
+      "loss": 0.0014,
+      "step": 333
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.526350900600401e-05,
+      "loss": 0.3445,
+      "step": 334
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.524683122081388e-05,
+      "loss": 0.549,
+      "step": 335
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.523015343562375e-05,
+      "loss": 0.1901,
+      "step": 336
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.521347565043362e-05,
+      "loss": 0.0062,
+      "step": 337
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.51967978652435e-05,
+      "loss": 0.0079,
+      "step": 338
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.5180120080053375e-05,
+      "loss": 0.0111,
+      "step": 339
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.5163442294863245e-05,
+      "loss": 0.1573,
+      "step": 340
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.5146764509673115e-05,
+      "loss": 0.009,
+      "step": 341
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.513008672448299e-05,
+      "loss": 0.0105,
+      "step": 342
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.511340893929286e-05,
+      "loss": 0.0112,
+      "step": 343
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.509673115410274e-05,
+      "loss": 0.0114,
+      "step": 344
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.508005336891261e-05,
+      "loss": 0.0058,
+      "step": 345
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.506337558372248e-05,
+      "loss": 0.1183,
+      "step": 346
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.504669779853236e-05,
+      "loss": 0.0251,
+      "step": 347
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.503002001334223e-05,
+      "loss": 0.0023,
+      "step": 348
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.5013342228152105e-05,
+      "loss": 0.7293,
+      "step": 349
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.4996664442961976e-05,
+      "loss": 0.0017,
+      "step": 350
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.497998665777185e-05,
+      "loss": 0.3525,
+      "step": 351
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.496330887258172e-05,
+      "loss": 0.396,
+      "step": 352
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.494663108739159e-05,
+      "loss": 0.0018,
+      "step": 353
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.492995330220147e-05,
+      "loss": 0.0025,
+      "step": 354
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.491327551701134e-05,
+      "loss": 0.2422,
+      "step": 355
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.489659773182122e-05,
+      "loss": 0.5386,
+      "step": 356
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.487991994663109e-05,
+      "loss": 0.0061,
+      "step": 357
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.486324216144096e-05,
+      "loss": 0.0042,
+      "step": 358
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.4846564376250836e-05,
+      "loss": 0.1915,
+      "step": 359
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.482988659106071e-05,
+      "loss": 0.0021,
+      "step": 360
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.481320880587058e-05,
+      "loss": 0.0025,
+      "step": 361
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.4796531020680454e-05,
+      "loss": 0.3431,
+      "step": 362
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.4779853235490324e-05,
+      "loss": 0.0039,
+      "step": 363
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.47631754503002e-05,
+      "loss": 0.0014,
+      "step": 364
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.474649766511008e-05,
+      "loss": 0.0007,
+      "step": 365
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.472981987991995e-05,
+      "loss": 0.0019,
+      "step": 366
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.471314209472982e-05,
+      "loss": 0.0013,
+      "step": 367
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.4696464309539696e-05,
+      "loss": 0.259,
+      "step": 368
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.467978652434957e-05,
+      "loss": 0.0013,
+      "step": 369
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.4663108739159444e-05,
+      "loss": 0.0094,
+      "step": 370
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.4646430953969314e-05,
+      "loss": 0.001,
+      "step": 371
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.4629753168779184e-05,
+      "loss": 0.001,
+      "step": 372
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.461307538358906e-05,
+      "loss": 0.001,
+      "step": 373
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.459639759839894e-05,
+      "loss": 0.0008,
+      "step": 374
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.457971981320881e-05,
+      "loss": 0.0008,
+      "step": 375
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.456304202801868e-05,
+      "loss": 0.0025,
+      "step": 376
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.4546364242828556e-05,
+      "loss": 0.0852,
+      "step": 377
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.452968645763843e-05,
+      "loss": 0.0007,
+      "step": 378
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.4513008672448304e-05,
+      "loss": 0.0011,
+      "step": 379
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.4496330887258174e-05,
+      "loss": 0.001,
+      "step": 380
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.4479653102068045e-05,
+      "loss": 0.7787,
+      "step": 381
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.446297531687792e-05,
+      "loss": 0.1214,
+      "step": 382
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.444629753168779e-05,
+      "loss": 0.4831,
+      "step": 383
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.442961974649767e-05,
+      "loss": 0.0055,
+      "step": 384
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.441294196130754e-05,
+      "loss": 0.0025,
+      "step": 385
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.439626417611742e-05,
+      "loss": 0.3078,
+      "step": 386
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.437958639092729e-05,
+      "loss": 0.23,
+      "step": 387
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.436290860573716e-05,
+      "loss": 0.3858,
+      "step": 388
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.4346230820547034e-05,
+      "loss": 0.0956,
+      "step": 389
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.4329553035356905e-05,
+      "loss": 0.0262,
+      "step": 390
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.431287525016678e-05,
+      "loss": 0.13,
+      "step": 391
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.429619746497665e-05,
+      "loss": 0.25,
+      "step": 392
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.427951967978652e-05,
+      "loss": 0.2173,
+      "step": 393
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.42628418945964e-05,
+      "loss": 0.0115,
+      "step": 394
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.424616410940628e-05,
+      "loss": 0.0083,
+      "step": 395
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.422948632421615e-05,
+      "loss": 0.0148,
+      "step": 396
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.421280853902602e-05,
+      "loss": 0.35,
+      "step": 397
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.419613075383589e-05,
+      "loss": 0.2463,
+      "step": 398
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.4179452968645765e-05,
+      "loss": 0.0189,
+      "step": 399
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.416277518345564e-05,
+      "loss": 0.0039,
+      "step": 400
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.414609739826551e-05,
+      "loss": 0.002,
+      "step": 401
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.412941961307538e-05,
+      "loss": 0.3199,
+      "step": 402
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.411274182788525e-05,
+      "loss": 0.3847,
+      "step": 403
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.409606404269514e-05,
+      "loss": 0.0027,
+      "step": 404
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.407938625750501e-05,
+      "loss": 0.017,
+      "step": 405
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.406270847231488e-05,
+      "loss": 0.0036,
+      "step": 406
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.404603068712475e-05,
+      "loss": 0.0061,
+      "step": 407
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.4029352901934625e-05,
+      "loss": 0.0023,
+      "step": 408
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.40126751167445e-05,
+      "loss": 0.0017,
+      "step": 409
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.399599733155437e-05,
+      "loss": 0.0012,
+      "step": 410
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.397931954636424e-05,
+      "loss": 0.0015,
+      "step": 411
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.396264176117412e-05,
+      "loss": 0.0304,
+      "step": 412
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.394596397598399e-05,
+      "loss": 0.0011,
+      "step": 413
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.392928619079387e-05,
+      "loss": 0.0015,
+      "step": 414
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.391260840560374e-05,
+      "loss": 0.004,
+      "step": 415
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.389593062041361e-05,
+      "loss": 0.001,
+      "step": 416
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.3879252835223486e-05,
+      "loss": 0.0006,
+      "step": 417
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.3862575050033356e-05,
+      "loss": 0.0005,
+      "step": 418
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.384589726484323e-05,
+      "loss": 0.0009,
+      "step": 419
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.3829219479653104e-05,
+      "loss": 0.0005,
+      "step": 420
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.381254169446298e-05,
+      "loss": 0.514,
+      "step": 421
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.379586390927285e-05,
+      "loss": 0.4038,
+      "step": 422
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.377918612408272e-05,
+      "loss": 0.001,
+      "step": 423
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.37625083388926e-05,
+      "loss": 0.0012,
+      "step": 424
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.374583055370247e-05,
+      "loss": 0.0015,
+      "step": 425
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.3729152768512346e-05,
+      "loss": 0.0022,
+      "step": 426
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.3712474983322216e-05,
+      "loss": 0.0016,
+      "step": 427
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.369579719813209e-05,
+      "loss": 0.0033,
+      "step": 428
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.3679119412941964e-05,
+      "loss": 0.0021,
+      "step": 429
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.366244162775184e-05,
+      "loss": 0.3637,
+      "step": 430
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.364576384256171e-05,
+      "loss": 0.0083,
+      "step": 431
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.362908605737158e-05,
+      "loss": 0.4366,
+      "step": 432
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.361240827218145e-05,
+      "loss": 0.0013,
+      "step": 433
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.359573048699133e-05,
+      "loss": 0.001,
+      "step": 434
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.3579052701801206e-05,
+      "loss": 0.001,
+      "step": 435
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.3562374916611077e-05,
+      "loss": 0.0015,
+      "step": 436
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.354569713142095e-05,
+      "loss": 0.0017,
+      "step": 437
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.352901934623082e-05,
+      "loss": 0.0033,
+      "step": 438
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.35123415610407e-05,
+      "loss": 0.0019,
+      "step": 439
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.349566377585057e-05,
+      "loss": 0.0024,
+      "step": 440
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.347898599066044e-05,
+      "loss": 0.4103,
+      "step": 441
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.346230820547031e-05,
+      "loss": 0.0013,
+      "step": 442
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.344563042028019e-05,
+      "loss": 0.0013,
+      "step": 443
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.3428952635090067e-05,
+      "loss": 0.0099,
+      "step": 444
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.341227484989994e-05,
+      "loss": 0.0417,
+      "step": 445
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.339559706470981e-05,
+      "loss": 0.0015,
+      "step": 446
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.337891927951968e-05,
+      "loss": 0.2401,
+      "step": 447
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.3362241494329555e-05,
+      "loss": 0.0016,
+      "step": 448
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.334556370913943e-05,
+      "loss": 0.5658,
+      "step": 449
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.33288859239493e-05,
+      "loss": 0.0024,
+      "step": 450
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.331220813875917e-05,
+      "loss": 0.0034,
+      "step": 451
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.329553035356905e-05,
+      "loss": 0.1005,
+      "step": 452
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.327885256837892e-05,
+      "loss": 0.0029,
+      "step": 453
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.32621747831888e-05,
+      "loss": 0.0038,
+      "step": 454
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.324549699799867e-05,
+      "loss": 0.0022,
+      "step": 455
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.322881921280854e-05,
+      "loss": 0.0799,
+      "step": 456
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.3212141427618415e-05,
+      "loss": 0.0025,
+      "step": 457
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.3195463642428285e-05,
+      "loss": 0.0011,
+      "step": 458
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.317878585723816e-05,
+      "loss": 0.0011,
+      "step": 459
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.316210807204803e-05,
+      "loss": 0.0059,
+      "step": 460
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.314543028685791e-05,
+      "loss": 0.3185,
+      "step": 461
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.312875250166778e-05,
+      "loss": 0.0014,
+      "step": 462
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.311207471647765e-05,
+      "loss": 0.0011,
+      "step": 463
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.309539693128753e-05,
+      "loss": 0.2308,
+      "step": 464
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.3078719146097405e-05,
+      "loss": 0.0073,
+      "step": 465
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.3062041360907275e-05,
+      "loss": 0.0031,
+      "step": 466
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.3045363575717146e-05,
+      "loss": 0.0015,
+      "step": 467
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.3028685790527016e-05,
+      "loss": 0.1125,
+      "step": 468
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.301200800533689e-05,
+      "loss": 0.0007,
+      "step": 469
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.299533022014677e-05,
+      "loss": 0.0015,
+      "step": 470
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.297865243495664e-05,
+      "loss": 0.0015,
+      "step": 471
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.296197464976651e-05,
+      "loss": 0.0006,
+      "step": 472
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.294529686457638e-05,
+      "loss": 0.0029,
+      "step": 473
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.2928619079386265e-05,
+      "loss": 0.4703,
+      "step": 474
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.2911941294196136e-05,
+      "loss": 0.0013,
+      "step": 475
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.2895263509006006e-05,
+      "loss": 0.009,
+      "step": 476
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.2878585723815876e-05,
+      "loss": 0.0026,
+      "step": 477
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.286190793862575e-05,
+      "loss": 0.0101,
+      "step": 478
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.284523015343563e-05,
+      "loss": 0.0016,
+      "step": 479
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.28285523682455e-05,
+      "loss": 0.2461,
+      "step": 480
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.281187458305537e-05,
+      "loss": 0.003,
+      "step": 481
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.279519679786524e-05,
+      "loss": 0.0261,
+      "step": 482
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.277851901267512e-05,
+      "loss": 0.0015,
+      "step": 483
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.2761841227484996e-05,
+      "loss": 0.0012,
+      "step": 484
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.2745163442294866e-05,
+      "loss": 0.4948,
+      "step": 485
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.2728485657104737e-05,
+      "loss": 0.0018,
+      "step": 486
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.2711807871914614e-05,
+      "loss": 0.2557,
+      "step": 487
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.2695130086724484e-05,
+      "loss": 0.0014,
+      "step": 488
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.267845230153436e-05,
+      "loss": 0.0019,
+      "step": 489
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.266177451634423e-05,
+      "loss": 0.0024,
+      "step": 490
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.26450967311541e-05,
+      "loss": 0.0023,
+      "step": 491
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.262841894596398e-05,
+      "loss": 0.0018,
+      "step": 492
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.261174116077385e-05,
+      "loss": 0.6833,
+      "step": 493
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.2595063375583726e-05,
+      "loss": 0.3308,
+      "step": 494
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.25783855903936e-05,
+      "loss": 0.0035,
+      "step": 495
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.2561707805203474e-05,
+      "loss": 0.0101,
+      "step": 496
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.2545030020013344e-05,
+      "loss": 0.0101,
+      "step": 497
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.2528352234823215e-05,
+      "loss": 0.2967,
+      "step": 498
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.251167444963309e-05,
+      "loss": 0.0173,
+      "step": 499
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.249499666444296e-05,
+      "loss": 0.0099,
+      "step": 500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.247831887925284e-05,
+      "loss": 0.0297,
+      "step": 501
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.246164109406271e-05,
+      "loss": 0.2158,
+      "step": 502
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.244496330887258e-05,
+      "loss": 0.0324,
+      "step": 503
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.242828552368246e-05,
+      "loss": 0.0329,
+      "step": 504
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.2411607738492334e-05,
+      "loss": 0.0296,
+      "step": 505
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.2394929953302205e-05,
+      "loss": 0.0229,
+      "step": 506
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.2378252168112075e-05,
+      "loss": 0.0172,
+      "step": 507
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.2361574382921945e-05,
+      "loss": 0.0104,
+      "step": 508
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.234489659773182e-05,
+      "loss": 0.0072,
+      "step": 509
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.23282188125417e-05,
+      "loss": 0.2734,
+      "step": 510
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.231154102735157e-05,
+      "loss": 0.0041,
+      "step": 511
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.229486324216144e-05,
+      "loss": 0.0042,
+      "step": 512
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.227818545697132e-05,
+      "loss": 0.0049,
+      "step": 513
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.2261507671781194e-05,
+      "loss": 0.005,
+      "step": 514
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.2244829886591065e-05,
+      "loss": 0.0032,
+      "step": 515
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.2228152101400935e-05,
+      "loss": 0.0026,
+      "step": 516
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.2211474316210806e-05,
+      "loss": 0.0023,
+      "step": 517
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.219479653102068e-05,
+      "loss": 0.002,
+      "step": 518
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.217811874583056e-05,
+      "loss": 0.0014,
+      "step": 519
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.216144096064043e-05,
+      "loss": 0.001,
+      "step": 520
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.21447631754503e-05,
+      "loss": 0.001,
+      "step": 521
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.212808539026018e-05,
+      "loss": 0.2008,
+      "step": 522
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.211140760507005e-05,
+      "loss": 0.0008,
+      "step": 523
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.2094729819879925e-05,
+      "loss": 0.0007,
+      "step": 524
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.2078052034689795e-05,
+      "loss": 0.4649,
+      "step": 525
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.2061374249499666e-05,
+      "loss": 0.0008,
+      "step": 526
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.204469646430954e-05,
+      "loss": 0.0009,
+      "step": 527
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.202801867911941e-05,
+      "loss": 0.001,
+      "step": 528
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.201134089392929e-05,
+      "loss": 0.0011,
+      "step": 529
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.199466310873916e-05,
+      "loss": 0.2925,
+      "step": 530
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.197798532354904e-05,
+      "loss": 0.0017,
+      "step": 531
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.196130753835891e-05,
+      "loss": 0.3747,
+      "step": 532
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.194462975316878e-05,
+      "loss": 0.0018,
+      "step": 533
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.1927951967978656e-05,
+      "loss": 0.002,
+      "step": 534
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.1911274182788526e-05,
+      "loss": 0.0019,
+      "step": 535
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.18945963975984e-05,
+      "loss": 0.3065,
+      "step": 536
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.1877918612408274e-05,
+      "loss": 0.0022,
+      "step": 537
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.1861240827218144e-05,
+      "loss": 0.0018,
+      "step": 538
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.184456304202802e-05,
+      "loss": 0.0351,
+      "step": 539
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.18278852568379e-05,
+      "loss": 0.0027,
+      "step": 540
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.181120747164777e-05,
+      "loss": 0.0042,
+      "step": 541
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.179452968645764e-05,
+      "loss": 0.0024,
+      "step": 542
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.177785190126751e-05,
+      "loss": 0.0021,
+      "step": 543
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.1761174116077386e-05,
+      "loss": 0.3221,
+      "step": 544
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.1744496330887264e-05,
+      "loss": 0.003,
+      "step": 545
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.1727818545697134e-05,
+      "loss": 0.0036,
+      "step": 546
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.1711140760507004e-05,
+      "loss": 0.0025,
+      "step": 547
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.169446297531688e-05,
+      "loss": 0.0016,
+      "step": 548
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.167778519012676e-05,
+      "loss": 0.0014,
+      "step": 549
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.166110740493663e-05,
+      "loss": 0.0082,
+      "step": 550
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.16444296197465e-05,
+      "loss": 0.006,
+      "step": 551
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.162775183455637e-05,
+      "loss": 0.0012,
+      "step": 552
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.161107404936625e-05,
+      "loss": 0.0012,
+      "step": 553
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.1594396264176124e-05,
+      "loss": 1.4553,
+      "step": 554
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.1577718478985994e-05,
+      "loss": 0.7971,
+      "step": 555
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.1561040693795865e-05,
+      "loss": 0.4361,
+      "step": 556
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.154436290860574e-05,
+      "loss": 0.0019,
+      "step": 557
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.152768512341561e-05,
+      "loss": 0.0028,
+      "step": 558
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.151100733822549e-05,
+      "loss": 0.3036,
+      "step": 559
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.149432955303536e-05,
+      "loss": 0.2602,
+      "step": 560
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.147765176784523e-05,
+      "loss": 0.0205,
+      "step": 561
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.146097398265511e-05,
+      "loss": 0.0221,
+      "step": 562
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.144429619746498e-05,
+      "loss": 0.02,
+      "step": 563
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.1427618412274854e-05,
+      "loss": 0.0087,
+      "step": 564
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.1410940627084725e-05,
+      "loss": 0.2379,
+      "step": 565
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.13942628418946e-05,
+      "loss": 0.2308,
+      "step": 566
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.137758505670447e-05,
+      "loss": 0.3252,
+      "step": 567
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.136090727151434e-05,
+      "loss": 0.2486,
+      "step": 568
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.134422948632422e-05,
+      "loss": 0.0038,
+      "step": 569
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.132755170113409e-05,
+      "loss": 0.0058,
+      "step": 570
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.131087391594397e-05,
+      "loss": 0.0206,
+      "step": 571
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.129419613075384e-05,
+      "loss": 0.0077,
+      "step": 572
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.127751834556371e-05,
+      "loss": 0.1546,
+      "step": 573
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.1260840560373585e-05,
+      "loss": 0.2352,
+      "step": 574
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.124416277518346e-05,
+      "loss": 0.0077,
+      "step": 575
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.122748498999333e-05,
+      "loss": 0.0058,
+      "step": 576
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.12108072048032e-05,
+      "loss": 0.3266,
+      "step": 577
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.119412941961307e-05,
+      "loss": 0.2963,
+      "step": 578
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.117745163442295e-05,
+      "loss": 0.0055,
+      "step": 579
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.116077384923283e-05,
+      "loss": 0.2947,
+      "step": 580
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.11440960640427e-05,
+      "loss": 0.2467,
+      "step": 581
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.112741827885257e-05,
+      "loss": 0.035,
+      "step": 582
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.111074049366244e-05,
+      "loss": 0.0521,
+      "step": 583
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.109406270847232e-05,
+      "loss": 0.022,
+      "step": 584
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.107738492328219e-05,
+      "loss": 0.0209,
+      "step": 585
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.106070713809206e-05,
+      "loss": 0.0152,
+      "step": 586
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.1044029352901934e-05,
+      "loss": 0.0073,
+      "step": 587
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.102735156771181e-05,
+      "loss": 0.1828,
+      "step": 588
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.101067378252169e-05,
+      "loss": 0.0823,
+      "step": 589
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.099399599733156e-05,
+      "loss": 0.7055,
+      "step": 590
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.097731821214143e-05,
+      "loss": 0.4526,
+      "step": 591
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.09606404269513e-05,
+      "loss": 0.1376,
+      "step": 592
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.0943962641761176e-05,
+      "loss": 0.0043,
+      "step": 593
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.092728485657105e-05,
+      "loss": 0.2429,
+      "step": 594
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.0910607071380923e-05,
+      "loss": 0.0056,
+      "step": 595
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.0893929286190794e-05,
+      "loss": 0.0057,
+      "step": 596
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.087725150100067e-05,
+      "loss": 0.1615,
+      "step": 597
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.086057371581054e-05,
+      "loss": 0.0057,
+      "step": 598
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.084389593062042e-05,
+      "loss": 0.0055,
+      "step": 599
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.082721814543029e-05,
+      "loss": 0.1324,
+      "step": 600
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.0810540360240166e-05,
+      "loss": 0.0033,
+      "step": 601
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.0793862575050036e-05,
+      "loss": 0.0043,
+      "step": 602
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.0777184789859907e-05,
+      "loss": 0.1004,
+      "step": 603
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.0760507004669784e-05,
+      "loss": 0.0039,
+      "step": 604
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.0743829219479654e-05,
+      "loss": 0.0068,
+      "step": 605
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.072715143428953e-05,
+      "loss": 0.0976,
+      "step": 606
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.07104736490994e-05,
+      "loss": 0.0449,
+      "step": 607
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.069379586390927e-05,
+      "loss": 0.0031,
+      "step": 608
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.067711807871915e-05,
+      "loss": 0.0062,
+      "step": 609
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.0660440293529026e-05,
+      "loss": 0.0195,
+      "step": 610
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.0643762508338897e-05,
+      "loss": 0.2125,
+      "step": 611
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.062708472314877e-05,
+      "loss": 0.0018,
+      "step": 612
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.061040693795864e-05,
+      "loss": 0.0139,
+      "step": 613
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.0593729152768514e-05,
+      "loss": 0.002,
+      "step": 614
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.057705136757839e-05,
+      "loss": 0.0026,
+      "step": 615
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.056037358238826e-05,
+      "loss": 0.0014,
+      "step": 616
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.054369579719813e-05,
+      "loss": 0.2354,
+      "step": 617
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.0527018012008e-05,
+      "loss": 0.0016,
+      "step": 618
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.051034022681788e-05,
+      "loss": 0.2635,
+      "step": 619
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.049366244162776e-05,
+      "loss": 0.0012,
+      "step": 620
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.047698465643763e-05,
+      "loss": 0.4177,
+      "step": 621
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.04603068712475e-05,
+      "loss": 0.0009,
+      "step": 622
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.0443629086057375e-05,
+      "loss": 0.001,
+      "step": 623
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.0426951300867245e-05,
+      "loss": 0.0109,
+      "step": 624
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.041027351567712e-05,
+      "loss": 0.3153,
+      "step": 625
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.039359573048699e-05,
+      "loss": 0.0011,
+      "step": 626
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.037691794529686e-05,
+      "loss": 0.0012,
+      "step": 627
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.036024016010674e-05,
+      "loss": 0.0016,
+      "step": 628
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.034356237491661e-05,
+      "loss": 0.0011,
+      "step": 629
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.032688458972649e-05,
+      "loss": 0.0026,
+      "step": 630
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.031020680453636e-05,
+      "loss": 0.2383,
+      "step": 631
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.0293529019346235e-05,
+      "loss": 0.3166,
+      "step": 632
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.0276851234156105e-05,
+      "loss": 0.0016,
+      "step": 633
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.0260173448965976e-05,
+      "loss": 0.0013,
+      "step": 634
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.024349566377585e-05,
+      "loss": 0.001,
+      "step": 635
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.022681787858572e-05,
+      "loss": 0.0021,
+      "step": 636
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.02101400933956e-05,
+      "loss": 0.0009,
+      "step": 637
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.019346230820547e-05,
+      "loss": 0.1609,
+      "step": 638
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.017678452301534e-05,
+      "loss": 0.0014,
+      "step": 639
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.016010673782522e-05,
+      "loss": 0.001,
+      "step": 640
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.0143428952635095e-05,
+      "loss": 0.0017,
+      "step": 641
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.0126751167444966e-05,
+      "loss": 0.0102,
+      "step": 642
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.0110073382254836e-05,
+      "loss": 0.0062,
+      "step": 643
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.0093395597064706e-05,
+      "loss": 0.0007,
+      "step": 644
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.007671781187458e-05,
+      "loss": 0.0992,
+      "step": 645
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.006004002668446e-05,
+      "loss": 0.0008,
+      "step": 646
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.004336224149433e-05,
+      "loss": 0.0009,
+      "step": 647
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.00266844563042e-05,
+      "loss": 0.0007,
+      "step": 648
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.001000667111408e-05,
+      "loss": 0.0007,
+      "step": 649
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.9993328885923955e-05,
+      "loss": 0.0009,
+      "step": 650
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.9976651100733826e-05,
+      "loss": 0.0006,
+      "step": 651
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.9959973315543696e-05,
+      "loss": 0.0037,
+      "step": 652
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.9943295530353567e-05,
+      "loss": 0.0007,
+      "step": 653
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.9926617745163444e-05,
+      "loss": 0.0006,
+      "step": 654
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.990993995997332e-05,
+      "loss": 0.0007,
+      "step": 655
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.989326217478319e-05,
+      "loss": 0.0008,
+      "step": 656
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.987658438959306e-05,
+      "loss": 0.0007,
+      "step": 657
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.985990660440294e-05,
+      "loss": 0.0005,
+      "step": 658
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.984322881921281e-05,
+      "loss": 0.0042,
+      "step": 659
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.9826551034022686e-05,
+      "loss": 0.0006,
+      "step": 660
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.9809873248832556e-05,
+      "loss": 0.0005,
+      "step": 661
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.979319546364243e-05,
+      "loss": 0.0006,
+      "step": 662
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.9776517678452304e-05,
+      "loss": 0.3488,
+      "step": 663
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.9759839893262174e-05,
+      "loss": 0.0013,
+      "step": 664
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.974316210807205e-05,
+      "loss": 0.0006,
+      "step": 665
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.972648432288192e-05,
+      "loss": 0.0004,
+      "step": 666
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.97098065376918e-05,
+      "loss": 0.0004,
+      "step": 667
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.969312875250167e-05,
+      "loss": 0.0004,
+      "step": 668
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.967645096731154e-05,
+      "loss": 0.0005,
+      "step": 669
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.965977318212142e-05,
+      "loss": 0.0005,
+      "step": 670
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.964309539693129e-05,
+      "loss": 0.0005,
+      "step": 671
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.9626417611741164e-05,
+      "loss": 0.0007,
+      "step": 672
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.9609739826551035e-05,
+      "loss": 0.4719,
+      "step": 673
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.9593062041360905e-05,
+      "loss": 0.0004,
+      "step": 674
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.957638425617078e-05,
+      "loss": 0.0005,
+      "step": 675
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.955970647098066e-05,
+      "loss": 0.0044,
+      "step": 676
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.954302868579053e-05,
+      "loss": 0.0122,
+      "step": 677
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.95263509006004e-05,
+      "loss": 0.0024,
+      "step": 678
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.950967311541027e-05,
+      "loss": 0.0005,
+      "step": 679
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.949299533022015e-05,
+      "loss": 0.0014,
+      "step": 680
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.9476317545030025e-05,
+      "loss": 0.0006,
+      "step": 681
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.9459639759839895e-05,
+      "loss": 0.4023,
+      "step": 682
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.9442961974649765e-05,
+      "loss": 0.0008,
+      "step": 683
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.942628418945964e-05,
+      "loss": 0.0005,
+      "step": 684
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.940960640426952e-05,
+      "loss": 0.0005,
+      "step": 685
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.939292861907939e-05,
+      "loss": 0.5024,
+      "step": 686
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.937625083388926e-05,
+      "loss": 0.0581,
+      "step": 687
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.935957304869913e-05,
+      "loss": 0.0013,
+      "step": 688
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.934289526350901e-05,
+      "loss": 0.1772,
+      "step": 689
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.9326217478318885e-05,
+      "loss": 0.0008,
+      "step": 690
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.9309539693128755e-05,
+      "loss": 0.3078,
+      "step": 691
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.9292861907938625e-05,
+      "loss": 0.0005,
+      "step": 692
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.92761841227485e-05,
+      "loss": 0.0004,
+      "step": 693
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 3.925950633755837e-05,
+      "loss": 0.0005,
+      "step": 694
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 3.924282855236825e-05,
+      "loss": 0.0005,
+      "step": 695
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 3.922615076717812e-05,
+      "loss": 0.0005,
+      "step": 696
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 3.920947298198799e-05,
+      "loss": 0.2643,
+      "step": 697
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 3.919279519679787e-05,
+      "loss": 0.0006,
+      "step": 698
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 3.917611741160774e-05,
+      "loss": 0.0029,
+      "step": 699
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 3.9159439626417615e-05,
+      "loss": 0.0168,
+      "step": 700
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 3.9142761841227486e-05,
+      "loss": 0.424,
+      "step": 701
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 3.912608405603736e-05,
+      "loss": 0.0005,
+      "step": 702
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 3.910940627084723e-05,
+      "loss": 0.3861,
+      "step": 703
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 3.9092728485657104e-05,
+      "loss": 0.0026,
+      "step": 704
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 3.907605070046698e-05,
+      "loss": 0.0005,
+      "step": 705
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 3.905937291527685e-05,
+      "loss": 0.0005,
+      "step": 706
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 3.904269513008673e-05,
+      "loss": 0.0035,
+      "step": 707
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 3.90260173448966e-05,
+      "loss": 0.4003,
+      "step": 708
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.900933955970647e-05,
+      "loss": 0.7569,
+      "step": 709
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.8992661774516346e-05,
+      "loss": 0.0008,
+      "step": 710
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.897598398932622e-05,
+      "loss": 0.0013,
+      "step": 711
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.8959306204136094e-05,
+      "loss": 0.002,
+      "step": 712
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.8942628418945964e-05,
+      "loss": 0.0026,
+      "step": 713
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.8925950633755834e-05,
+      "loss": 0.0027,
+      "step": 714
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.890927284856571e-05,
+      "loss": 0.2189,
+      "step": 715
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.889259506337559e-05,
+      "loss": 0.0047,
+      "step": 716
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.887591727818546e-05,
+      "loss": 0.0238,
+      "step": 717
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.885923949299533e-05,
+      "loss": 0.0033,
+      "step": 718
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.88425617078052e-05,
+      "loss": 0.0012,
+      "step": 719
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.8825883922615083e-05,
+      "loss": 0.0011,
+      "step": 720
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.8809206137424954e-05,
+      "loss": 0.0011,
+      "step": 721
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.8792528352234824e-05,
+      "loss": 0.0008,
+      "step": 722
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.8775850567044695e-05,
+      "loss": 0.0007,
+      "step": 723
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.875917278185457e-05,
+      "loss": 0.0006,
+      "step": 724
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.874249499666445e-05,
+      "loss": 0.4621,
+      "step": 725
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.872581721147432e-05,
+      "loss": 0.0008,
+      "step": 726
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.870913942628419e-05,
+      "loss": 0.0009,
+      "step": 727
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.8692461641094067e-05,
+      "loss": 0.0012,
+      "step": 728
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.867578385590394e-05,
+      "loss": 0.0013,
+      "step": 729
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.8659106070713814e-05,
+      "loss": 0.0014,
+      "step": 730
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.8642428285523684e-05,
+      "loss": 0.0039,
+      "step": 731
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.8625750500333555e-05,
+      "loss": 0.3797,
+      "step": 732
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.860907271514343e-05,
+      "loss": 0.1586,
+      "step": 733
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.85923949299533e-05,
+      "loss": 0.004,
+      "step": 734
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.857571714476318e-05,
+      "loss": 0.0882,
+      "step": 735
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.855903935957305e-05,
+      "loss": 0.0102,
+      "step": 736
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.854236157438293e-05,
+      "loss": 0.007,
+      "step": 737
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.85256837891928e-05,
+      "loss": 0.0104,
+      "step": 738
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.850900600400267e-05,
+      "loss": 0.0331,
+      "step": 739
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 3.8492328218812545e-05,
+      "loss": 0.0067,
+      "step": 740
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 3.8475650433622415e-05,
+      "loss": 0.0061,
+      "step": 741
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 3.845897264843229e-05,
+      "loss": 0.015,
+      "step": 742
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 3.844229486324216e-05,
+      "loss": 0.0061,
+      "step": 743
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 3.842561707805203e-05,
+      "loss": 0.0051,
+      "step": 744
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 3.840893929286191e-05,
+      "loss": 0.0215,
+      "step": 745
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 3.839226150767179e-05,
+      "loss": 0.2269,
+      "step": 746
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 3.837558372248166e-05,
+      "loss": 0.3612,
+      "step": 747
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 3.835890593729153e-05,
+      "loss": 0.0037,
+      "step": 748
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 3.83422281521014e-05,
+      "loss": 0.0022,
+      "step": 749
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 3.8325550366911275e-05,
+      "loss": 0.0021,
+      "step": 750
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 3.830887258172115e-05,
+      "loss": 0.0021,
+      "step": 751
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 3.829219479653102e-05,
+      "loss": 0.0015,
+      "step": 752
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 3.827551701134089e-05,
+      "loss": 0.3934,
+      "step": 753
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 3.8258839226150764e-05,
+      "loss": 0.0032,
+      "step": 754
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 3.824216144096065e-05,
+      "loss": 0.0018,
+      "step": 755
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 3.822548365577052e-05,
+      "loss": 0.6576,
+      "step": 756
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 3.820880587058039e-05,
+      "loss": 0.0019,
+      "step": 757
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 3.819212808539026e-05,
+      "loss": 0.0031,
+      "step": 758
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 3.8175450300200136e-05,
+      "loss": 0.0032,
+      "step": 759
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 3.815877251501001e-05,
+      "loss": 0.0127,
+      "step": 760
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 3.814209472981988e-05,
+      "loss": 0.0461,
+      "step": 761
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 3.8125416944629753e-05,
+      "loss": 0.0025,
+      "step": 762
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 3.8108739159439624e-05,
+      "loss": 0.0029,
+      "step": 763
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 3.80920613742495e-05,
+      "loss": 0.0026,
+      "step": 764
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 3.807538358905938e-05,
+      "loss": 0.0025,
+      "step": 765
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 3.805870580386925e-05,
+      "loss": 0.3283,
+      "step": 766
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 3.804202801867912e-05,
+      "loss": 0.2321,
+      "step": 767
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 3.8025350233488996e-05,
+      "loss": 0.0021,
+      "step": 768
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 3.8008672448298866e-05,
+      "loss": 0.0028,
+      "step": 769
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 3.799199466310874e-05,
+      "loss": 0.0027,
+      "step": 770
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 3.7975316877918614e-05,
+      "loss": 0.0027,
+      "step": 771
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 3.7958639092728484e-05,
+      "loss": 0.004,
+      "step": 772
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 3.794196130753836e-05,
+      "loss": 0.0029,
+      "step": 773
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 3.792528352234823e-05,
+      "loss": 0.0026,
+      "step": 774
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 3.790860573715811e-05,
+      "loss": 0.0027,
+      "step": 775
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 3.789192795196798e-05,
+      "loss": 0.0026,
+      "step": 776
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 3.7875250166777856e-05,
+      "loss": 0.0022,
+      "step": 777
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 3.7858572381587727e-05,
+      "loss": 1.0656,
+      "step": 778
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 3.78418945963976e-05,
+      "loss": 0.0027,
+      "step": 779
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 3.7825216811207474e-05,
+      "loss": 0.0019,
+      "step": 780
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 3.780853902601735e-05,
+      "loss": 0.0027,
+      "step": 781
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 3.779186124082722e-05,
+      "loss": 0.0042,
+      "step": 782
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 3.777518345563709e-05,
+      "loss": 0.0038,
+      "step": 783
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 3.775850567044696e-05,
+      "loss": 0.2915,
+      "step": 784
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 3.774182788525684e-05,
+      "loss": 0.002,
+      "step": 785
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 3.7725150100066716e-05,
+      "loss": 0.0151,
+      "step": 786
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 3.770847231487659e-05,
+      "loss": 0.0052,
+      "step": 787
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 3.769179452968646e-05,
+      "loss": 0.0041,
+      "step": 788
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 3.767511674449633e-05,
+      "loss": 0.0037,
+      "step": 789
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 3.765843895930621e-05,
+      "loss": 0.0019,
+      "step": 790
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 3.764176117411608e-05,
+      "loss": 0.0021,
+      "step": 791
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 3.762508338892595e-05,
+      "loss": 0.3614,
+      "step": 792
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 3.760840560373582e-05,
+      "loss": 0.0021,
+      "step": 793
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 3.75917278185457e-05,
+      "loss": 0.209,
+      "step": 794
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 3.757505003335558e-05,
+      "loss": 0.0024,
+      "step": 795
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 3.755837224816545e-05,
+      "loss": 0.3805,
+      "step": 796
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 3.754169446297532e-05,
+      "loss": 0.0074,
+      "step": 797
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 3.752501667778519e-05,
+      "loss": 0.0203,
+      "step": 798
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 3.7508338892595065e-05,
+      "loss": 0.0076,
+      "step": 799
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 3.749166110740494e-05,
+      "loss": 0.0038,
+      "step": 800
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 3.747498332221481e-05,
+      "loss": 0.1372,
+      "step": 801
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 3.745830553702468e-05,
+      "loss": 0.0043,
+      "step": 802
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 3.744162775183456e-05,
+      "loss": 0.0049,
+      "step": 803
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 3.742494996664443e-05,
+      "loss": 0.011,
+      "step": 804
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 3.740827218145431e-05,
+      "loss": 0.0053,
+      "step": 805
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 3.739159439626418e-05,
+      "loss": 0.0057,
+      "step": 806
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 3.737491661107405e-05,
+      "loss": 0.2745,
+      "step": 807
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 3.7358238825883925e-05,
+      "loss": 0.0197,
+      "step": 808
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 3.7341561040693796e-05,
+      "loss": 0.0024,
+      "step": 809
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 3.732488325550367e-05,
+      "loss": 0.575,
+      "step": 810
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 3.730820547031354e-05,
+      "loss": 0.0014,
+      "step": 811
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 3.729152768512342e-05,
+      "loss": 0.0015,
+      "step": 812
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 3.727484989993329e-05,
+      "loss": 0.0022,
+      "step": 813
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 3.725817211474316e-05,
+      "loss": 0.0023,
+      "step": 814
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 3.724149432955304e-05,
+      "loss": 0.0015,
+      "step": 815
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 3.722481654436291e-05,
+      "loss": 0.3778,
+      "step": 816
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 3.7208138759172785e-05,
+      "loss": 0.0021,
+      "step": 817
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 3.7191460973982656e-05,
+      "loss": 0.0021,
+      "step": 818
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 3.7174783188792526e-05,
+      "loss": 0.0022,
+      "step": 819
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 3.71581054036024e-05,
+      "loss": 0.6581,
+      "step": 820
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 3.714142761841228e-05,
+      "loss": 0.0019,
+      "step": 821
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 3.712474983322215e-05,
+      "loss": 0.0027,
+      "step": 822
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 3.710807204803202e-05,
+      "loss": 0.002,
+      "step": 823
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 3.709139426284189e-05,
+      "loss": 0.2313,
+      "step": 824
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 3.707471647765177e-05,
+      "loss": 0.321,
+      "step": 825
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 3.7058038692461646e-05,
+      "loss": 0.0033,
+      "step": 826
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 3.7041360907271516e-05,
+      "loss": 0.008,
+      "step": 827
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 3.7024683122081386e-05,
+      "loss": 0.0043,
+      "step": 828
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 3.7008005336891264e-05,
+      "loss": 0.0076,
+      "step": 829
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 3.699132755170114e-05,
+      "loss": 0.0022,
+      "step": 830
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 3.697464976651101e-05,
+      "loss": 0.2972,
+      "step": 831
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 3.695797198132088e-05,
+      "loss": 0.0019,
+      "step": 832
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 3.694129419613075e-05,
+      "loss": 0.002,
+      "step": 833
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 3.692461641094063e-05,
+      "loss": 0.0022,
+      "step": 834
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 3.6907938625750506e-05,
+      "loss": 0.0026,
+      "step": 835
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 3.6891260840560376e-05,
+      "loss": 0.0027,
+      "step": 836
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 3.687458305537025e-05,
+      "loss": 0.0019,
+      "step": 837
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 3.6857905270180124e-05,
+      "loss": 0.2285,
+      "step": 838
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 3.6841227484989994e-05,
+      "loss": 0.4608,
+      "step": 839
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 3.682454969979987e-05,
+      "loss": 0.0026,
+      "step": 840
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 3.680787191460974e-05,
+      "loss": 0.004,
+      "step": 841
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 3.679119412941961e-05,
+      "loss": 0.0075,
+      "step": 842
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 3.677451634422949e-05,
+      "loss": 0.0146,
+      "step": 843
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 3.675783855903936e-05,
+      "loss": 0.006,
+      "step": 844
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 3.674116077384924e-05,
+      "loss": 0.3795,
+      "step": 845
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 3.672448298865911e-05,
+      "loss": 0.0054,
+      "step": 846
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 3.6707805203468984e-05,
+      "loss": 0.0058,
+      "step": 847
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 3.6691127418278855e-05,
+      "loss": 0.187,
+      "step": 848
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 3.6674449633088725e-05,
+      "loss": 0.0036,
+      "step": 849
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 3.66577718478986e-05,
+      "loss": 0.0045,
+      "step": 850
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 3.664109406270847e-05,
+      "loss": 0.2647,
+      "step": 851
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 3.662441627751835e-05,
+      "loss": 0.0049,
+      "step": 852
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 3.660773849232822e-05,
+      "loss": 0.0054,
+      "step": 853
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 3.659106070713809e-05,
+      "loss": 0.0038,
+      "step": 854
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 3.657438292194797e-05,
+      "loss": 0.006,
+      "step": 855
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 3.6557705136757844e-05,
+      "loss": 0.0042,
+      "step": 856
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 3.6541027351567715e-05,
+      "loss": 0.0043,
+      "step": 857
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 3.6524349566377585e-05,
+      "loss": 0.0042,
+      "step": 858
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 3.6507671781187455e-05,
+      "loss": 0.0035,
+      "step": 859
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 3.649099399599733e-05,
+      "loss": 0.003,
+      "step": 860
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 3.647431621080721e-05,
+      "loss": 0.002,
+      "step": 861
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 3.645763842561708e-05,
+      "loss": 0.2869,
+      "step": 862
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 3.644096064042695e-05,
+      "loss": 0.0019,
+      "step": 863
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 3.642428285523683e-05,
+      "loss": 0.3849,
+      "step": 864
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 3.6407605070046705e-05,
+      "loss": 0.0019,
+      "step": 865
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 3.6390927284856575e-05,
+      "loss": 0.0019,
+      "step": 866
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 3.6374249499666445e-05,
+      "loss": 0.0019,
+      "step": 867
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 3.6357571714476316e-05,
+      "loss": 0.0019,
+      "step": 868
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 3.634089392928619e-05,
+      "loss": 0.0021,
+      "step": 869
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 3.632421614409607e-05,
+      "loss": 0.0017,
+      "step": 870
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 3.630753835890594e-05,
+      "loss": 0.0015,
+      "step": 871
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 3.629086057371581e-05,
+      "loss": 0.0018,
+      "step": 872
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 3.627418278852569e-05,
+      "loss": 0.0019,
+      "step": 873
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 3.625750500333556e-05,
+      "loss": 0.0023,
+      "step": 874
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 3.6240827218145435e-05,
+      "loss": 0.325,
+      "step": 875
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 3.6224149432955306e-05,
+      "loss": 0.414,
+      "step": 876
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.6207471647765176e-05,
+      "loss": 0.0017,
+      "step": 877
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.619079386257505e-05,
+      "loss": 0.0016,
+      "step": 878
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.6174116077384924e-05,
+      "loss": 0.0018,
+      "step": 879
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.61574382921948e-05,
+      "loss": 0.0017,
+      "step": 880
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.614076050700467e-05,
+      "loss": 0.3202,
+      "step": 881
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.612408272181455e-05,
+      "loss": 0.0018,
+      "step": 882
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.610740493662442e-05,
+      "loss": 0.0022,
+      "step": 883
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.609072715143429e-05,
+      "loss": 0.0067,
+      "step": 884
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.6074049366244166e-05,
+      "loss": 0.0244,
+      "step": 885
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.6057371581054036e-05,
+      "loss": 0.1974,
+      "step": 886
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.6040693795863913e-05,
+      "loss": 0.0022,
+      "step": 887
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.6024016010673784e-05,
+      "loss": 0.0025,
+      "step": 888
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.6007338225483654e-05,
+      "loss": 0.0064,
+      "step": 889
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.599066044029353e-05,
+      "loss": 0.0023,
+      "step": 890
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.597398265510341e-05,
+      "loss": 0.0037,
+      "step": 891
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.595730486991328e-05,
+      "loss": 0.0045,
+      "step": 892
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.594062708472315e-05,
+      "loss": 0.1247,
+      "step": 893
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.592394929953302e-05,
+      "loss": 0.0025,
+      "step": 894
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.5907271514342897e-05,
+      "loss": 0.2346,
+      "step": 895
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.5890593729152774e-05,
+      "loss": 0.1991,
+      "step": 896
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.5873915943962644e-05,
+      "loss": 0.0019,
+      "step": 897
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.5857238158772514e-05,
+      "loss": 0.0028,
+      "step": 898
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.5840560373582385e-05,
+      "loss": 0.4147,
+      "step": 899
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.582388258839227e-05,
+      "loss": 0.2646,
+      "step": 900
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.580720480320214e-05,
+      "loss": 0.0043,
+      "step": 901
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.579052701801201e-05,
+      "loss": 0.3339,
+      "step": 902
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.577384923282188e-05,
+      "loss": 0.5998,
+      "step": 903
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.575717144763176e-05,
+      "loss": 0.0047,
+      "step": 904
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.5740493662441634e-05,
+      "loss": 0.1014,
+      "step": 905
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.5723815877251504e-05,
+      "loss": 0.1743,
+      "step": 906
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.5707138092061375e-05,
+      "loss": 0.3604,
+      "step": 907
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.5690460306871245e-05,
+      "loss": 0.0104,
+      "step": 908
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.567378252168112e-05,
+      "loss": 0.0052,
+      "step": 909
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.5657104736491e-05,
+      "loss": 0.01,
+      "step": 910
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.564042695130087e-05,
+      "loss": 0.0062,
+      "step": 911
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.562374916611074e-05,
+      "loss": 0.006,
+      "step": 912
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.560707138092062e-05,
+      "loss": 0.0038,
+      "step": 913
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.559039359573049e-05,
+      "loss": 0.0044,
+      "step": 914
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.5573715810540365e-05,
+      "loss": 0.354,
+      "step": 915
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.5557038025350235e-05,
+      "loss": 0.0056,
+      "step": 916
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.554036024016011e-05,
+      "loss": 0.0022,
+      "step": 917
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.552368245496998e-05,
+      "loss": 0.0854,
+      "step": 918
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.550700466977985e-05,
+      "loss": 0.0041,
+      "step": 919
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.549032688458973e-05,
+      "loss": 0.0021,
+      "step": 920
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.54736490993996e-05,
+      "loss": 0.1785,
+      "step": 921
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.545697131420948e-05,
+      "loss": 0.0032,
+      "step": 922
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.544029352901935e-05,
+      "loss": 0.003,
+      "step": 923
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.542361574382922e-05,
+      "loss": 0.0028,
+      "step": 924
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.5406937958639095e-05,
+      "loss": 0.0037,
+      "step": 925
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.539026017344897e-05,
+      "loss": 0.1162,
+      "step": 926
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.537358238825884e-05,
+      "loss": 0.0031,
+      "step": 927
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.535690460306871e-05,
+      "loss": 0.0027,
+      "step": 928
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.5340226817878583e-05,
+      "loss": 0.1153,
+      "step": 929
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.532354903268846e-05,
+      "loss": 0.0157,
+      "step": 930
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.530687124749834e-05,
+      "loss": 0.0027,
+      "step": 931
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.529019346230821e-05,
+      "loss": 0.002,
+      "step": 932
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.527351567711808e-05,
+      "loss": 0.0029,
+      "step": 933
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.525683789192795e-05,
+      "loss": 0.0017,
+      "step": 934
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.524016010673783e-05,
+      "loss": 0.0032,
+      "step": 935
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.52234823215477e-05,
+      "loss": 0.0018,
+      "step": 936
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.5206804536357573e-05,
+      "loss": 0.0018,
+      "step": 937
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.5190126751167444e-05,
+      "loss": 0.0018,
+      "step": 938
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.517344896597732e-05,
+      "loss": 0.0018,
+      "step": 939
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.51567711807872e-05,
+      "loss": 0.0019,
+      "step": 940
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.514009339559707e-05,
+      "loss": 0.002,
+      "step": 941
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.512341561040694e-05,
+      "loss": 0.3323,
+      "step": 942
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.510673782521681e-05,
+      "loss": 0.0017,
+      "step": 943
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.5090060040026686e-05,
+      "loss": 0.0015,
+      "step": 944
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.507338225483656e-05,
+      "loss": 0.0011,
+      "step": 945
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.5056704469646434e-05,
+      "loss": 0.0019,
+      "step": 946
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.5040026684456304e-05,
+      "loss": 0.0011,
+      "step": 947
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.502334889926618e-05,
+      "loss": 0.0011,
+      "step": 948
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.500667111407605e-05,
+      "loss": 0.0011,
+      "step": 949
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.498999332888593e-05,
+      "loss": 0.0019,
+      "step": 950
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.49733155436958e-05,
+      "loss": 0.0014,
+      "step": 951
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.495663775850567e-05,
+      "loss": 0.0011,
+      "step": 952
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.4939959973315546e-05,
+      "loss": 0.0013,
+      "step": 953
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.492328218812542e-05,
+      "loss": 0.2257,
+      "step": 954
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.4906604402935294e-05,
+      "loss": 0.0021,
+      "step": 955
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.4889926617745164e-05,
+      "loss": 0.001,
+      "step": 956
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.487324883255504e-05,
+      "loss": 0.0011,
+      "step": 957
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.485657104736491e-05,
+      "loss": 0.0012,
+      "step": 958
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.483989326217478e-05,
+      "loss": 0.0008,
+      "step": 959
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.482321547698466e-05,
+      "loss": 0.4449,
+      "step": 960
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.480653769179453e-05,
+      "loss": 0.2154,
+      "step": 961
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.478985990660441e-05,
+      "loss": 0.001,
+      "step": 962
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.477318212141428e-05,
+      "loss": 0.0011,
+      "step": 963
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.475650433622415e-05,
+      "loss": 0.0715,
+      "step": 964
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.4739826551034025e-05,
+      "loss": 0.0011,
+      "step": 965
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.47231487658439e-05,
+      "loss": 0.001,
+      "step": 966
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.470647098065377e-05,
+      "loss": 0.3556,
+      "step": 967
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.468979319546364e-05,
+      "loss": 0.0216,
+      "step": 968
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.467311541027351e-05,
+      "loss": 0.0012,
+      "step": 969
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.46564376250834e-05,
+      "loss": 0.0014,
+      "step": 970
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.463975983989327e-05,
+      "loss": 0.0014,
+      "step": 971
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.462308205470314e-05,
+      "loss": 0.3395,
+      "step": 972
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.460640426951301e-05,
+      "loss": 0.0016,
+      "step": 973
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.4589726484322885e-05,
+      "loss": 0.0314,
+      "step": 974
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.457304869913276e-05,
+      "loss": 0.0241,
+      "step": 975
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.455637091394263e-05,
+      "loss": 0.0969,
+      "step": 976
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.45396931287525e-05,
+      "loss": 0.2976,
+      "step": 977
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.452301534356237e-05,
+      "loss": 0.0259,
+      "step": 978
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.450633755837225e-05,
+      "loss": 0.0035,
+      "step": 979
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.448965977318213e-05,
+      "loss": 0.0033,
+      "step": 980
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.4472981987992e-05,
+      "loss": 0.0027,
+      "step": 981
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.445630420280187e-05,
+      "loss": 0.0024,
+      "step": 982
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.4439626417611745e-05,
+      "loss": 0.0463,
+      "step": 983
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.4422948632421615e-05,
+      "loss": 0.0057,
+      "step": 984
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.440627084723149e-05,
+      "loss": 0.0024,
+      "step": 985
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.438959306204136e-05,
+      "loss": 0.0015,
+      "step": 986
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.437291527685123e-05,
+      "loss": 0.428,
+      "step": 987
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.435623749166111e-05,
+      "loss": 0.0015,
+      "step": 988
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.433955970647098e-05,
+      "loss": 0.0028,
+      "step": 989
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.432288192128085e-05,
+      "loss": 0.0033,
+      "step": 990
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.430620413609073e-05,
+      "loss": 0.3294,
+      "step": 991
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.4289526350900605e-05,
+      "loss": 0.0041,
+      "step": 992
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.4272848565710476e-05,
+      "loss": 0.2262,
+      "step": 993
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.4256170780520346e-05,
+      "loss": 0.0016,
+      "step": 994
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.4239492995330216e-05,
+      "loss": 0.0017,
+      "step": 995
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.4222815210140094e-05,
+      "loss": 0.0021,
+      "step": 996
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.420613742494997e-05,
+      "loss": 0.0016,
+      "step": 997
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.418945963975984e-05,
+      "loss": 0.0012,
+      "step": 998
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.417278185456971e-05,
+      "loss": 0.007,
+      "step": 999
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.415610406937959e-05,
+      "loss": 0.3263,
+      "step": 1000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.4139426284189466e-05,
+      "loss": 0.0042,
+      "step": 1001
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.4122748498999336e-05,
+      "loss": 0.0096,
+      "step": 1002
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.4106070713809206e-05,
+      "loss": 0.0012,
+      "step": 1003
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.408939292861908e-05,
+      "loss": 0.0157,
+      "step": 1004
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.4072715143428954e-05,
+      "loss": 0.0025,
+      "step": 1005
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.405603735823883e-05,
+      "loss": 0.0021,
+      "step": 1006
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.40393595730487e-05,
+      "loss": 0.0033,
+      "step": 1007
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.402268178785857e-05,
+      "loss": 0.2794,
+      "step": 1008
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.400600400266845e-05,
+      "loss": 0.0013,
+      "step": 1009
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.398932621747832e-05,
+      "loss": 0.001,
+      "step": 1010
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.3972648432288196e-05,
+      "loss": 0.6856,
+      "step": 1011
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.395597064709807e-05,
+      "loss": 0.0012,
+      "step": 1012
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.393929286190794e-05,
+      "loss": 0.0012,
+      "step": 1013
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.3922615076717814e-05,
+      "loss": 0.004,
+      "step": 1014
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.3905937291527685e-05,
+      "loss": 0.0012,
+      "step": 1015
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.388925950633756e-05,
+      "loss": 0.1405,
+      "step": 1016
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.387258172114743e-05,
+      "loss": 0.001,
+      "step": 1017
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.385590393595731e-05,
+      "loss": 0.0021,
+      "step": 1018
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.383922615076718e-05,
+      "loss": 0.0012,
+      "step": 1019
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.382254836557705e-05,
+      "loss": 0.0023,
+      "step": 1020
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.380587058038693e-05,
+      "loss": 0.0012,
+      "step": 1021
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.37891927951968e-05,
+      "loss": 0.2768,
+      "step": 1022
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.3772515010006674e-05,
+      "loss": 0.0013,
+      "step": 1023
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.3755837224816545e-05,
+      "loss": 0.0013,
+      "step": 1024
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.3739159439626415e-05,
+      "loss": 0.0021,
+      "step": 1025
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.372248165443629e-05,
+      "loss": 0.0014,
+      "step": 1026
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.370580386924617e-05,
+      "loss": 0.0012,
+      "step": 1027
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.368912608405604e-05,
+      "loss": 0.0011,
+      "step": 1028
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.367244829886591e-05,
+      "loss": 0.2213,
+      "step": 1029
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.365577051367578e-05,
+      "loss": 0.0011,
+      "step": 1030
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.363909272848566e-05,
+      "loss": 0.0814,
+      "step": 1031
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.3622414943295535e-05,
+      "loss": 0.0011,
+      "step": 1032
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.3605737158105405e-05,
+      "loss": 0.0689,
+      "step": 1033
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.3589059372915275e-05,
+      "loss": 0.0023,
+      "step": 1034
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.3572381587725146e-05,
+      "loss": 0.0327,
+      "step": 1035
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.355570380253503e-05,
+      "loss": 0.0011,
+      "step": 1036
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.35390260173449e-05,
+      "loss": 0.0006,
+      "step": 1037
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.352234823215477e-05,
+      "loss": 0.179,
+      "step": 1038
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.350567044696464e-05,
+      "loss": 0.0015,
+      "step": 1039
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.348899266177452e-05,
+      "loss": 0.0012,
+      "step": 1040
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.3472314876584395e-05,
+      "loss": 0.3368,
+      "step": 1041
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.3455637091394265e-05,
+      "loss": 0.0007,
+      "step": 1042
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.3438959306204136e-05,
+      "loss": 0.0007,
+      "step": 1043
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.342228152101401e-05,
+      "loss": 0.3316,
+      "step": 1044
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.340560373582388e-05,
+      "loss": 0.0009,
+      "step": 1045
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.338892595063376e-05,
+      "loss": 0.0029,
+      "step": 1046
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.337224816544363e-05,
+      "loss": 0.6129,
+      "step": 1047
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.33555703802535e-05,
+      "loss": 0.2436,
+      "step": 1048
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.333889259506338e-05,
+      "loss": 0.0008,
+      "step": 1049
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.332221480987325e-05,
+      "loss": 0.0137,
+      "step": 1050
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.3305537024683126e-05,
+      "loss": 0.0085,
+      "step": 1051
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.3288859239492996e-05,
+      "loss": 0.0901,
+      "step": 1052
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.327218145430287e-05,
+      "loss": 0.0525,
+      "step": 1053
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.3255503669112743e-05,
+      "loss": 0.0037,
+      "step": 1054
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.3238825883922614e-05,
+      "loss": 0.0171,
+      "step": 1055
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.322214809873249e-05,
+      "loss": 0.0008,
+      "step": 1056
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.320547031354236e-05,
+      "loss": 0.0009,
+      "step": 1057
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.318879252835224e-05,
+      "loss": 0.0046,
+      "step": 1058
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.317211474316211e-05,
+      "loss": 0.0014,
+      "step": 1059
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.315543695797198e-05,
+      "loss": 0.0023,
+      "step": 1060
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.3138759172781856e-05,
+      "loss": 0.0012,
+      "step": 1061
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.3122081387591733e-05,
+      "loss": 0.0007,
+      "step": 1062
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.3105403602401604e-05,
+      "loss": 0.2839,
+      "step": 1063
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.3088725817211474e-05,
+      "loss": 0.0007,
+      "step": 1064
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.3072048032021344e-05,
+      "loss": 0.0013,
+      "step": 1065
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.305537024683122e-05,
+      "loss": 0.3601,
+      "step": 1066
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.30386924616411e-05,
+      "loss": 0.0006,
+      "step": 1067
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.302201467645097e-05,
+      "loss": 0.0011,
+      "step": 1068
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.300533689126084e-05,
+      "loss": 0.1935,
+      "step": 1069
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.298865910607071e-05,
+      "loss": 0.0007,
+      "step": 1070
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.2971981320880594e-05,
+      "loss": 0.445,
+      "step": 1071
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.2955303535690464e-05,
+      "loss": 0.0007,
+      "step": 1072
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.2938625750500334e-05,
+      "loss": 0.0023,
+      "step": 1073
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.2921947965310205e-05,
+      "loss": 0.0032,
+      "step": 1074
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.290527018012008e-05,
+      "loss": 0.0015,
+      "step": 1075
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.288859239492996e-05,
+      "loss": 0.002,
+      "step": 1076
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.287191460973983e-05,
+      "loss": 0.0013,
+      "step": 1077
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.28552368245497e-05,
+      "loss": 0.0022,
+      "step": 1078
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.283855903935957e-05,
+      "loss": 0.0015,
+      "step": 1079
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.282188125416945e-05,
+      "loss": 0.2998,
+      "step": 1080
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.2805203468979324e-05,
+      "loss": 0.0014,
+      "step": 1081
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.2788525683789195e-05,
+      "loss": 0.0339,
+      "step": 1082
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.2771847898599065e-05,
+      "loss": 0.0015,
+      "step": 1083
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.275517011340894e-05,
+      "loss": 0.0559,
+      "step": 1084
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.273849232821881e-05,
+      "loss": 0.0454,
+      "step": 1085
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.272181454302869e-05,
+      "loss": 0.0017,
+      "step": 1086
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.270513675783856e-05,
+      "loss": 0.0013,
+      "step": 1087
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.268845897264843e-05,
+      "loss": 0.2128,
+      "step": 1088
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.267178118745831e-05,
+      "loss": 0.0012,
+      "step": 1089
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.265510340226818e-05,
+      "loss": 0.0021,
+      "step": 1090
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.2638425617078055e-05,
+      "loss": 0.0015,
+      "step": 1091
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.2621747831887925e-05,
+      "loss": 0.0064,
+      "step": 1092
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.26050700466978e-05,
+      "loss": 0.0042,
+      "step": 1093
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.258839226150767e-05,
+      "loss": 0.0016,
+      "step": 1094
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.257171447631754e-05,
+      "loss": 0.0015,
+      "step": 1095
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.255503669112742e-05,
+      "loss": 0.001,
+      "step": 1096
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.253835890593729e-05,
+      "loss": 0.0013,
+      "step": 1097
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.252168112074717e-05,
+      "loss": 0.0011,
+      "step": 1098
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.250500333555704e-05,
+      "loss": 0.0129,
+      "step": 1099
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.248832555036691e-05,
+      "loss": 0.0016,
+      "step": 1100
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.2471647765176786e-05,
+      "loss": 0.0831,
+      "step": 1101
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.245496997998666e-05,
+      "loss": 0.0047,
+      "step": 1102
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.243829219479653e-05,
+      "loss": 0.002,
+      "step": 1103
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.2421614409606403e-05,
+      "loss": 0.0009,
+      "step": 1104
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.2404936624416274e-05,
+      "loss": 0.0009,
+      "step": 1105
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.238825883922616e-05,
+      "loss": 0.304,
+      "step": 1106
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.237158105403603e-05,
+      "loss": 0.001,
+      "step": 1107
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.23549032688459e-05,
+      "loss": 0.0011,
+      "step": 1108
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.233822548365577e-05,
+      "loss": 0.0014,
+      "step": 1109
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.2321547698465646e-05,
+      "loss": 0.0007,
+      "step": 1110
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.230486991327552e-05,
+      "loss": 0.0034,
+      "step": 1111
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.228819212808539e-05,
+      "loss": 0.0007,
+      "step": 1112
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.2271514342895264e-05,
+      "loss": 0.0008,
+      "step": 1113
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.2254836557705134e-05,
+      "loss": 0.0011,
+      "step": 1114
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.223815877251501e-05,
+      "loss": 0.0025,
+      "step": 1115
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.222148098732489e-05,
+      "loss": 0.0017,
+      "step": 1116
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.220480320213476e-05,
+      "loss": 0.0009,
+      "step": 1117
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.218812541694463e-05,
+      "loss": 0.0719,
+      "step": 1118
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.2171447631754506e-05,
+      "loss": 0.0006,
+      "step": 1119
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.2154769846564376e-05,
+      "loss": 0.0012,
+      "step": 1120
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.2138092061374254e-05,
+      "loss": 0.0016,
+      "step": 1121
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.2121414276184124e-05,
+      "loss": 0.0006,
+      "step": 1122
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.2104736490993994e-05,
+      "loss": 0.0007,
+      "step": 1123
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.208805870580387e-05,
+      "loss": 0.0006,
+      "step": 1124
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.207138092061374e-05,
+      "loss": 0.3019,
+      "step": 1125
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.205470313542362e-05,
+      "loss": 0.0007,
+      "step": 1126
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.203802535023349e-05,
+      "loss": 0.0006,
+      "step": 1127
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.2021347565043366e-05,
+      "loss": 0.0006,
+      "step": 1128
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.200466977985324e-05,
+      "loss": 0.008,
+      "step": 1129
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.198799199466311e-05,
+      "loss": 0.0007,
+      "step": 1130
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.1971314209472984e-05,
+      "loss": 0.0005,
+      "step": 1131
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.1954636424282855e-05,
+      "loss": 0.0005,
+      "step": 1132
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.193795863909273e-05,
+      "loss": 0.0006,
+      "step": 1133
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.19212808539026e-05,
+      "loss": 0.0007,
+      "step": 1134
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.190460306871247e-05,
+      "loss": 0.1846,
+      "step": 1135
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.188792528352235e-05,
+      "loss": 0.3874,
+      "step": 1136
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.187124749833223e-05,
+      "loss": 0.007,
+      "step": 1137
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.18545697131421e-05,
+      "loss": 0.0022,
+      "step": 1138
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.183789192795197e-05,
+      "loss": 0.0004,
+      "step": 1139
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.182121414276184e-05,
+      "loss": 0.0744,
+      "step": 1140
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.1804536357571715e-05,
+      "loss": 0.3236,
+      "step": 1141
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.178785857238159e-05,
+      "loss": 0.0006,
+      "step": 1142
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.177118078719146e-05,
+      "loss": 0.0008,
+      "step": 1143
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.175450300200133e-05,
+      "loss": 0.0016,
+      "step": 1144
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.173782521681121e-05,
+      "loss": 0.0005,
+      "step": 1145
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.172114743162109e-05,
+      "loss": 0.499,
+      "step": 1146
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.170446964643096e-05,
+      "loss": 0.0007,
+      "step": 1147
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.168779186124083e-05,
+      "loss": 0.0005,
+      "step": 1148
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.16711140760507e-05,
+      "loss": 0.0024,
+      "step": 1149
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.1654436290860575e-05,
+      "loss": 0.001,
+      "step": 1150
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.163775850567045e-05,
+      "loss": 0.0008,
+      "step": 1151
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.162108072048032e-05,
+      "loss": 0.0018,
+      "step": 1152
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.160440293529019e-05,
+      "loss": 0.5355,
+      "step": 1153
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.158772515010007e-05,
+      "loss": 0.003,
+      "step": 1154
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.157104736490994e-05,
+      "loss": 0.0011,
+      "step": 1155
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.155436957971982e-05,
+      "loss": 0.0015,
+      "step": 1156
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.153769179452969e-05,
+      "loss": 0.0019,
+      "step": 1157
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.152101400933956e-05,
+      "loss": 0.0011,
+      "step": 1158
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.1504336224149435e-05,
+      "loss": 0.2264,
+      "step": 1159
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.1487658438959306e-05,
+      "loss": 0.0023,
+      "step": 1160
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.147098065376918e-05,
+      "loss": 0.0116,
+      "step": 1161
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.145430286857905e-05,
+      "loss": 0.0044,
+      "step": 1162
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.143762508338893e-05,
+      "loss": 0.0022,
+      "step": 1163
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.14209472981988e-05,
+      "loss": 0.0038,
+      "step": 1164
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.140426951300867e-05,
+      "loss": 0.0048,
+      "step": 1165
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.138759172781855e-05,
+      "loss": 0.2036,
+      "step": 1166
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.137091394262842e-05,
+      "loss": 0.001,
+      "step": 1167
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.1354236157438296e-05,
+      "loss": 0.3823,
+      "step": 1168
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.1337558372248166e-05,
+      "loss": 0.003,
+      "step": 1169
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.1320880587058036e-05,
+      "loss": 0.1421,
+      "step": 1170
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.1304202801867914e-05,
+      "loss": 0.0023,
+      "step": 1171
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.128752501667779e-05,
+      "loss": 0.0071,
+      "step": 1172
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.127084723148766e-05,
+      "loss": 0.0045,
+      "step": 1173
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.125416944629753e-05,
+      "loss": 0.001,
+      "step": 1174
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.12374916611074e-05,
+      "loss": 0.0008,
+      "step": 1175
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.122081387591728e-05,
+      "loss": 0.0217,
+      "step": 1176
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.1204136090727156e-05,
+      "loss": 0.0005,
+      "step": 1177
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.1187458305537026e-05,
+      "loss": 0.261,
+      "step": 1178
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.11707805203469e-05,
+      "loss": 0.0009,
+      "step": 1179
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.1154102735156774e-05,
+      "loss": 0.0005,
+      "step": 1180
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.113742494996665e-05,
+      "loss": 0.0005,
+      "step": 1181
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.112074716477652e-05,
+      "loss": 0.0005,
+      "step": 1182
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.110406937958639e-05,
+      "loss": 0.0008,
+      "step": 1183
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.108739159439626e-05,
+      "loss": 0.0007,
+      "step": 1184
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.107071380920614e-05,
+      "loss": 0.2512,
+      "step": 1185
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.1054036024016016e-05,
+      "loss": 0.0004,
+      "step": 1186
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.1037358238825887e-05,
+      "loss": 0.0008,
+      "step": 1187
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.102068045363576e-05,
+      "loss": 0.0775,
+      "step": 1188
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.1004002668445634e-05,
+      "loss": 0.0009,
+      "step": 1189
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.0987324883255504e-05,
+      "loss": 0.0005,
+      "step": 1190
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.097064709806538e-05,
+      "loss": 0.0013,
+      "step": 1191
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.095396931287525e-05,
+      "loss": 0.0013,
+      "step": 1192
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.093729152768512e-05,
+      "loss": 0.0005,
+      "step": 1193
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.0920613742495e-05,
+      "loss": 0.0013,
+      "step": 1194
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.090393595730487e-05,
+      "loss": 0.0005,
+      "step": 1195
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.088725817211475e-05,
+      "loss": 0.0005,
+      "step": 1196
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.087058038692462e-05,
+      "loss": 0.0005,
+      "step": 1197
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.0853902601734494e-05,
+      "loss": 0.0008,
+      "step": 1198
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.0837224816544365e-05,
+      "loss": 0.0006,
+      "step": 1199
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.0820547031354235e-05,
+      "loss": 0.0023,
+      "step": 1200
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.080386924616411e-05,
+      "loss": 0.0007,
+      "step": 1201
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.078719146097398e-05,
+      "loss": 0.0005,
+      "step": 1202
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.077051367578386e-05,
+      "loss": 0.0005,
+      "step": 1203
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.075383589059373e-05,
+      "loss": 0.0041,
+      "step": 1204
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.07371581054036e-05,
+      "loss": 0.0004,
+      "step": 1205
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.072048032021348e-05,
+      "loss": 0.0008,
+      "step": 1206
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.0703802535023355e-05,
+      "loss": 0.0033,
+      "step": 1207
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.0687124749833225e-05,
+      "loss": 0.0009,
+      "step": 1208
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.0670446964643095e-05,
+      "loss": 0.0006,
+      "step": 1209
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.0653769179452966e-05,
+      "loss": 0.0004,
+      "step": 1210
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.063709139426284e-05,
+      "loss": 0.0004,
+      "step": 1211
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.062041360907272e-05,
+      "loss": 0.0004,
+      "step": 1212
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.060373582388259e-05,
+      "loss": 0.0004,
+      "step": 1213
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.058705803869246e-05,
+      "loss": 0.0004,
+      "step": 1214
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.057038025350233e-05,
+      "loss": 0.0003,
+      "step": 1215
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.0553702468312215e-05,
+      "loss": 0.0003,
+      "step": 1216
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.0537024683122085e-05,
+      "loss": 0.0003,
+      "step": 1217
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.0520346897931956e-05,
+      "loss": 0.0004,
+      "step": 1218
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.050366911274183e-05,
+      "loss": 0.0005,
+      "step": 1219
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.04869913275517e-05,
+      "loss": 0.001,
+      "step": 1220
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.0470313542361577e-05,
+      "loss": 0.0004,
+      "step": 1221
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.045363575717145e-05,
+      "loss": 0.0003,
+      "step": 1222
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.043695797198132e-05,
+      "loss": 0.0003,
+      "step": 1223
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.0420280186791195e-05,
+      "loss": 0.0003,
+      "step": 1224
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.040360240160107e-05,
+      "loss": 0.5147,
+      "step": 1225
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.0386924616410946e-05,
+      "loss": 0.0007,
+      "step": 1226
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.0370246831220816e-05,
+      "loss": 0.0007,
+      "step": 1227
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.035356904603069e-05,
+      "loss": 0.1189,
+      "step": 1228
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.033689126084056e-05,
+      "loss": 0.0008,
+      "step": 1229
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.0320213475650434e-05,
+      "loss": 0.0003,
+      "step": 1230
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.030353569046031e-05,
+      "loss": 0.0004,
+      "step": 1231
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.028685790527018e-05,
+      "loss": 0.0003,
+      "step": 1232
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.0270180120080055e-05,
+      "loss": 0.0003,
+      "step": 1233
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.025350233488993e-05,
+      "loss": 0.0003,
+      "step": 1234
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.02368245496998e-05,
+      "loss": 0.0003,
+      "step": 1235
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.0220146764509676e-05,
+      "loss": 0.0004,
+      "step": 1236
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.020346897931955e-05,
+      "loss": 0.0003,
+      "step": 1237
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.018679119412942e-05,
+      "loss": 0.0003,
+      "step": 1238
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.0170113408939294e-05,
+      "loss": 0.7741,
+      "step": 1239
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.0153435623749164e-05,
+      "loss": 0.0003,
+      "step": 1240
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.0136757838559045e-05,
+      "loss": 0.0004,
+      "step": 1241
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.0120080053368915e-05,
+      "loss": 0.0178,
+      "step": 1242
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.010340226817879e-05,
+      "loss": 0.378,
+      "step": 1243
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.008672448298866e-05,
+      "loss": 0.3657,
+      "step": 1244
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.0070046697798533e-05,
+      "loss": 0.0004,
+      "step": 1245
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.005336891260841e-05,
+      "loss": 0.0003,
+      "step": 1246
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.003669112741828e-05,
+      "loss": 0.0003,
+      "step": 1247
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.0020013342228154e-05,
+      "loss": 0.0004,
+      "step": 1248
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.0003335557038025e-05,
+      "loss": 0.0003,
+      "step": 1249
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.99866577718479e-05,
+      "loss": 0.0009,
+      "step": 1250
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.9969979986657776e-05,
+      "loss": 0.0003,
+      "step": 1251
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.995330220146765e-05,
+      "loss": 0.0011,
+      "step": 1252
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.993662441627752e-05,
+      "loss": 0.0003,
+      "step": 1253
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.9919946631087393e-05,
+      "loss": 0.0042,
+      "step": 1254
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.9903268845897264e-05,
+      "loss": 0.0004,
+      "step": 1255
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.988659106070714e-05,
+      "loss": 0.0003,
+      "step": 1256
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.9869913275517015e-05,
+      "loss": 0.0003,
+      "step": 1257
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.9853235490326885e-05,
+      "loss": 0.0003,
+      "step": 1258
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.983655770513676e-05,
+      "loss": 0.0003,
+      "step": 1259
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.981987991994663e-05,
+      "loss": 0.0003,
+      "step": 1260
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.980320213475651e-05,
+      "loss": 0.0003,
+      "step": 1261
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.978652434956638e-05,
+      "loss": 0.0003,
+      "step": 1262
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.9769846564376254e-05,
+      "loss": 0.0003,
+      "step": 1263
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.9753168779186124e-05,
+      "loss": 0.0003,
+      "step": 1264
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.9736490993995998e-05,
+      "loss": 0.0015,
+      "step": 1265
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.9719813208805875e-05,
+      "loss": 0.0005,
+      "step": 1266
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.9703135423615745e-05,
+      "loss": 0.0003,
+      "step": 1267
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.968645763842562e-05,
+      "loss": 0.4433,
+      "step": 1268
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.966977985323549e-05,
+      "loss": 0.0003,
+      "step": 1269
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.9653102068045363e-05,
+      "loss": 0.4709,
+      "step": 1270
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.963642428285524e-05,
+      "loss": 0.0025,
+      "step": 1271
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.9619746497665114e-05,
+      "loss": 0.0004,
+      "step": 1272
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.9603068712474984e-05,
+      "loss": 0.4103,
+      "step": 1273
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.9586390927284858e-05,
+      "loss": 0.0004,
+      "step": 1274
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.956971314209473e-05,
+      "loss": 0.0454,
+      "step": 1275
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.9553035356904605e-05,
+      "loss": 0.0011,
+      "step": 1276
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.953635757171448e-05,
+      "loss": 0.0006,
+      "step": 1277
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.9519679786524353e-05,
+      "loss": 0.2445,
+      "step": 1278
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.9503002001334223e-05,
+      "loss": 0.0017,
+      "step": 1279
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.9486324216144097e-05,
+      "loss": 0.0011,
+      "step": 1280
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.9469646430953974e-05,
+      "loss": 0.0022,
+      "step": 1281
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.9452968645763845e-05,
+      "loss": 0.0011,
+      "step": 1282
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.9436290860573718e-05,
+      "loss": 0.2002,
+      "step": 1283
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.941961307538359e-05,
+      "loss": 0.0022,
+      "step": 1284
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.9402935290193462e-05,
+      "loss": 0.001,
+      "step": 1285
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.938625750500334e-05,
+      "loss": 0.0017,
+      "step": 1286
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.9369579719813213e-05,
+      "loss": 0.0071,
+      "step": 1287
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.9352901934623084e-05,
+      "loss": 0.001,
+      "step": 1288
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.9336224149432957e-05,
+      "loss": 0.4633,
+      "step": 1289
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.9319546364242828e-05,
+      "loss": 0.0008,
+      "step": 1290
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.9302868579052705e-05,
+      "loss": 0.0008,
+      "step": 1291
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.928619079386258e-05,
+      "loss": 0.0012,
+      "step": 1292
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.926951300867245e-05,
+      "loss": 0.0007,
+      "step": 1293
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.9252835223482323e-05,
+      "loss": 0.5621,
+      "step": 1294
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.9236157438292193e-05,
+      "loss": 0.0007,
+      "step": 1295
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.9219479653102074e-05,
+      "loss": 0.0302,
+      "step": 1296
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.9202801867911944e-05,
+      "loss": 0.0086,
+      "step": 1297
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.9186124082721818e-05,
+      "loss": 0.0009,
+      "step": 1298
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.9169446297531688e-05,
+      "loss": 0.0978,
+      "step": 1299
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.9152768512341562e-05,
+      "loss": 0.001,
+      "step": 1300
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.913609072715144e-05,
+      "loss": 0.0008,
+      "step": 1301
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.911941294196131e-05,
+      "loss": 0.0014,
+      "step": 1302
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.9102735156771183e-05,
+      "loss": 0.001,
+      "step": 1303
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.9086057371581053e-05,
+      "loss": 0.0015,
+      "step": 1304
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.9069379586390927e-05,
+      "loss": 0.0012,
+      "step": 1305
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.9052701801200804e-05,
+      "loss": 0.0011,
+      "step": 1306
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.9036024016010678e-05,
+      "loss": 0.3579,
+      "step": 1307
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.9019346230820548e-05,
+      "loss": 0.0006,
+      "step": 1308
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.9002668445630422e-05,
+      "loss": 0.0006,
+      "step": 1309
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.8985990660440292e-05,
+      "loss": 0.0007,
+      "step": 1310
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.896931287525017e-05,
+      "loss": 0.0007,
+      "step": 1311
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.8952635090060043e-05,
+      "loss": 0.0024,
+      "step": 1312
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.8935957304869914e-05,
+      "loss": 0.0864,
+      "step": 1313
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.8919279519679787e-05,
+      "loss": 0.0011,
+      "step": 1314
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.8902601734489658e-05,
+      "loss": 0.0005,
+      "step": 1315
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.8885923949299538e-05,
+      "loss": 0.0009,
+      "step": 1316
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.886924616410941e-05,
+      "loss": 0.001,
+      "step": 1317
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.8852568378919282e-05,
+      "loss": 0.0011,
+      "step": 1318
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.8835890593729153e-05,
+      "loss": 0.0925,
+      "step": 1319
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.8819212808539026e-05,
+      "loss": 0.0008,
+      "step": 1320
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.8802535023348903e-05,
+      "loss": 0.001,
+      "step": 1321
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.8785857238158774e-05,
+      "loss": 0.0402,
+      "step": 1322
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.8769179452968648e-05,
+      "loss": 0.0237,
+      "step": 1323
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.875250166777852e-05,
+      "loss": 0.0094,
+      "step": 1324
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.873582388258839e-05,
+      "loss": 0.003,
+      "step": 1325
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.871914609739827e-05,
+      "loss": 0.0005,
+      "step": 1326
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.8702468312208143e-05,
+      "loss": 0.0004,
+      "step": 1327
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.8685790527018013e-05,
+      "loss": 0.0005,
+      "step": 1328
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.8669112741827887e-05,
+      "loss": 0.6644,
+      "step": 1329
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.8652434956637757e-05,
+      "loss": 0.0007,
+      "step": 1330
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.8635757171447634e-05,
+      "loss": 0.0141,
+      "step": 1331
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.8619079386257508e-05,
+      "loss": 0.0894,
+      "step": 1332
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.860240160106738e-05,
+      "loss": 0.0006,
+      "step": 1333
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.8585723815877252e-05,
+      "loss": 0.0008,
+      "step": 1334
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.8569046030687126e-05,
+      "loss": 0.001,
+      "step": 1335
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.8552368245497003e-05,
+      "loss": 0.0011,
+      "step": 1336
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.8535690460306873e-05,
+      "loss": 0.0006,
+      "step": 1337
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.8519012675116747e-05,
+      "loss": 0.0004,
+      "step": 1338
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.8502334889926617e-05,
+      "loss": 0.0007,
+      "step": 1339
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.848565710473649e-05,
+      "loss": 0.0006,
+      "step": 1340
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.8468979319546368e-05,
+      "loss": 0.0048,
+      "step": 1341
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.8452301534356242e-05,
+      "loss": 0.0005,
+      "step": 1342
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.8435623749166112e-05,
+      "loss": 0.0004,
+      "step": 1343
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.8418945963975986e-05,
+      "loss": 0.0011,
+      "step": 1344
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.8402268178785856e-05,
+      "loss": 0.0004,
+      "step": 1345
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.8385590393595733e-05,
+      "loss": 0.0029,
+      "step": 1346
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.8368912608405607e-05,
+      "loss": 0.0004,
+      "step": 1347
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.8352234823215478e-05,
+      "loss": 0.0008,
+      "step": 1348
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.833555703802535e-05,
+      "loss": 0.0013,
+      "step": 1349
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.831887925283522e-05,
+      "loss": 0.0006,
+      "step": 1350
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.8302201467645102e-05,
+      "loss": 0.0004,
+      "step": 1351
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.8285523682454973e-05,
+      "loss": 0.0004,
+      "step": 1352
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.8268845897264846e-05,
+      "loss": 0.292,
+      "step": 1353
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.8252168112074717e-05,
+      "loss": 0.0005,
+      "step": 1354
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.823549032688459e-05,
+      "loss": 0.0021,
+      "step": 1355
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.8218812541694467e-05,
+      "loss": 0.0004,
+      "step": 1356
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.8202134756504338e-05,
+      "loss": 0.0016,
+      "step": 1357
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.818545697131421e-05,
+      "loss": 0.0003,
+      "step": 1358
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.8168779186124082e-05,
+      "loss": 0.0024,
+      "step": 1359
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.8152101400933956e-05,
+      "loss": 0.0011,
+      "step": 1360
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.8135423615743833e-05,
+      "loss": 0.0003,
+      "step": 1361
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.8118745830553707e-05,
+      "loss": 0.0005,
+      "step": 1362
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.8102068045363577e-05,
+      "loss": 0.0362,
+      "step": 1363
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.808539026017345e-05,
+      "loss": 0.5146,
+      "step": 1364
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.806871247498332e-05,
+      "loss": 0.2625,
+      "step": 1365
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.8052034689793195e-05,
+      "loss": 0.0005,
+      "step": 1366
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.8035356904603072e-05,
+      "loss": 0.5093,
+      "step": 1367
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.8018679119412942e-05,
+      "loss": 0.0013,
+      "step": 1368
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.8002001334222816e-05,
+      "loss": 0.0012,
+      "step": 1369
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.798532354903269e-05,
+      "loss": 0.0035,
+      "step": 1370
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.796864576384256e-05,
+      "loss": 0.0019,
+      "step": 1371
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.7951967978652437e-05,
+      "loss": 0.0057,
+      "step": 1372
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.793529019346231e-05,
+      "loss": 0.0009,
+      "step": 1373
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.791861240827218e-05,
+      "loss": 0.0052,
+      "step": 1374
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.7901934623082055e-05,
+      "loss": 0.0012,
+      "step": 1375
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.7885256837891925e-05,
+      "loss": 0.0017,
+      "step": 1376
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.7868579052701806e-05,
+      "loss": 0.1493,
+      "step": 1377
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.7851901267511676e-05,
+      "loss": 0.0019,
+      "step": 1378
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.783522348232155e-05,
+      "loss": 0.0018,
+      "step": 1379
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.781854569713142e-05,
+      "loss": 0.0013,
+      "step": 1380
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.7801867911941294e-05,
+      "loss": 0.0014,
+      "step": 1381
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.778519012675117e-05,
+      "loss": 0.0014,
+      "step": 1382
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.776851234156104e-05,
+      "loss": 0.0019,
+      "step": 1383
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.7751834556370915e-05,
+      "loss": 0.2027,
+      "step": 1384
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.7735156771180786e-05,
+      "loss": 0.0021,
+      "step": 1385
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.771847898599066e-05,
+      "loss": 0.1003,
+      "step": 1386
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.7701801200800537e-05,
+      "loss": 0.0013,
+      "step": 1387
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.768512341561041e-05,
+      "loss": 0.0018,
+      "step": 1388
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.766844563042028e-05,
+      "loss": 0.0018,
+      "step": 1389
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.7651767845230154e-05,
+      "loss": 0.0028,
+      "step": 1390
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.7635090060040025e-05,
+      "loss": 0.0019,
+      "step": 1391
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.7618412274849902e-05,
+      "loss": 0.0385,
+      "step": 1392
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.7601734489659776e-05,
+      "loss": 0.0015,
+      "step": 1393
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.7585056704469646e-05,
+      "loss": 0.1801,
+      "step": 1394
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.756837891927952e-05,
+      "loss": 0.0015,
+      "step": 1395
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.755170113408939e-05,
+      "loss": 0.0012,
+      "step": 1396
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.753502334889927e-05,
+      "loss": 0.0017,
+      "step": 1397
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.751834556370914e-05,
+      "loss": 0.0013,
+      "step": 1398
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.7501667778519015e-05,
+      "loss": 0.0015,
+      "step": 1399
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.7484989993328885e-05,
+      "loss": 0.0911,
+      "step": 1400
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.746831220813876e-05,
+      "loss": 0.0022,
+      "step": 1401
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.7451634422948636e-05,
+      "loss": 0.0018,
+      "step": 1402
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.7434956637758506e-05,
+      "loss": 0.0042,
+      "step": 1403
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.741827885256838e-05,
+      "loss": 0.0024,
+      "step": 1404
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.740160106737825e-05,
+      "loss": 0.0015,
+      "step": 1405
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.7384923282188124e-05,
+      "loss": 0.0013,
+      "step": 1406
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.7368245496998e-05,
+      "loss": 0.0211,
+      "step": 1407
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.7351567711807875e-05,
+      "loss": 0.0008,
+      "step": 1408
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.7334889926617745e-05,
+      "loss": 0.0026,
+      "step": 1409
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.731821214142762e-05,
+      "loss": 0.0018,
+      "step": 1410
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.730153435623749e-05,
+      "loss": 0.1578,
+      "step": 1411
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.7284856571047366e-05,
+      "loss": 0.0008,
+      "step": 1412
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.726817878585724e-05,
+      "loss": 0.0009,
+      "step": 1413
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.7251501000667114e-05,
+      "loss": 0.0008,
+      "step": 1414
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.7234823215476984e-05,
+      "loss": 0.0193,
+      "step": 1415
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.7218145430286858e-05,
+      "loss": 0.0101,
+      "step": 1416
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.7201467645096735e-05,
+      "loss": 0.0024,
+      "step": 1417
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.7184789859906606e-05,
+      "loss": 0.0006,
+      "step": 1418
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.716811207471648e-05,
+      "loss": 0.0008,
+      "step": 1419
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.715143428952635e-05,
+      "loss": 0.0007,
+      "step": 1420
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.7134756504336223e-05,
+      "loss": 0.3646,
+      "step": 1421
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.71180787191461e-05,
+      "loss": 0.0493,
+      "step": 1422
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.7101400933955974e-05,
+      "loss": 0.4855,
+      "step": 1423
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.7084723148765845e-05,
+      "loss": 0.3818,
+      "step": 1424
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.706804536357572e-05,
+      "loss": 0.0008,
+      "step": 1425
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.705136757838559e-05,
+      "loss": 0.0008,
+      "step": 1426
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.7034689793195466e-05,
+      "loss": 0.0018,
+      "step": 1427
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.701801200800534e-05,
+      "loss": 0.0028,
+      "step": 1428
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.700133422281521e-05,
+      "loss": 0.0011,
+      "step": 1429
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.6984656437625084e-05,
+      "loss": 0.0009,
+      "step": 1430
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.6967978652434954e-05,
+      "loss": 0.0013,
+      "step": 1431
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.6951300867244835e-05,
+      "loss": 0.0023,
+      "step": 1432
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.6934623082054705e-05,
+      "loss": 0.3609,
+      "step": 1433
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.691794529686458e-05,
+      "loss": 0.0021,
+      "step": 1434
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.690126751167445e-05,
+      "loss": 0.001,
+      "step": 1435
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.6884589726484323e-05,
+      "loss": 0.0009,
+      "step": 1436
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.68679119412942e-05,
+      "loss": 0.0012,
+      "step": 1437
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.685123415610407e-05,
+      "loss": 0.0077,
+      "step": 1438
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.6834556370913944e-05,
+      "loss": 0.0011,
+      "step": 1439
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.6817878585723814e-05,
+      "loss": 0.0015,
+      "step": 1440
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.6801200800533688e-05,
+      "loss": 0.0008,
+      "step": 1441
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.6784523015343565e-05,
+      "loss": 0.0017,
+      "step": 1442
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.676784523015344e-05,
+      "loss": 0.0017,
+      "step": 1443
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.675116744496331e-05,
+      "loss": 0.4123,
+      "step": 1444
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.6734489659773183e-05,
+      "loss": 0.0009,
+      "step": 1445
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.6717811874583053e-05,
+      "loss": 0.0029,
+      "step": 1446
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.670113408939293e-05,
+      "loss": 0.0187,
+      "step": 1447
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.6684456304202804e-05,
+      "loss": 0.0009,
+      "step": 1448
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.6667778519012675e-05,
+      "loss": 0.0007,
+      "step": 1449
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.6651100733822548e-05,
+      "loss": 0.0013,
+      "step": 1450
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.6634422948632422e-05,
+      "loss": 0.0055,
+      "step": 1451
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.66177451634423e-05,
+      "loss": 0.2949,
+      "step": 1452
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.660106737825217e-05,
+      "loss": 0.0123,
+      "step": 1453
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.6584389593062043e-05,
+      "loss": 0.0008,
+      "step": 1454
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.6567711807871914e-05,
+      "loss": 0.001,
+      "step": 1455
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.6551034022681787e-05,
+      "loss": 0.0072,
+      "step": 1456
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.6534356237491664e-05,
+      "loss": 0.0009,
+      "step": 1457
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.6517678452301535e-05,
+      "loss": 0.0006,
+      "step": 1458
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.650100066711141e-05,
+      "loss": 0.0007,
+      "step": 1459
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.6484322881921282e-05,
+      "loss": 0.8522,
+      "step": 1460
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.6467645096731153e-05,
+      "loss": 0.0007,
+      "step": 1461
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.645096731154103e-05,
+      "loss": 0.0005,
+      "step": 1462
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.6434289526350904e-05,
+      "loss": 0.0018,
+      "step": 1463
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.6417611741160774e-05,
+      "loss": 0.0006,
+      "step": 1464
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.6400933955970648e-05,
+      "loss": 0.0007,
+      "step": 1465
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.6384256170780518e-05,
+      "loss": 0.0421,
+      "step": 1466
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.63675783855904e-05,
+      "loss": 0.0009,
+      "step": 1467
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.635090060040027e-05,
+      "loss": 0.0006,
+      "step": 1468
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.6334222815210143e-05,
+      "loss": 0.0008,
+      "step": 1469
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.6317545030020013e-05,
+      "loss": 0.0008,
+      "step": 1470
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.6300867244829887e-05,
+      "loss": 0.3614,
+      "step": 1471
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.6284189459639764e-05,
+      "loss": 0.0006,
+      "step": 1472
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.6267511674449634e-05,
+      "loss": 0.0012,
+      "step": 1473
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.6250833889259508e-05,
+      "loss": 0.0036,
+      "step": 1474
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.6234156104069378e-05,
+      "loss": 0.0008,
+      "step": 1475
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.6217478318879252e-05,
+      "loss": 0.0012,
+      "step": 1476
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.620080053368913e-05,
+      "loss": 0.0007,
+      "step": 1477
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.6184122748499003e-05,
+      "loss": 0.0007,
+      "step": 1478
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.6167444963308873e-05,
+      "loss": 0.0008,
+      "step": 1479
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.6150767178118747e-05,
+      "loss": 0.0007,
+      "step": 1480
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.6134089392928617e-05,
+      "loss": 0.0006,
+      "step": 1481
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.6117411607738494e-05,
+      "loss": 0.0011,
+      "step": 1482
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.6100733822548368e-05,
+      "loss": 0.0007,
+      "step": 1483
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.608405603735824e-05,
+      "loss": 0.0006,
+      "step": 1484
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.6067378252168112e-05,
+      "loss": 0.0006,
+      "step": 1485
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.6050700466977983e-05,
+      "loss": 0.0008,
+      "step": 1486
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.6034022681787863e-05,
+      "loss": 0.0006,
+      "step": 1487
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.6017344896597734e-05,
+      "loss": 0.0005,
+      "step": 1488
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.6000667111407607e-05,
+      "loss": 0.0006,
+      "step": 1489
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.5983989326217478e-05,
+      "loss": 0.0012,
+      "step": 1490
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.596731154102735e-05,
+      "loss": 0.0007,
+      "step": 1491
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.595063375583723e-05,
+      "loss": 0.0005,
+      "step": 1492
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.59339559706471e-05,
+      "loss": 0.6162,
+      "step": 1493
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.5917278185456973e-05,
+      "loss": 0.0006,
+      "step": 1494
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.5900600400266843e-05,
+      "loss": 0.2499,
+      "step": 1495
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.5883922615076717e-05,
+      "loss": 0.0004,
+      "step": 1496
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.5867244829886594e-05,
+      "loss": 0.0004,
+      "step": 1497
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.5850567044696468e-05,
+      "loss": 0.0004,
+      "step": 1498
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.5833889259506338e-05,
+      "loss": 0.0007,
+      "step": 1499
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.581721147431621e-05,
+      "loss": 0.445,
+      "step": 1500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.5800533689126082e-05,
+      "loss": 0.2245,
+      "step": 1501
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.578385590393596e-05,
+      "loss": 0.0008,
+      "step": 1502
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.5767178118745833e-05,
+      "loss": 0.0016,
+      "step": 1503
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.5750500333555707e-05,
+      "loss": 0.0006,
+      "step": 1504
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.5733822548365577e-05,
+      "loss": 0.0008,
+      "step": 1505
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.571714476317545e-05,
+      "loss": 0.0005,
+      "step": 1506
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.5700466977985328e-05,
+      "loss": 0.0006,
+      "step": 1507
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.5683789192795198e-05,
+      "loss": 0.0006,
+      "step": 1508
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.5667111407605072e-05,
+      "loss": 0.0007,
+      "step": 1509
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.5650433622414942e-05,
+      "loss": 0.0006,
+      "step": 1510
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.5633755837224816e-05,
+      "loss": 0.0006,
+      "step": 1511
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.5617078052034693e-05,
+      "loss": 0.1923,
+      "step": 1512
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.5600400266844567e-05,
+      "loss": 0.1771,
+      "step": 1513
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.5583722481654437e-05,
+      "loss": 0.0007,
+      "step": 1514
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.556704469646431e-05,
+      "loss": 0.0013,
+      "step": 1515
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.555036691127418e-05,
+      "loss": 0.0006,
+      "step": 1516
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.553368912608406e-05,
+      "loss": 0.0007,
+      "step": 1517
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.5517011340893932e-05,
+      "loss": 0.0008,
+      "step": 1518
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.5500333555703803e-05,
+      "loss": 0.0009,
+      "step": 1519
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.5483655770513676e-05,
+      "loss": 0.0007,
+      "step": 1520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.5466977985323547e-05,
+      "loss": 0.0007,
+      "step": 1521
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.5450300200133427e-05,
+      "loss": 0.0006,
+      "step": 1522
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.5433622414943297e-05,
+      "loss": 0.0219,
+      "step": 1523
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.541694462975317e-05,
+      "loss": 0.0005,
+      "step": 1524
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.9922874958976042,
+      "eval_loss": 0.03057965449988842,
+      "eval_runtime": 105.8871,
+      "eval_samples_per_second": 57.552,
+      "eval_steps_per_second": 3.598,
+      "step": 1524
+    }
+  ],
+  "max_steps": 3048,
+  "num_train_epochs": 2,
+  "total_flos": 6376163426720160.0,
+  "trial_name": null,
+  "trial_params": null
+}