diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,24521 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.56,
+  "eval_steps": 500,
+  "global_step": 3500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 5e-09,
+      "loss": 1.1108,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1e-08,
+      "loss": 0.9378,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5e-08,
+      "loss": 1.2227,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 2e-08,
+      "loss": 0.8609,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5e-08,
+      "loss": 0.8715,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 3e-08,
+      "loss": 1.0075,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5e-08,
+      "loss": 0.9488,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 4e-08,
+      "loss": 1.1053,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5e-08,
+      "loss": 1.0498,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 5e-08,
+      "loss": 0.7981,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 5.4999999999999996e-08,
+      "loss": 1.0449,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 6e-08,
+      "loss": 0.9039,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 6.5e-08,
+      "loss": 0.9535,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 7e-08,
+      "loss": 1.0223,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 7.5e-08,
+      "loss": 1.1507,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 8e-08,
+      "loss": 1.0437,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 8.500000000000001e-08,
+      "loss": 1.0746,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 9e-08,
+      "loss": 1.0501,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 9.499999999999999e-08,
+      "loss": 1.0481,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1e-07,
+      "loss": 1.1637,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0499999999999999e-07,
+      "loss": 1.0122,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0999999999999999e-07,
+      "loss": 0.9446,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.15e-07,
+      "loss": 0.9718,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2e-07,
+      "loss": 1.0156,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.25e-07,
+      "loss": 1.0177,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3e-07,
+      "loss": 0.951,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.35e-07,
+      "loss": 0.9293,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4e-07,
+      "loss": 0.9501,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.45e-07,
+      "loss": 0.9664,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5e-07,
+      "loss": 1.1829,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.55e-07,
+      "loss": 1.1321,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6e-07,
+      "loss": 0.9481,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.65e-07,
+      "loss": 0.9958,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7000000000000001e-07,
+      "loss": 1.0273,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.75e-07,
+      "loss": 0.9583,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8e-07,
+      "loss": 0.9532,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.85e-07,
+      "loss": 0.8938,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8999999999999998e-07,
+      "loss": 1.1758,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9499999999999999e-07,
+      "loss": 1.0412,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 2e-07,
+      "loss": 0.9385,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0499999999999997e-07,
+      "loss": 0.9743,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0999999999999997e-07,
+      "loss": 1.1089,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1499999999999998e-07,
+      "loss": 0.9012,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1999999999999998e-07,
+      "loss": 1.1171,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 2.25e-07,
+      "loss": 0.9651,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3e-07,
+      "loss": 0.9881,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3499999999999997e-07,
+      "loss": 1.0264,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4e-07,
+      "loss": 0.9755,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 2.45e-07,
+      "loss": 0.9889,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5e-07,
+      "loss": 1.131,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 2.55e-07,
+      "loss": 0.8718,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6e-07,
+      "loss": 1.0497,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 2.65e-07,
+      "loss": 0.9653,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7e-07,
+      "loss": 0.7756,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 2.75e-07,
+      "loss": 0.9691,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8e-07,
+      "loss": 1.0708,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8499999999999997e-07,
+      "loss": 0.9941,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9e-07,
+      "loss": 1.0888,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 2.95e-07,
+      "loss": 1.0335,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 3e-07,
+      "loss": 1.1476,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 3.05e-07,
+      "loss": 1.0659,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1e-07,
+      "loss": 1.0062,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 3.15e-07,
+      "loss": 1.0126,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2e-07,
+      "loss": 1.0422,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 3.25e-07,
+      "loss": 1.0672,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3e-07,
+      "loss": 1.0779,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 3.35e-07,
+      "loss": 1.0264,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4000000000000003e-07,
+      "loss": 1.1484,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 3.45e-07,
+      "loss": 1.0174,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5e-07,
+      "loss": 0.9791,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 3.55e-07,
+      "loss": 0.8383,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6e-07,
+      "loss": 0.8979,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 3.65e-07,
+      "loss": 1.0383,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7e-07,
+      "loss": 0.8781,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 3.75e-07,
+      "loss": 0.8842,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7999999999999996e-07,
+      "loss": 1.027,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8499999999999997e-07,
+      "loss": 1.1345,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8999999999999997e-07,
+      "loss": 0.9651,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 3.95e-07,
+      "loss": 1.0664,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 4e-07,
+      "loss": 1.0735,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 4.05e-07,
+      "loss": 1.0544,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0999999999999994e-07,
+      "loss": 1.0329,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1499999999999994e-07,
+      "loss": 1.0057,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1999999999999995e-07,
+      "loss": 1.0078,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2499999999999995e-07,
+      "loss": 1.1718,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2999999999999996e-07,
+      "loss": 1.3407,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3499999999999996e-07,
+      "loss": 0.977,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3999999999999997e-07,
+      "loss": 1.1009,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 4.45e-07,
+      "loss": 1.0956,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5e-07,
+      "loss": 1.0901,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 4.55e-07,
+      "loss": 0.9896,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6e-07,
+      "loss": 0.9707,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 4.65e-07,
+      "loss": 1.0049,
+      "step": 93
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6999999999999995e-07,
+      "loss": 1.0691,
+      "step": 94
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7499999999999995e-07,
+      "loss": 1.048,
+      "step": 95
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.8e-07,
+      "loss": 0.9962,
+      "step": 96
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.85e-07,
+      "loss": 1.1493,
+      "step": 97
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.9e-07,
+      "loss": 1.0445,
+      "step": 98
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.95e-07,
+      "loss": 0.9708,
+      "step": 99
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 5e-07,
+      "loss": 0.8209,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999999673818355e-07,
+      "loss": 1.0859,
+      "step": 101
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999998695273509e-07,
+      "loss": 1.2999,
+      "step": 102
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999997064365714e-07,
+      "loss": 1.0456,
+      "step": 103
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999994781095397e-07,
+      "loss": 0.9679,
+      "step": 104
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999991845463154e-07,
+      "loss": 1.0527,
+      "step": 105
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.99998825746975e-07,
+      "loss": 1.0347,
+      "step": 106
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999984017116122e-07,
+      "loss": 1.0801,
+      "step": 107
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999979124403377e-07,
+      "loss": 1.2352,
+      "step": 108
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999973579332791e-07,
+      "loss": 0.9178,
+      "step": 109
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999967381905813e-07,
+      "loss": 0.9543,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999960532124056e-07,
+      "loss": 0.9818,
+      "step": 111
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999953029989311e-07,
+      "loss": 0.9451,
+      "step": 112
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999944875503534e-07,
+      "loss": 1.0938,
+      "step": 113
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999936068668854e-07,
+      "loss": 1.0041,
+      "step": 114
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999926609487567e-07,
+      "loss": 1.0892,
+      "step": 115
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999916497962144e-07,
+      "loss": 1.0252,
+      "step": 116
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999905734095222e-07,
+      "loss": 1.0766,
+      "step": 117
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.99989431788961e-07,
+      "loss": 1.158,
+      "step": 118
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999882249348287e-07,
+      "loss": 0.9395,
+      "step": 119
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999869528474402e-07,
+      "loss": 1.1848,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999856155271275e-07,
+      "loss": 0.9202,
+      "step": 121
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999842129742396e-07,
+      "loss": 0.9208,
+      "step": 122
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999827451891424e-07,
+      "loss": 0.9899,
+      "step": 123
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.99981212172219e-07,
+      "loss": 0.9138,
+      "step": 124
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999796139238694e-07,
+      "loss": 0.9894,
+      "step": 125
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999779504445104e-07,
+      "loss": 0.9745,
+      "step": 126
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999762217345765e-07,
+      "loss": 0.9744,
+      "step": 127
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999744277945187e-07,
+      "loss": 0.923,
+      "step": 128
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999725686248048e-07,
+      "loss": 1.1324,
+      "step": 129
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999706442259204e-07,
+      "loss": 0.9951,
+      "step": 130
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999686545983672e-07,
+      "loss": 1.1528,
+      "step": 131
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999665997426649e-07,
+      "loss": 1.0345,
+      "step": 132
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999644796593492e-07,
+      "loss": 0.9347,
+      "step": 133
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999622943489737e-07,
+      "loss": 1.068,
+      "step": 134
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999600438121084e-07,
+      "loss": 1.0121,
+      "step": 135
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999577280493407e-07,
+      "loss": 1.1516,
+      "step": 136
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999553470612748e-07,
+      "loss": 1.0839,
+      "step": 137
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999529008485321e-07,
+      "loss": 0.9805,
+      "step": 138
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999503894117509e-07,
+      "loss": 0.9992,
+      "step": 139
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999478127515865e-07,
+      "loss": 1.0346,
+      "step": 140
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999451708687113e-07,
+      "loss": 0.9175,
+      "step": 141
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999424637638148e-07,
+      "loss": 1.018,
+      "step": 142
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999396914376032e-07,
+      "loss": 0.9929,
+      "step": 143
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999368538908e-07,
+      "loss": 0.8938,
+      "step": 144
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999339511241457e-07,
+      "loss": 0.8234,
+      "step": 145
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999309831383978e-07,
+      "loss": 1.0533,
+      "step": 146
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999279499343306e-07,
+      "loss": 1.1269,
+      "step": 147
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999248515127357e-07,
+      "loss": 1.0035,
+      "step": 148
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999216878744217e-07,
+      "loss": 0.9474,
+      "step": 149
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.99918459020214e-07,
+      "loss": 1.1385,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999151649509553e-07,
+      "loss": 0.9895,
+      "step": 151
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999118056675052e-07,
+      "loss": 0.9931,
+      "step": 152
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.9990838117074e-07,
+      "loss": 0.9969,
+      "step": 153
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999048914615535e-07,
+      "loss": 1.2068,
+      "step": 154
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.999013365408563e-07,
+      "loss": 1.0389,
+      "step": 155
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 4.998977164095762e-07,
+      "loss": 1.2054,
+      "step": 156
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.998940310686575e-07,
+      "loss": 0.9756,
+      "step": 157
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.998902805190623e-07,
+      "loss": 1.0564,
+      "step": 158
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.998864647617689e-07,
+      "loss": 1.0957,
+      "step": 159
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.998825837977732e-07,
+      "loss": 0.993,
+      "step": 160
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.99878637628088e-07,
+      "loss": 1.0784,
+      "step": 161
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.998746262537428e-07,
+      "loss": 0.9139,
+      "step": 162
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.998705496757846e-07,
+      "loss": 0.9966,
+      "step": 163
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.998664078952769e-07,
+      "loss": 1.0368,
+      "step": 164
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.998622009133007e-07,
+      "loss": 1.019,
+      "step": 165
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.998579287309537e-07,
+      "loss": 1.0249,
+      "step": 166
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.998535913493508e-07,
+      "loss": 1.1294,
+      "step": 167
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.998491887696235e-07,
+      "loss": 0.9636,
+      "step": 168
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.998447209929211e-07,
+      "loss": 1.1231,
+      "step": 169
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.998401880204092e-07,
+      "loss": 0.9907,
+      "step": 170
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.998355898532706e-07,
+      "loss": 0.9157,
+      "step": 171
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.998309264927052e-07,
+      "loss": 0.9028,
+      "step": 172
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.998261979399299e-07,
+      "loss": 1.1656,
+      "step": 173
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.998214041961788e-07,
+      "loss": 1.0084,
+      "step": 174
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.998165452627024e-07,
+      "loss": 0.9744,
+      "step": 175
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.99811621140769e-07,
+      "loss": 1.0205,
+      "step": 176
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.998066318316634e-07,
+      "loss": 0.955,
+      "step": 177
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.998015773366874e-07,
+      "loss": 1.0414,
+      "step": 178
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.997964576571601e-07,
+      "loss": 0.9526,
+      "step": 179
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.997912727944173e-07,
+      "loss": 0.9496,
+      "step": 180
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.997860227498122e-07,
+      "loss": 1.0763,
+      "step": 181
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.997807075247145e-07,
+      "loss": 1.0803,
+      "step": 182
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.997753271205115e-07,
+      "loss": 0.9938,
+      "step": 183
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.997698815386069e-07,
+      "loss": 0.9488,
+      "step": 184
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.997643707804218e-07,
+      "loss": 1.1208,
+      "step": 185
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.997587948473944e-07,
+      "loss": 1.0772,
+      "step": 186
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.997531537409794e-07,
+      "loss": 1.1029,
+      "step": 187
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.99747447462649e-07,
+      "loss": 0.9019,
+      "step": 188
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.997416760138921e-07,
+      "loss": 1.0554,
+      "step": 189
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.99735839396215e-07,
+      "loss": 1.0222,
+      "step": 190
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.997299376111404e-07,
+      "loss": 0.939,
+      "step": 191
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.997239706602086e-07,
+      "loss": 1.0052,
+      "step": 192
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.997179385449765e-07,
+      "loss": 0.9364,
+      "step": 193
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.997118412670182e-07,
+      "loss": 0.9319,
+      "step": 194
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.997056788279247e-07,
+      "loss": 0.8464,
+      "step": 195
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.996994512293041e-07,
+      "loss": 0.8623,
+      "step": 196
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.996931584727816e-07,
+      "loss": 0.8746,
+      "step": 197
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.996868005599991e-07,
+      "loss": 1.0098,
+      "step": 198
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.996803774926158e-07,
+      "loss": 0.9856,
+      "step": 199
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.996738892723074e-07,
+      "loss": 1.07,
+      "step": 200
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.996673359007675e-07,
+      "loss": 0.8634,
+      "step": 201
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.996607173797058e-07,
+      "loss": 0.9852,
+      "step": 202
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.996540337108496e-07,
+      "loss": 1.005,
+      "step": 203
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.996472848959428e-07,
+      "loss": 0.8518,
+      "step": 204
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.996404709367466e-07,
+      "loss": 0.9017,
+      "step": 205
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.99633591835039e-07,
+      "loss": 0.9915,
+      "step": 206
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.99626647592615e-07,
+      "loss": 1.1084,
+      "step": 207
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.996196382112868e-07,
+      "loss": 1.0512,
+      "step": 208
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.996125636928834e-07,
+      "loss": 0.965,
+      "step": 209
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.996054240392509e-07,
+      "loss": 1.0634,
+      "step": 210
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.995982192522523e-07,
+      "loss": 1.1211,
+      "step": 211
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.995909493337677e-07,
+      "loss": 1.081,
+      "step": 212
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.995836142856941e-07,
+      "loss": 1.1389,
+      "step": 213
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.995762141099456e-07,
+      "loss": 0.935,
+      "step": 214
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.995687488084532e-07,
+      "loss": 1.007,
+      "step": 215
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.995612183831651e-07,
+      "loss": 0.9157,
+      "step": 216
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.99553622836046e-07,
+      "loss": 1.0768,
+      "step": 217
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 4.995459621690783e-07,
+      "loss": 1.0462,
+      "step": 218
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.995382363842608e-07,
+      "loss": 1.0269,
+      "step": 219
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.995304454836095e-07,
+      "loss": 1.0122,
+      "step": 220
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.995225894691574e-07,
+      "loss": 1.0307,
+      "step": 221
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.995146683429545e-07,
+      "loss": 0.9139,
+      "step": 222
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.995066821070679e-07,
+      "loss": 0.9914,
+      "step": 223
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.994986307635814e-07,
+      "loss": 1.1283,
+      "step": 224
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.994905143145961e-07,
+      "loss": 1.0613,
+      "step": 225
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.994823327622298e-07,
+      "loss": 1.0297,
+      "step": 226
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.994740861086176e-07,
+      "loss": 0.9454,
+      "step": 227
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.994657743559114e-07,
+      "loss": 1.0245,
+      "step": 228
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.994573975062799e-07,
+      "loss": 0.9892,
+      "step": 229
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.994489555619093e-07,
+      "loss": 1.1015,
+      "step": 230
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.994404485250024e-07,
+      "loss": 1.1257,
+      "step": 231
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.994318763977788e-07,
+      "loss": 0.9388,
+      "step": 232
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.994232391824757e-07,
+      "loss": 1.1389,
+      "step": 233
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.994145368813469e-07,
+      "loss": 0.9472,
+      "step": 234
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.994057694966631e-07,
+      "loss": 0.977,
+      "step": 235
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.993969370307121e-07,
+      "loss": 0.8947,
+      "step": 236
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.993880394857988e-07,
+      "loss": 1.2934,
+      "step": 237
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.993790768642448e-07,
+      "loss": 0.9112,
+      "step": 238
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.993700491683891e-07,
+      "loss": 1.0144,
+      "step": 239
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.993609564005872e-07,
+      "loss": 0.8699,
+      "step": 240
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.99351798563212e-07,
+      "loss": 0.8624,
+      "step": 241
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.99342575658653e-07,
+      "loss": 1.06,
+      "step": 242
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.993332876893172e-07,
+      "loss": 0.9683,
+      "step": 243
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.993239346576278e-07,
+      "loss": 1.2182,
+      "step": 244
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.993145165660258e-07,
+      "loss": 1.0217,
+      "step": 245
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.993050334169686e-07,
+      "loss": 1.0232,
+      "step": 246
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.99295485212931e-07,
+      "loss": 0.988,
+      "step": 247
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.992858719564043e-07,
+      "loss": 1.1706,
+      "step": 248
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.992761936498971e-07,
+      "loss": 0.98,
+      "step": 249
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.992664502959351e-07,
+      "loss": 0.9186,
+      "step": 250
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.992566418970606e-07,
+      "loss": 1.0298,
+      "step": 251
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.992467684558331e-07,
+      "loss": 1.0313,
+      "step": 252
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.992368299748289e-07,
+      "loss": 0.808,
+      "step": 253
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.992268264566417e-07,
+      "loss": 0.9911,
+      "step": 254
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.992167579038816e-07,
+      "loss": 0.9627,
+      "step": 255
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.992066243191761e-07,
+      "loss": 0.9853,
+      "step": 256
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.991964257051694e-07,
+      "loss": 0.9144,
+      "step": 257
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.99186162064523e-07,
+      "loss": 0.8683,
+      "step": 258
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.991758333999148e-07,
+      "loss": 1.0895,
+      "step": 259
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.991654397140402e-07,
+      "loss": 1.1043,
+      "step": 260
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.991549810096114e-07,
+      "loss": 1.0593,
+      "step": 261
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.991444572893575e-07,
+      "loss": 0.9056,
+      "step": 262
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.991338685560248e-07,
+      "loss": 1.0233,
+      "step": 263
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.991232148123761e-07,
+      "loss": 1.0125,
+      "step": 264
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.991124960611916e-07,
+      "loss": 1.1255,
+      "step": 265
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.991017123052682e-07,
+      "loss": 1.0842,
+      "step": 266
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.9909086354742e-07,
+      "loss": 1.1586,
+      "step": 267
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.99079949790478e-07,
+      "loss": 1.0952,
+      "step": 268
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.990689710372899e-07,
+      "loss": 1.0778,
+      "step": 269
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.990579272907206e-07,
+      "loss": 0.9299,
+      "step": 270
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.99046818553652e-07,
+      "loss": 0.9646,
+      "step": 271
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.990356448289827e-07,
+      "loss": 1.0505,
+      "step": 272
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.990244061196286e-07,
+      "loss": 1.1298,
+      "step": 273
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.990131024285224e-07,
+      "loss": 1.1034,
+      "step": 274
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.990017337586137e-07,
+      "loss": 0.9777,
+      "step": 275
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.98990300112869e-07,
+      "loss": 0.9534,
+      "step": 276
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.98978801494272e-07,
+      "loss": 0.947,
+      "step": 277
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.98967237905823e-07,
+      "loss": 1.0191,
+      "step": 278
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.989556093505398e-07,
+      "loss": 1.0003,
+      "step": 279
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.989439158314566e-07,
+      "loss": 0.9869,
+      "step": 280
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 4.989321573516247e-07,
+      "loss": 1.0413,
+      "step": 281
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.989203339141127e-07,
+      "loss": 1.0557,
+      "step": 282
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.989084455220056e-07,
+      "loss": 1.0703,
+      "step": 283
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.988964921784057e-07,
+      "loss": 1.0759,
+      "step": 284
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.988844738864322e-07,
+      "loss": 0.9923,
+      "step": 285
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.988723906492211e-07,
+      "loss": 1.0508,
+      "step": 286
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.988602424699256e-07,
+      "loss": 1.2,
+      "step": 287
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.988480293517158e-07,
+      "loss": 1.1024,
+      "step": 288
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.988357512977784e-07,
+      "loss": 1.0653,
+      "step": 289
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.988234083113174e-07,
+      "loss": 0.9167,
+      "step": 290
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.988110003955538e-07,
+      "loss": 1.0799,
+      "step": 291
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.987985275537252e-07,
+      "loss": 1.1001,
+      "step": 292
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.987859897890864e-07,
+      "loss": 0.9928,
+      "step": 293
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.98773387104909e-07,
+      "loss": 0.926,
+      "step": 294
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.987607195044818e-07,
+      "loss": 0.9956,
+      "step": 295
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.987479869911102e-07,
+      "loss": 0.9637,
+      "step": 296
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.987351895681166e-07,
+      "loss": 1.0385,
+      "step": 297
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.987223272388406e-07,
+      "loss": 1.0181,
+      "step": 298
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.987094000066385e-07,
+      "loss": 0.9543,
+      "step": 299
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.986964078748837e-07,
+      "loss": 0.9897,
+      "step": 300
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.986833508469662e-07,
+      "loss": 0.9254,
+      "step": 301
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.986702289262934e-07,
+      "loss": 1.0356,
+      "step": 302
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.986570421162893e-07,
+      "loss": 1.0327,
+      "step": 303
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.98643790420395e-07,
+      "loss": 1.2576,
+      "step": 304
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.986304738420683e-07,
+      "loss": 1.069,
+      "step": 305
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.986170923847843e-07,
+      "loss": 0.7588,
+      "step": 306
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.986036460520348e-07,
+      "loss": 1.102,
+      "step": 307
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.985901348473283e-07,
+      "loss": 1.0011,
+      "step": 308
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.985765587741909e-07,
+      "loss": 1.0041,
+      "step": 309
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.985629178361649e-07,
+      "loss": 1.034,
+      "step": 310
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.9854921203681e-07,
+      "loss": 1.0158,
+      "step": 311
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.985354413797026e-07,
+      "loss": 1.1892,
+      "step": 312
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.985216058684361e-07,
+      "loss": 0.9833,
+      "step": 313
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.985077055066208e-07,
+      "loss": 0.9527,
+      "step": 314
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.984937402978839e-07,
+      "loss": 1.0235,
+      "step": 315
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.984797102458697e-07,
+      "loss": 1.1287,
+      "step": 316
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.98465615354239e-07,
+      "loss": 0.9738,
+      "step": 317
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.984514556266701e-07,
+      "loss": 1.0872,
+      "step": 318
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.984372310668578e-07,
+      "loss": 0.933,
+      "step": 319
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.984229416785138e-07,
+      "loss": 0.9213,
+      "step": 320
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.984085874653671e-07,
+      "loss": 1.1749,
+      "step": 321
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.983941684311633e-07,
+      "loss": 0.987,
+      "step": 322
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.983796845796646e-07,
+      "loss": 0.9715,
+      "step": 323
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.98365135914651e-07,
+      "loss": 0.8994,
+      "step": 324
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.983505224399187e-07,
+      "loss": 0.922,
+      "step": 325
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.983358441592811e-07,
+      "loss": 1.111,
+      "step": 326
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.983211010765682e-07,
+      "loss": 1.0257,
+      "step": 327
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.983062931956275e-07,
+      "loss": 0.9399,
+      "step": 328
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.982914205203226e-07,
+      "loss": 1.1079,
+      "step": 329
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.982764830545348e-07,
+      "loss": 1.1544,
+      "step": 330
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.982614808021619e-07,
+      "loss": 0.974,
+      "step": 331
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.982464137671185e-07,
+      "loss": 1.2419,
+      "step": 332
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.982312819533366e-07,
+      "loss": 0.9224,
+      "step": 333
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.982160853647644e-07,
+      "loss": 1.0398,
+      "step": 334
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.982008240053676e-07,
+      "loss": 0.9245,
+      "step": 335
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.981854978791286e-07,
+      "loss": 0.9132,
+      "step": 336
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.981701069900464e-07,
+      "loss": 0.9198,
+      "step": 337
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.981546513421377e-07,
+      "loss": 1.0996,
+      "step": 338
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.98139130939435e-07,
+      "loss": 1.1049,
+      "step": 339
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.981235457859887e-07,
+      "loss": 1.0019,
+      "step": 340
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.981078958858656e-07,
+      "loss": 0.977,
+      "step": 341
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.980921812431492e-07,
+      "loss": 1.2227,
+      "step": 342
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.980764018619404e-07,
+      "loss": 0.9953,
+      "step": 343
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.980605577463568e-07,
+      "loss": 0.9869,
+      "step": 344
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.980446489005326e-07,
+      "loss": 1.1174,
+      "step": 345
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.980286753286194e-07,
+      "loss": 1.0051,
+      "step": 346
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.980126370347853e-07,
+      "loss": 1.0339,
+      "step": 347
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.979965340232155e-07,
+      "loss": 0.9604,
+      "step": 348
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.979803662981119e-07,
+      "loss": 0.9215,
+      "step": 349
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.979641338636934e-07,
+      "loss": 1.1324,
+      "step": 350
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.979478367241958e-07,
+      "loss": 1.0466,
+      "step": 351
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.979314748838719e-07,
+      "loss": 1.0447,
+      "step": 352
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.979150483469911e-07,
+      "loss": 0.8388,
+      "step": 353
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.978985571178397e-07,
+      "loss": 0.9384,
+      "step": 354
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.978820012007213e-07,
+      "loss": 0.9549,
+      "step": 355
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.978653805999559e-07,
+      "loss": 1.0223,
+      "step": 356
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.978486953198806e-07,
+      "loss": 1.1354,
+      "step": 357
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.978319453648495e-07,
+      "loss": 1.0064,
+      "step": 358
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.978151307392331e-07,
+      "loss": 0.8928,
+      "step": 359
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.977982514474194e-07,
+      "loss": 1.0118,
+      "step": 360
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.977813074938128e-07,
+      "loss": 1.0305,
+      "step": 361
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.977642988828348e-07,
+      "loss": 1.0872,
+      "step": 362
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.977472256189237e-07,
+      "loss": 1.089,
+      "step": 363
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.977300877065346e-07,
+      "loss": 0.9659,
+      "step": 364
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.977128851501398e-07,
+      "loss": 1.0821,
+      "step": 365
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.976956179542282e-07,
+      "loss": 1.024,
+      "step": 366
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.976782861233054e-07,
+      "loss": 0.8932,
+      "step": 367
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.976608896618941e-07,
+      "loss": 1.0115,
+      "step": 368
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.976434285745339e-07,
+      "loss": 1.0936,
+      "step": 369
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.976259028657811e-07,
+      "loss": 1.0289,
+      "step": 370
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.976083125402091e-07,
+      "loss": 0.8974,
+      "step": 371
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.975906576024078e-07,
+      "loss": 0.9093,
+      "step": 372
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.975729380569845e-07,
+      "loss": 1.0713,
+      "step": 373
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.975551539085626e-07,
+      "loss": 1.0109,
+      "step": 374
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.975373051617831e-07,
+      "loss": 1.1374,
+      "step": 375
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.975193918213035e-07,
+      "loss": 0.9201,
+      "step": 376
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.975014138917982e-07,
+      "loss": 0.9781,
+      "step": 377
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.974833713779583e-07,
+      "loss": 1.0259,
+      "step": 378
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.974652642844921e-07,
+      "loss": 1.0376,
+      "step": 379
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.974470926161245e-07,
+      "loss": 1.0221,
+      "step": 380
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.974288563775974e-07,
+      "loss": 0.8828,
+      "step": 381
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.974105555736692e-07,
+      "loss": 1.0244,
+      "step": 382
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.973921902091157e-07,
+      "loss": 0.9756,
+      "step": 383
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.973737602887291e-07,
+      "loss": 0.9133,
+      "step": 384
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.973552658173185e-07,
+      "loss": 0.9819,
+      "step": 385
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.973367067997103e-07,
+      "loss": 1.1084,
+      "step": 386
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.973180832407471e-07,
+      "loss": 1.0914,
+      "step": 387
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.972993951452886e-07,
+      "loss": 1.0514,
+      "step": 388
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.972806425182115e-07,
+      "loss": 0.9057,
+      "step": 389
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.972618253644093e-07,
+      "loss": 1.0775,
+      "step": 390
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.972429436887921e-07,
+      "loss": 1.0777,
+      "step": 391
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.972239974962869e-07,
+      "loss": 0.9982,
+      "step": 392
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.972049867918378e-07,
+      "loss": 1.0403,
+      "step": 393
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.971859115804054e-07,
+      "loss": 0.8975,
+      "step": 394
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.971667718669676e-07,
+      "loss": 0.9883,
+      "step": 395
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.971475676565183e-07,
+      "loss": 0.9535,
+      "step": 396
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.971282989540692e-07,
+      "loss": 0.9227,
+      "step": 397
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.971089657646483e-07,
+      "loss": 0.9728,
+      "step": 398
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.970895680933003e-07,
+      "loss": 1.2107,
+      "step": 399
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.970701059450871e-07,
+      "loss": 0.8977,
+      "step": 400
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.970505793250872e-07,
+      "loss": 0.9873,
+      "step": 401
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.97030988238396e-07,
+      "loss": 0.8567,
+      "step": 402
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.970113326901257e-07,
+      "loss": 1.0891,
+      "step": 403
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.969916126854053e-07,
+      "loss": 1.0034,
+      "step": 404
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.969718282293807e-07,
+      "loss": 1.0105,
+      "step": 405
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.969519793272145e-07,
+      "loss": 1.0328,
+      "step": 406
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.969320659840862e-07,
+      "loss": 1.0527,
+      "step": 407
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.96912088205192e-07,
+      "loss": 1.1223,
+      "step": 408
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.968920459957453e-07,
+      "loss": 1.1309,
+      "step": 409
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.968719393609756e-07,
+      "loss": 1.0786,
+      "step": 410
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.9685176830613e-07,
+      "loss": 0.9664,
+      "step": 411
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.968315328364718e-07,
+      "loss": 1.0952,
+      "step": 412
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.968112329572815e-07,
+      "loss": 0.9693,
+      "step": 413
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.967908686738563e-07,
+      "loss": 0.9172,
+      "step": 414
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.9677043999151e-07,
+      "loss": 1.1841,
+      "step": 415
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.967499469155733e-07,
+      "loss": 0.981,
+      "step": 416
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.967293894513942e-07,
+      "loss": 0.9827,
+      "step": 417
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.967087676043366e-07,
+      "loss": 1.0054,
+      "step": 418
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.966880813797818e-07,
+      "loss": 0.9715,
+      "step": 419
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.966673307831279e-07,
+      "loss": 0.9207,
+      "step": 420
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.966465158197896e-07,
+      "loss": 0.9453,
+      "step": 421
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.966256364951986e-07,
+      "loss": 1.0919,
+      "step": 422
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.96604692814803e-07,
+      "loss": 0.948,
+      "step": 423
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.96583684784068e-07,
+      "loss": 1.0717,
+      "step": 424
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.965626124084759e-07,
+      "loss": 1.0105,
+      "step": 425
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.96541475693525e-07,
+      "loss": 0.9656,
+      "step": 426
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.96520274644731e-07,
+      "loss": 0.8958,
+      "step": 427
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.964990092676262e-07,
+      "loss": 1.0344,
+      "step": 428
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.964776795677598e-07,
+      "loss": 1.0944,
+      "step": 429
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.964562855506975e-07,
+      "loss": 1.0129,
+      "step": 430
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.964348272220221e-07,
+      "loss": 1.0259,
+      "step": 431
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.964133045873331e-07,
+      "loss": 0.8333,
+      "step": 432
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.963917176522465e-07,
+      "loss": 1.1487,
+      "step": 433
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.963700664223956e-07,
+      "loss": 0.9257,
+      "step": 434
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.9634835090343e-07,
+      "loss": 0.9626,
+      "step": 435
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.963265711010162e-07,
+      "loss": 1.0218,
+      "step": 436
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.963047270208378e-07,
+      "loss": 0.9979,
+      "step": 437
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.962828186685948e-07,
+      "loss": 1.0014,
+      "step": 438
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.962608460500039e-07,
+      "loss": 1.0499,
+      "step": 439
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.96238809170799e-07,
+      "loss": 1.0435,
+      "step": 440
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.962167080367303e-07,
+      "loss": 0.9849,
+      "step": 441
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.961945426535651e-07,
+      "loss": 0.9549,
+      "step": 442
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.961723130270874e-07,
+      "loss": 1.007,
+      "step": 443
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.961500191630979e-07,
+      "loss": 1.0137,
+      "step": 444
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.961276610674139e-07,
+      "loss": 1.0327,
+      "step": 445
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.9610523874587e-07,
+      "loss": 1.106,
+      "step": 446
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.96082752204317e-07,
+      "loss": 0.9061,
+      "step": 447
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.960602014486225e-07,
+      "loss": 0.8749,
+      "step": 448
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.960375864846712e-07,
+      "loss": 1.0795,
+      "step": 449
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.960149073183643e-07,
+      "loss": 0.9718,
+      "step": 450
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.959921639556199e-07,
+      "loss": 1.0112,
+      "step": 451
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.959693564023726e-07,
+      "loss": 1.138,
+      "step": 452
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.959464846645743e-07,
+      "loss": 0.9604,
+      "step": 453
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.959235487481927e-07,
+      "loss": 1.0363,
+      "step": 454
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.959005486592134e-07,
+      "loss": 0.9976,
+      "step": 455
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.958774844036378e-07,
+      "loss": 1.0656,
+      "step": 456
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.958543559874846e-07,
+      "loss": 1.0112,
+      "step": 457
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.958311634167889e-07,
+      "loss": 0.9421,
+      "step": 458
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.958079066976027e-07,
+      "loss": 0.9048,
+      "step": 459
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.957845858359949e-07,
+      "loss": 1.0176,
+      "step": 460
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.957612008380507e-07,
+      "loss": 0.9495,
+      "step": 461
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.957377517098727e-07,
+      "loss": 1.0253,
+      "step": 462
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.957142384575794e-07,
+      "loss": 1.234,
+      "step": 463
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.956906610873068e-07,
+      "loss": 1.0177,
+      "step": 464
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.956670196052072e-07,
+      "loss": 0.9613,
+      "step": 465
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.956433140174498e-07,
+      "loss": 0.9004,
+      "step": 466
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.956195443302202e-07,
+      "loss": 0.8422,
+      "step": 467
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.955957105497212e-07,
+      "loss": 1.0525,
+      "step": 468
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.955718126821722e-07,
+      "loss": 1.0549,
+      "step": 469
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.95547850733809e-07,
+      "loss": 1.0847,
+      "step": 470
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.955238247108845e-07,
+      "loss": 1.1729,
+      "step": 471
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.954997346196682e-07,
+      "loss": 1.1092,
+      "step": 472
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.954755804664462e-07,
+      "loss": 1.0468,
+      "step": 473
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.954513622575214e-07,
+      "loss": 0.9308,
+      "step": 474
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.954270799992138e-07,
+      "loss": 1.0839,
+      "step": 475
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.954027336978591e-07,
+      "loss": 0.838,
+      "step": 476
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.953783233598109e-07,
+      "loss": 0.8153,
+      "step": 477
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.953538489914387e-07,
+      "loss": 1.032,
+      "step": 478
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.953293105991289e-07,
+      "loss": 1.0384,
+      "step": 479
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.953047081892849e-07,
+      "loss": 0.9561,
+      "step": 480
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.952800417683265e-07,
+      "loss": 1.1624,
+      "step": 481
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.952553113426903e-07,
+      "loss": 1.0896,
+      "step": 482
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.952305169188295e-07,
+      "loss": 0.8951,
+      "step": 483
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.952056585032141e-07,
+      "loss": 1.0462,
+      "step": 484
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.951807361023309e-07,
+      "loss": 1.0284,
+      "step": 485
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.951557497226832e-07,
+      "loss": 0.9198,
+      "step": 486
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.951306993707912e-07,
+      "loss": 0.9937,
+      "step": 487
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.951055850531915e-07,
+      "loss": 1.009,
+      "step": 488
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.950804067764377e-07,
+      "loss": 0.8874,
+      "step": 489
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.950551645470998e-07,
+      "loss": 1.0382,
+      "step": 490
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.950298583717648e-07,
+      "loss": 0.9465,
+      "step": 491
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.950044882570361e-07,
+      "loss": 0.9974,
+      "step": 492
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.94979054209534e-07,
+      "loss": 1.2416,
+      "step": 493
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.949535562358953e-07,
+      "loss": 0.943,
+      "step": 494
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.949279943427737e-07,
+      "loss": 0.9963,
+      "step": 495
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.949023685368395e-07,
+      "loss": 0.9793,
+      "step": 496
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.948766788247794e-07,
+      "loss": 1.0342,
+      "step": 497
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.948509252132972e-07,
+      "loss": 1.0017,
+      "step": 498
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.94825107709113e-07,
+      "loss": 0.9664,
+      "step": 499
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.94799226318964e-07,
+      "loss": 1.0643,
+      "step": 500
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.947732810496037e-07,
+      "loss": 0.9888,
+      "step": 501
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.947472719078024e-07,
+      "loss": 1.1006,
+      "step": 502
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.947211989003471e-07,
+      "loss": 1.0456,
+      "step": 503
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.946950620340414e-07,
+      "loss": 0.9978,
+      "step": 504
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.946688613157056e-07,
+      "loss": 0.8233,
+      "step": 505
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.946425967521766e-07,
+      "loss": 0.8843,
+      "step": 506
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.946162683503081e-07,
+      "loss": 1.0397,
+      "step": 507
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.945898761169704e-07,
+      "loss": 1.0764,
+      "step": 508
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.945634200590502e-07,
+      "loss": 1.2706,
+      "step": 509
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.945369001834514e-07,
+      "loss": 1.0872,
+      "step": 510
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.94510316497094e-07,
+      "loss": 0.968,
+      "step": 511
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.944836690069151e-07,
+      "loss": 1.0411,
+      "step": 512
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.944569577198679e-07,
+      "loss": 0.9547,
+      "step": 513
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.94430182642923e-07,
+      "loss": 1.1596,
+      "step": 514
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.94403343783067e-07,
+      "loss": 1.0512,
+      "step": 515
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.943764411473033e-07,
+      "loss": 1.0141,
+      "step": 516
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.943494747426522e-07,
+      "loss": 0.9752,
+      "step": 517
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.943224445761504e-07,
+      "loss": 0.9632,
+      "step": 518
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.942953506548513e-07,
+      "loss": 1.0944,
+      "step": 519
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.942681929858248e-07,
+      "loss": 0.9134,
+      "step": 520
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.942409715761578e-07,
+      "loss": 1.0387,
+      "step": 521
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.942136864329533e-07,
+      "loss": 0.8817,
+      "step": 522
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.941863375633315e-07,
+      "loss": 1.0728,
+      "step": 523
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.941589249744289e-07,
+      "loss": 0.9469,
+      "step": 524
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.941314486733986e-07,
+      "loss": 0.9242,
+      "step": 525
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.941039086674105e-07,
+      "loss": 1.0802,
+      "step": 526
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.940763049636509e-07,
+      "loss": 1.0335,
+      "step": 527
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.94048637569323e-07,
+      "loss": 1.0692,
+      "step": 528
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.940209064916465e-07,
+      "loss": 1.0677,
+      "step": 529
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.939931117378575e-07,
+      "loss": 0.9398,
+      "step": 530
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 4.939652533152091e-07,
+      "loss": 1.0232,
+      "step": 531
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.939373312309708e-07,
+      "loss": 0.9853,
+      "step": 532
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.939093454924287e-07,
+      "loss": 0.976,
+      "step": 533
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.938812961068855e-07,
+      "loss": 1.0847,
+      "step": 534
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.938531830816606e-07,
+      "loss": 1.0347,
+      "step": 535
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.938250064240899e-07,
+      "loss": 1.0324,
+      "step": 536
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.937967661415262e-07,
+      "loss": 1.1935,
+      "step": 537
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.937684622413384e-07,
+      "loss": 1.129,
+      "step": 538
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.937400947309123e-07,
+      "loss": 0.9783,
+      "step": 539
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.937116636176505e-07,
+      "loss": 1.0641,
+      "step": 540
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.936831689089718e-07,
+      "loss": 1.0223,
+      "step": 541
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.936546106123116e-07,
+      "loss": 0.9886,
+      "step": 542
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.936259887351224e-07,
+      "loss": 0.9818,
+      "step": 543
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.935973032848726e-07,
+      "loss": 1.0409,
+      "step": 544
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.935685542690477e-07,
+      "loss": 1.0106,
+      "step": 545
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.935397416951497e-07,
+      "loss": 1.0249,
+      "step": 546
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.935108655706971e-07,
+      "loss": 1.0428,
+      "step": 547
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.934819259032248e-07,
+      "loss": 0.9928,
+      "step": 548
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.934529227002846e-07,
+      "loss": 0.9511,
+      "step": 549
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.934238559694447e-07,
+      "loss": 1.1731,
+      "step": 550
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.9339472571829e-07,
+      "loss": 1.1887,
+      "step": 551
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.933655319544219e-07,
+      "loss": 0.9896,
+      "step": 552
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.933362746854583e-07,
+      "loss": 0.9692,
+      "step": 553
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.933069539190338e-07,
+      "loss": 1.0899,
+      "step": 554
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.932775696627995e-07,
+      "loss": 1.1258,
+      "step": 555
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.932481219244231e-07,
+      "loss": 1.0369,
+      "step": 556
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.932186107115888e-07,
+      "loss": 1.0133,
+      "step": 557
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.931890360319976e-07,
+      "loss": 1.0479,
+      "step": 558
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.931593978933665e-07,
+      "loss": 0.9027,
+      "step": 559
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.931296963034298e-07,
+      "loss": 1.0451,
+      "step": 560
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.930999312699378e-07,
+      "loss": 1.1568,
+      "step": 561
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.930701028006576e-07,
+      "loss": 0.9237,
+      "step": 562
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.930402109033728e-07,
+      "loss": 0.9361,
+      "step": 563
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.930102555858836e-07,
+      "loss": 1.065,
+      "step": 564
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.929802368560065e-07,
+      "loss": 1.0951,
+      "step": 565
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.92950154721575e-07,
+      "loss": 0.9515,
+      "step": 566
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.929200091904388e-07,
+      "loss": 1.0701,
+      "step": 567
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.928898002704641e-07,
+      "loss": 1.0068,
+      "step": 568
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.92859527969534e-07,
+      "loss": 1.0826,
+      "step": 569
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.928291922955478e-07,
+      "loss": 1.1043,
+      "step": 570
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.927987932564215e-07,
+      "loss": 1.0747,
+      "step": 571
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.927683308600875e-07,
+      "loss": 1.2905,
+      "step": 572
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.927378051144948e-07,
+      "loss": 1.1999,
+      "step": 573
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.927072160276091e-07,
+      "loss": 0.9205,
+      "step": 574
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.926765636074124e-07,
+      "loss": 1.0877,
+      "step": 575
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.926458478619033e-07,
+      "loss": 0.8531,
+      "step": 576
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.926150687990969e-07,
+      "loss": 1.0126,
+      "step": 577
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.925842264270249e-07,
+      "loss": 1.2443,
+      "step": 578
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.925533207537355e-07,
+      "loss": 0.978,
+      "step": 579
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.925223517872933e-07,
+      "loss": 0.952,
+      "step": 580
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.924913195357795e-07,
+      "loss": 1.0756,
+      "step": 581
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.92460224007292e-07,
+      "loss": 0.9924,
+      "step": 582
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.924290652099448e-07,
+      "loss": 0.9456,
+      "step": 583
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.923978431518688e-07,
+      "loss": 1.0065,
+      "step": 584
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.923665578412111e-07,
+      "loss": 0.9037,
+      "step": 585
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.923352092861357e-07,
+      "loss": 1.0408,
+      "step": 586
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.923037974948227e-07,
+      "loss": 1.0946,
+      "step": 587
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.922723224754688e-07,
+      "loss": 1.019,
+      "step": 588
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.922407842362874e-07,
+      "loss": 0.8788,
+      "step": 589
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.922091827855082e-07,
+      "loss": 1.0193,
+      "step": 590
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.921775181313774e-07,
+      "loss": 1.0904,
+      "step": 591
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.921457902821578e-07,
+      "loss": 0.9689,
+      "step": 592
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.921139992461286e-07,
+      "loss": 1.0158,
+      "step": 593
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.920821450315855e-07,
+      "loss": 0.9623,
+      "step": 594
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.920502276468408e-07,
+      "loss": 0.9591,
+      "step": 595
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.920182471002231e-07,
+      "loss": 1.0045,
+      "step": 596
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.919862034000776e-07,
+      "loss": 0.9524,
+      "step": 597
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.91954096554766e-07,
+      "loss": 1.1391,
+      "step": 598
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.919219265726664e-07,
+      "loss": 0.8323,
+      "step": 599
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.918896934621734e-07,
+      "loss": 1.1219,
+      "step": 600
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.918573972316981e-07,
+      "loss": 0.754,
+      "step": 601
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.91825037889668e-07,
+      "loss": 0.9894,
+      "step": 602
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.917926154445271e-07,
+      "loss": 0.9451,
+      "step": 603
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.91760129904736e-07,
+      "loss": 1.0046,
+      "step": 604
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.917275812787715e-07,
+      "loss": 1.049,
+      "step": 605
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.916949695751273e-07,
+      "loss": 1.0342,
+      "step": 606
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.916622948023129e-07,
+      "loss": 1.0607,
+      "step": 607
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.916295569688548e-07,
+      "loss": 0.9334,
+      "step": 608
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.915967560832958e-07,
+      "loss": 0.9867,
+      "step": 609
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.915638921541951e-07,
+      "loss": 0.9532,
+      "step": 610
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.915309651901285e-07,
+      "loss": 1.0336,
+      "step": 611
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.91497975199688e-07,
+      "loss": 1.0291,
+      "step": 612
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.914649221914822e-07,
+      "loss": 0.8819,
+      "step": 613
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.914318061741361e-07,
+      "loss": 0.8821,
+      "step": 614
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.913986271562913e-07,
+      "loss": 1.0248,
+      "step": 615
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.913653851466056e-07,
+      "loss": 1.1008,
+      "step": 616
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.913320801537534e-07,
+      "loss": 1.2488,
+      "step": 617
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.912987121864254e-07,
+      "loss": 0.9689,
+      "step": 618
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.91265281253329e-07,
+      "loss": 0.918,
+      "step": 619
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.912317873631877e-07,
+      "loss": 1.1027,
+      "step": 620
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.911982305247416e-07,
+      "loss": 1.0963,
+      "step": 621
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.911646107467471e-07,
+      "loss": 0.9667,
+      "step": 622
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.911309280379773e-07,
+      "loss": 0.9905,
+      "step": 623
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.910971824072214e-07,
+      "loss": 0.9575,
+      "step": 624
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.910633738632852e-07,
+      "loss": 1.13,
+      "step": 625
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.91029502414991e-07,
+      "loss": 0.9378,
+      "step": 626
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.909955680711773e-07,
+      "loss": 0.9591,
+      "step": 627
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.90961570840699e-07,
+      "loss": 1.0168,
+      "step": 628
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.909275107324278e-07,
+      "loss": 1.0702,
+      "step": 629
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.908933877552513e-07,
+      "loss": 0.9767,
+      "step": 630
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.908592019180738e-07,
+      "loss": 1.0498,
+      "step": 631
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.908249532298157e-07,
+      "loss": 1.1876,
+      "step": 632
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.907906416994145e-07,
+      "loss": 1.0107,
+      "step": 633
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.907562673358234e-07,
+      "loss": 0.9215,
+      "step": 634
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.907218301480121e-07,
+      "loss": 1.0323,
+      "step": 635
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.90687330144967e-07,
+      "loss": 1.0486,
+      "step": 636
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.906527673356906e-07,
+      "loss": 0.963,
+      "step": 637
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.90618141729202e-07,
+      "loss": 0.9798,
+      "step": 638
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.905834533345366e-07,
+      "loss": 1.0481,
+      "step": 639
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.905487021607461e-07,
+      "loss": 1.1382,
+      "step": 640
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.905138882168987e-07,
+      "loss": 1.0147,
+      "step": 641
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.904790115120788e-07,
+      "loss": 1.1058,
+      "step": 642
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.904440720553875e-07,
+      "loss": 0.981,
+      "step": 643
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.90409069855942e-07,
+      "loss": 1.0657,
+      "step": 644
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.90374004922876e-07,
+      "loss": 1.1014,
+      "step": 645
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.903388772653395e-07,
+      "loss": 1.0712,
+      "step": 646
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.903036868924989e-07,
+      "loss": 1.0849,
+      "step": 647
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.90268433813537e-07,
+      "loss": 1.1238,
+      "step": 648
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.902331180376528e-07,
+      "loss": 0.9155,
+      "step": 649
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.901977395740619e-07,
+      "loss": 1.059,
+      "step": 650
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.901622984319962e-07,
+      "loss": 0.8837,
+      "step": 651
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.901267946207037e-07,
+      "loss": 1.0031,
+      "step": 652
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.900912281494491e-07,
+      "loss": 0.9921,
+      "step": 653
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.900555990275134e-07,
+      "loss": 0.9409,
+      "step": 654
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.900199072641936e-07,
+      "loss": 0.9712,
+      "step": 655
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.899841528688036e-07,
+      "loss": 1.0907,
+      "step": 656
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.899483358506729e-07,
+      "loss": 0.9199,
+      "step": 657
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.899124562191483e-07,
+      "loss": 0.9157,
+      "step": 658
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.898765139835921e-07,
+      "loss": 0.9957,
+      "step": 659
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.898405091533834e-07,
+      "loss": 0.9829,
+      "step": 660
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.898044417379173e-07,
+      "loss": 0.9698,
+      "step": 661
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.897683117466056e-07,
+      "loss": 0.8728,
+      "step": 662
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.897321191888762e-07,
+      "loss": 0.9944,
+      "step": 663
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.896958640741734e-07,
+      "loss": 0.9567,
+      "step": 664
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.896595464119577e-07,
+      "loss": 1.1733,
+      "step": 665
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.896231662117062e-07,
+      "loss": 0.984,
+      "step": 666
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.895867234829121e-07,
+      "loss": 1.2404,
+      "step": 667
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.895502182350848e-07,
+      "loss": 0.9711,
+      "step": 668
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.895136504777503e-07,
+      "loss": 1.028,
+      "step": 669
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.894770202204508e-07,
+      "loss": 0.9821,
+      "step": 670
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.894403274727446e-07,
+      "loss": 1.0347,
+      "step": 671
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.894035722442069e-07,
+      "loss": 1.1268,
+      "step": 672
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.893667545444284e-07,
+      "loss": 1.0601,
+      "step": 673
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.893298743830167e-07,
+      "loss": 1.0687,
+      "step": 674
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.892929317695956e-07,
+      "loss": 1.037,
+      "step": 675
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.892559267138048e-07,
+      "loss": 0.9359,
+      "step": 676
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.892188592253009e-07,
+      "loss": 1.1047,
+      "step": 677
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.891817293137565e-07,
+      "loss": 0.9256,
+      "step": 678
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.8914453698886e-07,
+      "loss": 0.9676,
+      "step": 679
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.891072822603172e-07,
+      "loss": 0.9329,
+      "step": 680
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.890699651378491e-07,
+      "loss": 0.9597,
+      "step": 681
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.890325856311936e-07,
+      "loss": 0.9406,
+      "step": 682
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.889951437501046e-07,
+      "loss": 0.9994,
+      "step": 683
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.889576395043526e-07,
+      "loss": 0.9919,
+      "step": 684
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.88920072903724e-07,
+      "loss": 0.948,
+      "step": 685
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.888824439580216e-07,
+      "loss": 0.8772,
+      "step": 686
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.888447526770646e-07,
+      "loss": 1.0739,
+      "step": 687
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.888069990706884e-07,
+      "loss": 0.9699,
+      "step": 688
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.887691831487444e-07,
+      "loss": 1.2236,
+      "step": 689
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.887313049211006e-07,
+      "loss": 0.9359,
+      "step": 690
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.886933643976414e-07,
+      "loss": 1.2383,
+      "step": 691
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.886553615882668e-07,
+      "loss": 0.9856,
+      "step": 692
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.886172965028935e-07,
+      "loss": 1.0286,
+      "step": 693
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.885791691514548e-07,
+      "loss": 1.2042,
+      "step": 694
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.885409795438994e-07,
+      "loss": 1.0263,
+      "step": 695
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.88502727690193e-07,
+      "loss": 1.0051,
+      "step": 696
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.884644136003171e-07,
+      "loss": 0.9362,
+      "step": 697
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.884260372842696e-07,
+      "loss": 1.2036,
+      "step": 698
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.883875987520647e-07,
+      "loss": 1.0981,
+      "step": 699
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.883490980137326e-07,
+      "loss": 0.9598,
+      "step": 700
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.8831053507932e-07,
+      "loss": 0.9957,
+      "step": 701
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.882719099588898e-07,
+      "loss": 1.0824,
+      "step": 702
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.882332226625208e-07,
+      "loss": 1.1218,
+      "step": 703
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.881944732003085e-07,
+      "loss": 0.9495,
+      "step": 704
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.881556615823642e-07,
+      "loss": 0.9468,
+      "step": 705
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.881167878188158e-07,
+      "loss": 0.8537,
+      "step": 706
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.88077851919807e-07,
+      "loss": 1.0409,
+      "step": 707
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.880388538954983e-07,
+      "loss": 0.8885,
+      "step": 708
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.879997937560655e-07,
+      "loss": 0.9229,
+      "step": 709
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.879606715117018e-07,
+      "loss": 0.8834,
+      "step": 710
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.879214871726155e-07,
+      "loss": 0.9395,
+      "step": 711
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.878822407490317e-07,
+      "loss": 1.0701,
+      "step": 712
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.878429322511918e-07,
+      "loss": 1.0387,
+      "step": 713
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.878035616893528e-07,
+      "loss": 1.0974,
+      "step": 714
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.877641290737883e-07,
+      "loss": 0.8945,
+      "step": 715
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.877246344147884e-07,
+      "loss": 1.0405,
+      "step": 716
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.876850777226587e-07,
+      "loss": 1.2557,
+      "step": 717
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.876454590077215e-07,
+      "loss": 1.0942,
+      "step": 718
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.876057782803152e-07,
+      "loss": 0.9102,
+      "step": 719
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.87566035550794e-07,
+      "loss": 0.9809,
+      "step": 720
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.875262308295289e-07,
+      "loss": 0.8665,
+      "step": 721
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.874863641269064e-07,
+      "loss": 0.9468,
+      "step": 722
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.874464354533299e-07,
+      "loss": 1.0073,
+      "step": 723
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.874064448192185e-07,
+      "loss": 0.9523,
+      "step": 724
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.873663922350073e-07,
+      "loss": 0.984,
+      "step": 725
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.873262777111482e-07,
+      "loss": 1.0379,
+      "step": 726
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.872861012581088e-07,
+      "loss": 1.0099,
+      "step": 727
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.872458628863728e-07,
+      "loss": 0.9317,
+      "step": 728
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.872055626064403e-07,
+      "loss": 1.23,
+      "step": 729
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.871652004288275e-07,
+      "loss": 1.0409,
+      "step": 730
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.871247763640667e-07,
+      "loss": 0.9909,
+      "step": 731
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.870842904227063e-07,
+      "loss": 0.8738,
+      "step": 732
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.870437426153112e-07,
+      "loss": 1.0283,
+      "step": 733
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.870031329524619e-07,
+      "loss": 0.9638,
+      "step": 734
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.869624614447552e-07,
+      "loss": 1.0567,
+      "step": 735
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.869217281028044e-07,
+      "loss": 0.9265,
+      "step": 736
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.868809329372385e-07,
+      "loss": 1.0355,
+      "step": 737
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.868400759587031e-07,
+      "loss": 1.0424,
+      "step": 738
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.867991571778591e-07,
+      "loss": 1.1963,
+      "step": 739
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.867581766053846e-07,
+      "loss": 0.9655,
+      "step": 740
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.86717134251973e-07,
+      "loss": 1.0624,
+      "step": 741
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.866760301283342e-07,
+      "loss": 1.1112,
+      "step": 742
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.86634864245194e-07,
+      "loss": 1.1432,
+      "step": 743
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.865936366132945e-07,
+      "loss": 1.1147,
+      "step": 744
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.865523472433941e-07,
+      "loss": 1.0794,
+      "step": 745
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.865109961462668e-07,
+      "loss": 0.9718,
+      "step": 746
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.86469583332703e-07,
+      "loss": 0.9163,
+      "step": 747
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.864281088135093e-07,
+      "loss": 1.0615,
+      "step": 748
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.863865725995081e-07,
+      "loss": 1.0498,
+      "step": 749
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.863449747015383e-07,
+      "loss": 0.9849,
+      "step": 750
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.863033151304546e-07,
+      "loss": 1.0116,
+      "step": 751
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.862615938971278e-07,
+      "loss": 1.0412,
+      "step": 752
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.862198110124448e-07,
+      "loss": 1.0478,
+      "step": 753
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.861779664873088e-07,
+      "loss": 1.1512,
+      "step": 754
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.861360603326389e-07,
+      "loss": 0.9955,
+      "step": 755
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.860940925593702e-07,
+      "loss": 1.0198,
+      "step": 756
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.860520631784542e-07,
+      "loss": 1.0174,
+      "step": 757
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.860099722008581e-07,
+      "loss": 1.0141,
+      "step": 758
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.859678196375654e-07,
+      "loss": 1.0321,
+      "step": 759
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.859256054995757e-07,
+      "loss": 0.9098,
+      "step": 760
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.858833297979044e-07,
+      "loss": 1.1386,
+      "step": 761
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.858409925435832e-07,
+      "loss": 1.0999,
+      "step": 762
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.857985937476599e-07,
+      "loss": 0.9935,
+      "step": 763
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.857561334211982e-07,
+      "loss": 0.9682,
+      "step": 764
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.857136115752781e-07,
+      "loss": 0.9657,
+      "step": 765
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.856710282209951e-07,
+      "loss": 0.8528,
+      "step": 766
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.856283833694615e-07,
+      "loss": 1.1425,
+      "step": 767
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.85585677031805e-07,
+      "loss": 0.8479,
+      "step": 768
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.855429092191698e-07,
+      "loss": 1.0024,
+      "step": 769
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.855000799427159e-07,
+      "loss": 1.0482,
+      "step": 770
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.854571892136194e-07,
+      "loss": 1.0434,
+      "step": 771
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.854142370430724e-07,
+      "loss": 1.1352,
+      "step": 772
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.853712234422831e-07,
+      "loss": 0.9685,
+      "step": 773
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.853281484224757e-07,
+      "loss": 0.9947,
+      "step": 774
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.852850119948904e-07,
+      "loss": 0.9527,
+      "step": 775
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.852418141707834e-07,
+      "loss": 1.079,
+      "step": 776
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.851985549614272e-07,
+      "loss": 1.1571,
+      "step": 777
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.851552343781099e-07,
+      "loss": 1.0321,
+      "step": 778
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.851118524321357e-07,
+      "loss": 0.9385,
+      "step": 779
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.850684091348252e-07,
+      "loss": 1.2134,
+      "step": 780
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.850249044975145e-07,
+      "loss": 0.8825,
+      "step": 781
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.849813385315559e-07,
+      "loss": 1.067,
+      "step": 782
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.84937711248318e-07,
+      "loss": 0.9777,
+      "step": 783
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.848940226591849e-07,
+      "loss": 1.1401,
+      "step": 784
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.848502727755571e-07,
+      "loss": 1.2363,
+      "step": 785
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.848064616088506e-07,
+      "loss": 1.1253,
+      "step": 786
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.847625891704981e-07,
+      "loss": 0.9506,
+      "step": 787
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.847186554719478e-07,
+      "loss": 0.997,
+      "step": 788
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.846746605246638e-07,
+      "loss": 1.2174,
+      "step": 789
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.846306043401268e-07,
+      "loss": 1.1032,
+      "step": 790
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.845864869298325e-07,
+      "loss": 0.968,
+      "step": 791
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.845423083052936e-07,
+      "loss": 1.0879,
+      "step": 792
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.844980684780379e-07,
+      "loss": 0.9136,
+      "step": 793
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.8445376745961e-07,
+      "loss": 1.0882,
+      "step": 794
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.844094052615697e-07,
+      "loss": 1.0164,
+      "step": 795
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.843649818954934e-07,
+      "loss": 1.1169,
+      "step": 796
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.843204973729728e-07,
+      "loss": 0.9683,
+      "step": 797
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.842759517056163e-07,
+      "loss": 0.9023,
+      "step": 798
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.842313449050477e-07,
+      "loss": 1.1033,
+      "step": 799
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.841866769829069e-07,
+      "loss": 1.1018,
+      "step": 800
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.841419479508499e-07,
+      "loss": 1.0958,
+      "step": 801
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.840971578205485e-07,
+      "loss": 1.0291,
+      "step": 802
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.840523066036905e-07,
+      "loss": 1.0117,
+      "step": 803
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.840073943119794e-07,
+      "loss": 0.9812,
+      "step": 804
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.839624209571352e-07,
+      "loss": 0.9407,
+      "step": 805
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.839173865508931e-07,
+      "loss": 1.0837,
+      "step": 806
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.83872291105005e-07,
+      "loss": 1.1099,
+      "step": 807
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.838271346312381e-07,
+      "loss": 0.9728,
+      "step": 808
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.837819171413757e-07,
+      "loss": 0.9626,
+      "step": 809
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.837366386472174e-07,
+      "loss": 1.0201,
+      "step": 810
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.836912991605782e-07,
+      "loss": 0.9793,
+      "step": 811
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.836458986932892e-07,
+      "loss": 0.9397,
+      "step": 812
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.836004372571976e-07,
+      "loss": 1.0691,
+      "step": 813
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.835549148641663e-07,
+      "loss": 1.2264,
+      "step": 814
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.83509331526074e-07,
+      "loss": 0.9722,
+      "step": 815
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.834636872548156e-07,
+      "loss": 1.0166,
+      "step": 816
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.834179820623017e-07,
+      "loss": 1.2145,
+      "step": 817
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.833722159604589e-07,
+      "loss": 0.9476,
+      "step": 818
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.833263889612297e-07,
+      "loss": 1.1504,
+      "step": 819
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.832805010765723e-07,
+      "loss": 0.9639,
+      "step": 820
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.83234552318461e-07,
+      "loss": 1.0794,
+      "step": 821
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.83188542698886e-07,
+      "loss": 1.0656,
+      "step": 822
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.831424722298531e-07,
+      "loss": 1.1132,
+      "step": 823
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.830963409233843e-07,
+      "loss": 0.9176,
+      "step": 824
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.830501487915174e-07,
+      "loss": 1.1011,
+      "step": 825
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.83003895846306e-07,
+      "loss": 1.085,
+      "step": 826
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.829575820998194e-07,
+      "loss": 1.1236,
+      "step": 827
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.829112075641432e-07,
+      "loss": 1.0963,
+      "step": 828
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.828647722513784e-07,
+      "loss": 1.0071,
+      "step": 829
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.828182761736423e-07,
+      "loss": 1.1139,
+      "step": 830
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.827717193430677e-07,
+      "loss": 1.0241,
+      "step": 831
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.827251017718033e-07,
+      "loss": 1.1175,
+      "step": 832
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.82678423472014e-07,
+      "loss": 1.1219,
+      "step": 833
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.826316844558801e-07,
+      "loss": 1.0481,
+      "step": 834
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.825848847355979e-07,
+      "loss": 0.9822,
+      "step": 835
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.825380243233796e-07,
+      "loss": 0.9738,
+      "step": 836
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.824911032314534e-07,
+      "loss": 1.0644,
+      "step": 837
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.824441214720628e-07,
+      "loss": 0.9967,
+      "step": 838
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.823970790574677e-07,
+      "loss": 1.0126,
+      "step": 839
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.823499759999436e-07,
+      "loss": 1.0103,
+      "step": 840
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.823028123117817e-07,
+      "loss": 0.9965,
+      "step": 841
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.822555880052892e-07,
+      "loss": 1.0667,
+      "step": 842
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.822083030927891e-07,
+      "loss": 1.0016,
+      "step": 843
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.821609575866201e-07,
+      "loss": 0.8908,
+      "step": 844
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.821135514991369e-07,
+      "loss": 1.1862,
+      "step": 845
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.820660848427098e-07,
+      "loss": 1.0201,
+      "step": 846
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.82018557629725e-07,
+      "loss": 1.031,
+      "step": 847
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.819709698725846e-07,
+      "loss": 1.0217,
+      "step": 848
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.819233215837062e-07,
+      "loss": 1.0128,
+      "step": 849
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.818756127755237e-07,
+      "loss": 1.194,
+      "step": 850
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.818278434604863e-07,
+      "loss": 1.0911,
+      "step": 851
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.817800136510591e-07,
+      "loss": 1.0431,
+      "step": 852
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.817321233597232e-07,
+      "loss": 0.8697,
+      "step": 853
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.816841725989753e-07,
+      "loss": 0.9965,
+      "step": 854
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.816361613813279e-07,
+      "loss": 0.9966,
+      "step": 855
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.815880897193094e-07,
+      "loss": 0.9753,
+      "step": 856
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.815399576254638e-07,
+      "loss": 0.9764,
+      "step": 857
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.81491765112351e-07,
+      "loss": 0.9691,
+      "step": 858
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.814435121925466e-07,
+      "loss": 1.074,
+      "step": 859
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.813951988786419e-07,
+      "loss": 0.927,
+      "step": 860
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.813468251832441e-07,
+      "loss": 1.0687,
+      "step": 861
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.81298391118976e-07,
+      "loss": 0.9282,
+      "step": 862
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.812498966984764e-07,
+      "loss": 0.987,
+      "step": 863
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.812013419343996e-07,
+      "loss": 1.0776,
+      "step": 864
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.811527268394157e-07,
+      "loss": 1.0361,
+      "step": 865
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.811040514262106e-07,
+      "loss": 1.0157,
+      "step": 866
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.810553157074859e-07,
+      "loss": 1.045,
+      "step": 867
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.810065196959591e-07,
+      "loss": 0.9267,
+      "step": 868
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.809576634043631e-07,
+      "loss": 1.0311,
+      "step": 869
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.809087468454469e-07,
+      "loss": 1.1152,
+      "step": 870
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.80859770031975e-07,
+      "loss": 0.9713,
+      "step": 871
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.808107329767274e-07,
+      "loss": 0.978,
+      "step": 872
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.807616356925005e-07,
+      "loss": 1.083,
+      "step": 873
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.807124781921058e-07,
+      "loss": 1.0212,
+      "step": 874
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.806632604883708e-07,
+      "loss": 0.9418,
+      "step": 875
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.806139825941385e-07,
+      "loss": 1.0993,
+      "step": 876
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.805646445222678e-07,
+      "loss": 1.0355,
+      "step": 877
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.805152462856332e-07,
+      "loss": 0.9544,
+      "step": 878
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.804657878971251e-07,
+      "loss": 0.9101,
+      "step": 879
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.804162693696493e-07,
+      "loss": 1.0702,
+      "step": 880
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.803666907161274e-07,
+      "loss": 1.0464,
+      "step": 881
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.803170519494968e-07,
+      "loss": 1.128,
+      "step": 882
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.802673530827105e-07,
+      "loss": 1.1888,
+      "step": 883
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.802175941287371e-07,
+      "loss": 0.9462,
+      "step": 884
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.80167775100561e-07,
+      "loss": 0.953,
+      "step": 885
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.801178960111823e-07,
+      "loss": 0.9528,
+      "step": 886
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.800679568736166e-07,
+      "loss": 1.0397,
+      "step": 887
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.800179577008955e-07,
+      "loss": 1.1189,
+      "step": 888
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.799678985060657e-07,
+      "loss": 1.1409,
+      "step": 889
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.799177793021904e-07,
+      "loss": 1.0502,
+      "step": 890
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.798676001023475e-07,
+      "loss": 1.0727,
+      "step": 891
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.798173609196313e-07,
+      "loss": 1.0217,
+      "step": 892
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.797670617671514e-07,
+      "loss": 1.0162,
+      "step": 893
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.797167026580331e-07,
+      "loss": 0.9111,
+      "step": 894
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.796662836054175e-07,
+      "loss": 0.9706,
+      "step": 895
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.79615804622461e-07,
+      "loss": 0.9667,
+      "step": 896
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.795652657223361e-07,
+      "loss": 0.9222,
+      "step": 897
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.795146669182304e-07,
+      "loss": 0.9623,
+      "step": 898
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.794640082233477e-07,
+      "loss": 1.0462,
+      "step": 899
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.79413289650907e-07,
+      "loss": 1.0011,
+      "step": 900
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.79362511214143e-07,
+      "loss": 1.0029,
+      "step": 901
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.793116729263063e-07,
+      "loss": 1.0439,
+      "step": 902
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.792607748006629e-07,
+      "loss": 1.0299,
+      "step": 903
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.792098168504943e-07,
+      "loss": 1.0739,
+      "step": 904
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.791587990890977e-07,
+      "loss": 1.1162,
+      "step": 905
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.79107721529786e-07,
+      "loss": 1.0449,
+      "step": 906
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.790565841858879e-07,
+      "loss": 0.9028,
+      "step": 907
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.79005387070747e-07,
+      "loss": 1.0231,
+      "step": 908
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.789541301977233e-07,
+      "loss": 1.1491,
+      "step": 909
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.789028135801918e-07,
+      "loss": 1.0664,
+      "step": 910
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.788514372315436e-07,
+      "loss": 0.9742,
+      "step": 911
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.788000011651848e-07,
+      "loss": 0.9542,
+      "step": 912
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.787485053945376e-07,
+      "loss": 0.9996,
+      "step": 913
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.786969499330396e-07,
+      "loss": 1.1351,
+      "step": 914
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.78645334794144e-07,
+      "loss": 1.0999,
+      "step": 915
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.785936599913192e-07,
+      "loss": 1.0168,
+      "step": 916
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.785419255380499e-07,
+      "loss": 0.9694,
+      "step": 917
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.784901314478357e-07,
+      "loss": 1.0354,
+      "step": 918
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.784382777341922e-07,
+      "loss": 1.0691,
+      "step": 919
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.783863644106502e-07,
+      "loss": 0.9054,
+      "step": 920
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.783343914907563e-07,
+      "loss": 0.9456,
+      "step": 921
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.782823589880728e-07,
+      "loss": 1.0782,
+      "step": 922
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.782302669161771e-07,
+      "loss": 1.1003,
+      "step": 923
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.781781152886623e-07,
+      "loss": 0.9141,
+      "step": 924
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.781259041191375e-07,
+      "loss": 0.9827,
+      "step": 925
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.780736334212266e-07,
+      "loss": 1.0352,
+      "step": 926
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.780213032085696e-07,
+      "loss": 1.1129,
+      "step": 927
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.779689134948216e-07,
+      "loss": 1.0418,
+      "step": 928
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.779164642936537e-07,
+      "loss": 1.0169,
+      "step": 929
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.778639556187522e-07,
+      "loss": 0.9729,
+      "step": 930
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.77811387483819e-07,
+      "loss": 1.0952,
+      "step": 931
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.777587599025713e-07,
+      "loss": 1.1345,
+      "step": 932
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.777060728887423e-07,
+      "loss": 1.0775,
+      "step": 933
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.776533264560803e-07,
+      "loss": 0.9491,
+      "step": 934
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.776005206183494e-07,
+      "loss": 1.097,
+      "step": 935
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.775476553893287e-07,
+      "loss": 1.1099,
+      "step": 936
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.774947307828134e-07,
+      "loss": 0.9147,
+      "step": 937
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.774417468126139e-07,
+      "loss": 1.0611,
+      "step": 938
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.77388703492556e-07,
+      "loss": 0.9906,
+      "step": 939
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.773356008364812e-07,
+      "loss": 0.8428,
+      "step": 940
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.772824388582464e-07,
+      "loss": 1.1568,
+      "step": 941
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.772292175717239e-07,
+      "loss": 0.9301,
+      "step": 942
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.771759369908016e-07,
+      "loss": 1.007,
+      "step": 943
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.771225971293828e-07,
+      "loss": 0.9458,
+      "step": 944
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.770691980013863e-07,
+      "loss": 1.0335,
+      "step": 945
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.770157396207463e-07,
+      "loss": 0.9398,
+      "step": 946
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.769622220014126e-07,
+      "loss": 1.048,
+      "step": 947
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.769086451573503e-07,
+      "loss": 1.0313,
+      "step": 948
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.768550091025401e-07,
+      "loss": 1.0231,
+      "step": 949
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.76801313850978e-07,
+      "loss": 0.9521,
+      "step": 950
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7674755941667553e-07,
+      "loss": 1.0687,
+      "step": 951
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7669374581365974e-07,
+      "loss": 1.0632,
+      "step": 952
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.766398730559729e-07,
+      "loss": 1.1324,
+      "step": 953
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.76585941157673e-07,
+      "loss": 0.962,
+      "step": 954
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.765319501328332e-07,
+      "loss": 1.0282,
+      "step": 955
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.764778999955423e-07,
+      "loss": 0.9964,
+      "step": 956
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.764237907599044e-07,
+      "loss": 1.0839,
+      "step": 957
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7636962244003904e-07,
+      "loss": 0.9089,
+      "step": 958
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7631539505008113e-07,
+      "loss": 0.9679,
+      "step": 959
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.762611086041812e-07,
+      "loss": 1.0205,
+      "step": 960
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.762067631165049e-07,
+      "loss": 1.0471,
+      "step": 961
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.761523586012334e-07,
+      "loss": 1.0802,
+      "step": 962
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.760978950725635e-07,
+      "loss": 1.0352,
+      "step": 963
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7604337254470706e-07,
+      "loss": 1.0698,
+      "step": 964
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.759887910318915e-07,
+      "loss": 0.9917,
+      "step": 965
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.759341505483596e-07,
+      "loss": 1.0902,
+      "step": 966
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.758794511083696e-07,
+      "loss": 1.0481,
+      "step": 967
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.75824692726195e-07,
+      "loss": 1.0712,
+      "step": 968
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.757698754161248e-07,
+      "loss": 1.1442,
+      "step": 969
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.757149991924633e-07,
+      "loss": 1.0078,
+      "step": 970
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7566006406953006e-07,
+      "loss": 1.0866,
+      "step": 971
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7560507006166036e-07,
+      "loss": 1.0289,
+      "step": 972
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.755500171832045e-07,
+      "loss": 1.0806,
+      "step": 973
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7549490544852823e-07,
+      "loss": 0.8827,
+      "step": 974
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.754397348720128e-07,
+      "loss": 0.9416,
+      "step": 975
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7538450546805473e-07,
+      "loss": 1.1389,
+      "step": 976
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7532921725106583e-07,
+      "loss": 1.0035,
+      "step": 977
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7527387023547326e-07,
+      "loss": 1.0718,
+      "step": 978
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7521846443571965e-07,
+      "loss": 0.9282,
+      "step": 979
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7516299986626277e-07,
+      "loss": 0.9203,
+      "step": 980
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7510747654157595e-07,
+      "loss": 0.9249,
+      "step": 981
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.750518944761477e-07,
+      "loss": 1.1242,
+      "step": 982
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7499625368448184e-07,
+      "loss": 0.951,
+      "step": 983
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.749405541810977e-07,
+      "loss": 1.0692,
+      "step": 984
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.748847959805297e-07,
+      "loss": 0.9706,
+      "step": 985
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7482897909732766e-07,
+      "loss": 1.056,
+      "step": 986
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.747731035460569e-07,
+      "loss": 0.9825,
+      "step": 987
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.747171693412977e-07,
+      "loss": 1.0949,
+      "step": 988
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.746611764976459e-07,
+      "loss": 1.0685,
+      "step": 989
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7460512502971263e-07,
+      "loss": 0.9853,
+      "step": 990
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7454901495212416e-07,
+      "loss": 1.1005,
+      "step": 991
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7449284627952215e-07,
+      "loss": 1.1005,
+      "step": 992
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.744366190265636e-07,
+      "loss": 0.9439,
+      "step": 993
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.743803332079208e-07,
+      "loss": 1.0693,
+      "step": 994
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7432398883828123e-07,
+      "loss": 0.9282,
+      "step": 995
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7426758593234764e-07,
+      "loss": 1.0953,
+      "step": 996
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7421112450483816e-07,
+      "loss": 0.9719,
+      "step": 997
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.741546045704861e-07,
+      "loss": 1.079,
+      "step": 998
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7409802614404005e-07,
+      "loss": 1.0509,
+      "step": 999
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.740413892402639e-07,
+      "loss": 1.1666,
+      "step": 1000
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.739846938739368e-07,
+      "loss": 1.0586,
+      "step": 1001
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.739279400598532e-07,
+      "loss": 1.0003,
+      "step": 1002
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7387112781282267e-07,
+      "loss": 0.969,
+      "step": 1003
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.738142571476701e-07,
+      "loss": 1.0787,
+      "step": 1004
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.737573280792356e-07,
+      "loss": 1.0416,
+      "step": 1005
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.737003406223747e-07,
+      "loss": 0.9989,
+      "step": 1006
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.736432947919578e-07,
+      "loss": 0.9241,
+      "step": 1007
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7358619060287087e-07,
+      "loss": 1.0772,
+      "step": 1008
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7352902807001495e-07,
+      "loss": 0.9392,
+      "step": 1009
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7347180720830627e-07,
+      "loss": 0.9029,
+      "step": 1010
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7341452803267643e-07,
+      "loss": 1.1357,
+      "step": 1011
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7335719055807223e-07,
+      "loss": 0.8775,
+      "step": 1012
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7329979479945535e-07,
+      "loss": 1.0081,
+      "step": 1013
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.732423407718032e-07,
+      "loss": 0.9958,
+      "step": 1014
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.731848284901081e-07,
+      "loss": 1.0363,
+      "step": 1015
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.731272579693776e-07,
+      "loss": 0.9874,
+      "step": 1016
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.730696292246343e-07,
+      "loss": 0.9627,
+      "step": 1017
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.730119422709164e-07,
+      "loss": 0.9931,
+      "step": 1018
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.729541971232769e-07,
+      "loss": 0.9507,
+      "step": 1019
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7289639379678404e-07,
+      "loss": 1.0971,
+      "step": 1020
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.728385323065215e-07,
+      "loss": 0.8672,
+      "step": 1021
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.727806126675879e-07,
+      "loss": 0.9152,
+      "step": 1022
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7272263489509713e-07,
+      "loss": 0.9942,
+      "step": 1023
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.726645990041781e-07,
+      "loss": 0.9639,
+      "step": 1024
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.726065050099751e-07,
+      "loss": 1.1629,
+      "step": 1025
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.725483529276475e-07,
+      "loss": 1.0825,
+      "step": 1026
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7249014277236975e-07,
+      "loss": 0.924,
+      "step": 1027
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7243187455933167e-07,
+      "loss": 1.1062,
+      "step": 1028
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.723735483037378e-07,
+      "loss": 0.9616,
+      "step": 1029
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7231516402080833e-07,
+      "loss": 1.2028,
+      "step": 1030
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7225672172577837e-07,
+      "loss": 1.0689,
+      "step": 1031
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.72198221433898e-07,
+      "loss": 0.9023,
+      "step": 1032
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7213966316043263e-07,
+      "loss": 1.0005,
+      "step": 1033
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.72081046920663e-07,
+      "loss": 0.9411,
+      "step": 1034
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.720223727298844e-07,
+      "loss": 0.9938,
+      "step": 1035
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.719636406034078e-07,
+      "loss": 1.1532,
+      "step": 1036
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7190485055655904e-07,
+      "loss": 0.984,
+      "step": 1037
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.718460026046791e-07,
+      "loss": 0.8958,
+      "step": 1038
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7178709676312414e-07,
+      "loss": 0.9634,
+      "step": 1039
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7172813304726524e-07,
+      "loss": 1.024,
+      "step": 1040
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.716691114724887e-07,
+      "loss": 0.9316,
+      "step": 1041
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.71610032054196e-07,
+      "loss": 0.9117,
+      "step": 1042
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7155089480780365e-07,
+      "loss": 1.1579,
+      "step": 1043
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.714916997487432e-07,
+      "loss": 1.0142,
+      "step": 1044
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.714324468924613e-07,
+      "loss": 1.1181,
+      "step": 1045
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.713731362544198e-07,
+      "loss": 0.9739,
+      "step": 1046
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7131376785009546e-07,
+      "loss": 1.1872,
+      "step": 1047
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7125434169498017e-07,
+      "loss": 0.9861,
+      "step": 1048
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7119485780458094e-07,
+      "loss": 1.1845,
+      "step": 1049
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.711353161944198e-07,
+      "loss": 1.1198,
+      "step": 1050
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.710757168800339e-07,
+      "loss": 1.0222,
+      "step": 1051
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7101605987697534e-07,
+      "loss": 1.0233,
+      "step": 1052
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7095634520081135e-07,
+      "loss": 0.9841,
+      "step": 1053
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.708965728671242e-07,
+      "loss": 1.1138,
+      "step": 1054
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.708367428915112e-07,
+      "loss": 1.0317,
+      "step": 1055
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7077685528958464e-07,
+      "loss": 0.9159,
+      "step": 1056
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7071691007697214e-07,
+      "loss": 0.8832,
+      "step": 1057
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7065690726931583e-07,
+      "loss": 0.9733,
+      "step": 1058
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.705968468822733e-07,
+      "loss": 0.9677,
+      "step": 1059
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.705367289315171e-07,
+      "loss": 1.1134,
+      "step": 1060
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.704765534327345e-07,
+      "loss": 0.9372,
+      "step": 1061
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.704163204016283e-07,
+      "loss": 0.9066,
+      "step": 1062
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7035602985391576e-07,
+      "loss": 1.1329,
+      "step": 1063
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.702956818053296e-07,
+      "loss": 0.9986,
+      "step": 1064
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.702352762716173e-07,
+      "loss": 0.8444,
+      "step": 1065
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.701748132685414e-07,
+      "loss": 1.0126,
+      "step": 1066
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.701142928118794e-07,
+      "loss": 0.9177,
+      "step": 1067
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7005371491742396e-07,
+      "loss": 1.1003,
+      "step": 1068
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6999307960098246e-07,
+      "loss": 0.9784,
+      "step": 1069
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6993238687837745e-07,
+      "loss": 1.1605,
+      "step": 1070
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.698716367654464e-07,
+      "loss": 0.951,
+      "step": 1071
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6981082927804173e-07,
+      "loss": 0.9466,
+      "step": 1072
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6974996443203096e-07,
+      "loss": 0.9597,
+      "step": 1073
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.696890422432965e-07,
+      "loss": 1.017,
+      "step": 1074
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6962806272773557e-07,
+      "loss": 1.0309,
+      "step": 1075
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6956702590126063e-07,
+      "loss": 1.1394,
+      "step": 1076
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.695059317797988e-07,
+      "loss": 1.0966,
+      "step": 1077
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.694447803792925e-07,
+      "loss": 1.0402,
+      "step": 1078
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6938357171569876e-07,
+      "loss": 0.9156,
+      "step": 1079
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6932230580498975e-07,
+      "loss": 0.9442,
+      "step": 1080
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.692609826631525e-07,
+      "loss": 1.0002,
+      "step": 1081
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6919960230618905e-07,
+      "loss": 0.9768,
+      "step": 1082
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.691381647501162e-07,
+      "loss": 1.2298,
+      "step": 1083
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6907667001096585e-07,
+      "loss": 1.1954,
+      "step": 1084
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.690151181047849e-07,
+      "loss": 0.9487,
+      "step": 1085
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6895350904763475e-07,
+      "loss": 1.0824,
+      "step": 1086
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6889184285559224e-07,
+      "loss": 1.2066,
+      "step": 1087
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.688301195447488e-07,
+      "loss": 1.0532,
+      "step": 1088
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.687683391312108e-07,
+      "loss": 0.9783,
+      "step": 1089
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.687065016310996e-07,
+      "loss": 0.9501,
+      "step": 1090
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.686446070605513e-07,
+      "loss": 1.0419,
+      "step": 1091
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6858265543571716e-07,
+      "loss": 1.0699,
+      "step": 1092
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.685206467727631e-07,
+      "loss": 1.0307,
+      "step": 1093
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6845858108786997e-07,
+      "loss": 1.0166,
+      "step": 1094
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.683964583972335e-07,
+      "loss": 1.1014,
+      "step": 1095
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6833427871706434e-07,
+      "loss": 0.8925,
+      "step": 1096
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.682720420635879e-07,
+      "loss": 1.2077,
+      "step": 1097
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6820974845304486e-07,
+      "loss": 0.9437,
+      "step": 1098
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6814739790169e-07,
+      "loss": 0.9478,
+      "step": 1099
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6808499042579376e-07,
+      "loss": 0.8249,
+      "step": 1100
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.680225260416408e-07,
+      "loss": 1.1044,
+      "step": 1101
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.679600047655312e-07,
+      "loss": 1.0547,
+      "step": 1102
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.678974266137793e-07,
+      "loss": 1.0896,
+      "step": 1103
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.678347916027148e-07,
+      "loss": 0.8514,
+      "step": 1104
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6777209974868194e-07,
+      "loss": 0.9512,
+      "step": 1105
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.677093510680399e-07,
+      "loss": 1.0739,
+      "step": 1106
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.676465455771625e-07,
+      "loss": 1.0823,
+      "step": 1107
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.675836832924387e-07,
+      "loss": 0.8494,
+      "step": 1108
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.675207642302721e-07,
+      "loss": 1.1751,
+      "step": 1109
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6745778840708107e-07,
+      "loss": 0.9982,
+      "step": 1110
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6739475583929884e-07,
+      "loss": 1.179,
+      "step": 1111
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6733166654337354e-07,
+      "loss": 1.0878,
+      "step": 1112
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6726852053576804e-07,
+      "loss": 1.0493,
+      "step": 1113
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6720531783295994e-07,
+      "loss": 0.9707,
+      "step": 1114
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.671420584514416e-07,
+      "loss": 1.027,
+      "step": 1115
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.670787424077205e-07,
+      "loss": 0.9255,
+      "step": 1116
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6701536971831845e-07,
+      "loss": 0.9128,
+      "step": 1117
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6695194039977237e-07,
+      "loss": 1.1871,
+      "step": 1118
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6688845446863376e-07,
+      "loss": 0.9964,
+      "step": 1119
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.668249119414691e-07,
+      "loss": 0.93,
+      "step": 1120
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6676131283485936e-07,
+      "loss": 1.0258,
+      "step": 1121
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6669765716540053e-07,
+      "loss": 0.9822,
+      "step": 1122
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6663394494970324e-07,
+      "loss": 1.1196,
+      "step": 1123
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.665701762043929e-07,
+      "loss": 1.0897,
+      "step": 1124
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6650635094610966e-07,
+      "loss": 0.8889,
+      "step": 1125
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.664424691915084e-07,
+      "loss": 1.1127,
+      "step": 1126
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6637853095725876e-07,
+      "loss": 1.0438,
+      "step": 1127
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6631453626004514e-07,
+      "loss": 1.0041,
+      "step": 1128
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.662504851165667e-07,
+      "loss": 1.0913,
+      "step": 1129
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6618637754353724e-07,
+      "loss": 0.9219,
+      "step": 1130
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.661222135576853e-07,
+      "loss": 1.0344,
+      "step": 1131
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.660579931757542e-07,
+      "loss": 1.0652,
+      "step": 1132
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.65993716414502e-07,
+      "loss": 1.0703,
+      "step": 1133
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.659293832907014e-07,
+      "loss": 1.0794,
+      "step": 1134
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6586499382113976e-07,
+      "loss": 1.063,
+      "step": 1135
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6580054802261925e-07,
+      "loss": 0.9897,
+      "step": 1136
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6573604591195674e-07,
+      "loss": 1.0533,
+      "step": 1137
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.656714875059837e-07,
+      "loss": 0.9232,
+      "step": 1138
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.656068728215463e-07,
+      "loss": 1.0356,
+      "step": 1139
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6554220187550566e-07,
+      "loss": 1.0567,
+      "step": 1140
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6547747468473705e-07,
+      "loss": 1.0065,
+      "step": 1141
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6541269126613093e-07,
+      "loss": 1.0341,
+      "step": 1142
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.653478516365921e-07,
+      "loss": 0.8671,
+      "step": 1143
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.652829558130403e-07,
+      "loss": 0.8443,
+      "step": 1144
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.652180038124097e-07,
+      "loss": 1.0223,
+      "step": 1145
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6515299565164936e-07,
+      "loss": 1.1633,
+      "step": 1146
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.650879313477226e-07,
+      "loss": 0.8576,
+      "step": 1147
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.650228109176079e-07,
+      "loss": 1.1244,
+      "step": 1148
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6495763437829793e-07,
+      "loss": 0.9631,
+      "step": 1149
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6489240174680026e-07,
+      "loss": 1.0434,
+      "step": 1150
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.648271130401371e-07,
+      "loss": 0.9463,
+      "step": 1151
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6476176827534526e-07,
+      "loss": 1.0866,
+      "step": 1152
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6469636746947603e-07,
+      "loss": 1.102,
+      "step": 1153
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6463091063959555e-07,
+      "loss": 0.9272,
+      "step": 1154
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.645653978027844e-07,
+      "loss": 1.0304,
+      "step": 1155
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.644998289761379e-07,
+      "loss": 0.973,
+      "step": 1156
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.644342041767658e-07,
+      "loss": 0.9331,
+      "step": 1157
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.643685234217928e-07,
+      "loss": 1.0575,
+      "step": 1158
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6430278672835785e-07,
+      "loss": 0.9137,
+      "step": 1159
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.642369941136147e-07,
+      "loss": 0.9178,
+      "step": 1160
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6417114559473146e-07,
+      "loss": 1.1616,
+      "step": 1161
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.641052411888913e-07,
+      "loss": 0.9768,
+      "step": 1162
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6403928091329125e-07,
+      "loss": 0.9456,
+      "step": 1163
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6397326478514377e-07,
+      "loss": 1.0654,
+      "step": 1164
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.639071928216751e-07,
+      "loss": 0.9718,
+      "step": 1165
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6384106504012665e-07,
+      "loss": 1.0452,
+      "step": 1166
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.63774881457754e-07,
+      "loss": 0.9318,
+      "step": 1167
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.637086420918276e-07,
+      "loss": 0.963,
+      "step": 1168
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.636423469596322e-07,
+      "loss": 0.9084,
+      "step": 1169
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6357599607846713e-07,
+      "loss": 1.0244,
+      "step": 1170
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.635095894656465e-07,
+      "loss": 1.0676,
+      "step": 1171
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6344312713849877e-07,
+      "loss": 0.9395,
+      "step": 1172
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.633766091143668e-07,
+      "loss": 1.0447,
+      "step": 1173
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6331003541060844e-07,
+      "loss": 0.8776,
+      "step": 1174
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.632434060445956e-07,
+      "loss": 0.9541,
+      "step": 1175
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.631767210337149e-07,
+      "loss": 1.0583,
+      "step": 1176
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6310998039536756e-07,
+      "loss": 1.0472,
+      "step": 1177
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6304318414696917e-07,
+      "loss": 0.9281,
+      "step": 1178
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6297633230595e-07,
+      "loss": 0.8657,
+      "step": 1179
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6290942488975457e-07,
+      "loss": 1.1261,
+      "step": 1180
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.628424619158422e-07,
+      "loss": 0.9081,
+      "step": 1181
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.627754434016865e-07,
+      "loss": 1.0312,
+      "step": 1182
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6270836936477564e-07,
+      "loss": 1.114,
+      "step": 1183
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6264123982261226e-07,
+      "loss": 1.0557,
+      "step": 1184
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.625740547927135e-07,
+      "loss": 0.932,
+      "step": 1185
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6250681429261104e-07,
+      "loss": 0.9766,
+      "step": 1186
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6243951833985085e-07,
+      "loss": 0.9263,
+      "step": 1187
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6237216695199367e-07,
+      "loss": 1.1236,
+      "step": 1188
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.623047601466143e-07,
+      "loss": 0.9601,
+      "step": 1189
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.622372979413025e-07,
+      "loss": 1.0282,
+      "step": 1190
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6216978035366205e-07,
+      "loss": 1.0283,
+      "step": 1191
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6210220740131135e-07,
+      "loss": 0.9756,
+      "step": 1192
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6203457910188324e-07,
+      "loss": 0.9281,
+      "step": 1193
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6196689547302514e-07,
+      "loss": 0.9328,
+      "step": 1194
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.618991565323986e-07,
+      "loss": 0.9859,
+      "step": 1195
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6183136229767985e-07,
+      "loss": 1.0267,
+      "step": 1196
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6176351278655957e-07,
+      "loss": 0.9594,
+      "step": 1197
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.616956080167426e-07,
+      "loss": 1.0237,
+      "step": 1198
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.616276480059484e-07,
+      "loss": 1.0538,
+      "step": 1199
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.61559632771911e-07,
+      "loss": 0.8726,
+      "step": 1200
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.614915623323785e-07,
+      "loss": 0.9752,
+      "step": 1201
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.614234367051135e-07,
+      "loss": 1.0751,
+      "step": 1202
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6135525590789326e-07,
+      "loss": 1.0438,
+      "step": 1203
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.612870199585091e-07,
+      "loss": 1.0673,
+      "step": 1204
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6121872887476686e-07,
+      "loss": 1.0664,
+      "step": 1205
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.611503826744869e-07,
+      "loss": 1.0694,
+      "step": 1206
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6108198137550377e-07,
+      "loss": 1.204,
+      "step": 1207
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6101352499566637e-07,
+      "loss": 1.0051,
+      "step": 1208
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6094501355283826e-07,
+      "loss": 1.0232,
+      "step": 1209
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6087644706489703e-07,
+      "loss": 1.1127,
+      "step": 1210
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6080782554973487e-07,
+      "loss": 0.9796,
+      "step": 1211
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.607391490252582e-07,
+      "loss": 0.9546,
+      "step": 1212
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6067041750938787e-07,
+      "loss": 1.1582,
+      "step": 1213
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.60601631020059e-07,
+      "loss": 1.1834,
+      "step": 1214
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6053278957522115e-07,
+      "loss": 1.11,
+      "step": 1215
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.604638931928382e-07,
+      "loss": 0.9932,
+      "step": 1216
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.603949418908881e-07,
+      "loss": 1.1546,
+      "step": 1217
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6032593568736376e-07,
+      "loss": 1.1941,
+      "step": 1218
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.602568746002717e-07,
+      "loss": 0.8793,
+      "step": 1219
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6018775864763325e-07,
+      "loss": 1.0475,
+      "step": 1220
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.601185878474839e-07,
+      "loss": 1.0297,
+      "step": 1221
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.600493622178734e-07,
+      "loss": 0.9226,
+      "step": 1222
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5998008177686585e-07,
+      "loss": 0.9918,
+      "step": 1223
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.599107465425396e-07,
+      "loss": 0.8541,
+      "step": 1224
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.598413565329875e-07,
+      "loss": 1.0462,
+      "step": 1225
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5977191176631636e-07,
+      "loss": 1.144,
+      "step": 1226
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5970241226064764e-07,
+      "loss": 1.0635,
+      "step": 1227
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.596328580341169e-07,
+      "loss": 1.034,
+      "step": 1228
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.595632491048738e-07,
+      "loss": 0.9913,
+      "step": 1229
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.594935854910827e-07,
+      "loss": 1.1409,
+      "step": 1230
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5942386721092186e-07,
+      "loss": 1.1548,
+      "step": 1231
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.593540942825839e-07,
+      "loss": 0.9978,
+      "step": 1232
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.592842667242759e-07,
+      "loss": 1.098,
+      "step": 1233
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.592143845542189e-07,
+      "loss": 1.1238,
+      "step": 1234
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5914444779064835e-07,
+      "loss": 0.9649,
+      "step": 1235
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5907445645181395e-07,
+      "loss": 0.9675,
+      "step": 1236
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.590044105559796e-07,
+      "loss": 1.0103,
+      "step": 1237
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5893431012142345e-07,
+      "loss": 1.0895,
+      "step": 1238
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.588641551664379e-07,
+      "loss": 1.0725,
+      "step": 1239
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.587939457093295e-07,
+      "loss": 1.0162,
+      "step": 1240
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.587236817684191e-07,
+      "loss": 0.9065,
+      "step": 1241
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.586533633620418e-07,
+      "loss": 0.925,
+      "step": 1242
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5858299050854674e-07,
+      "loss": 1.0005,
+      "step": 1243
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.585125632262976e-07,
+      "loss": 1.0797,
+      "step": 1244
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.584420815336718e-07,
+      "loss": 0.9047,
+      "step": 1245
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5837154544906133e-07,
+      "loss": 1.0085,
+      "step": 1246
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.583009549908723e-07,
+      "loss": 0.88,
+      "step": 1247
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.582303101775248e-07,
+      "loss": 1.0063,
+      "step": 1248
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.581596110274535e-07,
+      "loss": 0.8634,
+      "step": 1249
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5808885755910673e-07,
+      "loss": 0.9401,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5801804979094755e-07,
+      "loss": 1.0416,
+      "step": 1251
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.579471877414527e-07,
+      "loss": 1.1385,
+      "step": 1252
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5787627142911345e-07,
+      "loss": 1.0073,
+      "step": 1253
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.57805300872435e-07,
+      "loss": 0.8769,
+      "step": 1254
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5773427608993675e-07,
+      "loss": 1.0099,
+      "step": 1255
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.576631971001524e-07,
+      "loss": 1.1268,
+      "step": 1256
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.575920639216296e-07,
+      "loss": 1.0641,
+      "step": 1257
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.575208765729302e-07,
+      "loss": 0.926,
+      "step": 1258
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.574496350726302e-07,
+      "loss": 0.9161,
+      "step": 1259
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5737833943931993e-07,
+      "loss": 0.8158,
+      "step": 1260
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.573069896916034e-07,
+      "loss": 0.976,
+      "step": 1261
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5723558584809905e-07,
+      "loss": 1.12,
+      "step": 1262
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5716412792743955e-07,
+      "loss": 1.177,
+      "step": 1263
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.570926159482712e-07,
+      "loss": 0.9338,
+      "step": 1264
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.57021049929255e-07,
+      "loss": 1.0785,
+      "step": 1265
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5694942988906565e-07,
+      "loss": 1.0972,
+      "step": 1266
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.568777558463921e-07,
+      "loss": 0.9888,
+      "step": 1267
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.568060278199373e-07,
+      "loss": 0.9386,
+      "step": 1268
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5673424582841837e-07,
+      "loss": 1.0706,
+      "step": 1269
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.566624098905665e-07,
+      "loss": 0.9652,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.565905200251269e-07,
+      "loss": 0.9064,
+      "step": 1271
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.565185762508589e-07,
+      "loss": 0.8494,
+      "step": 1272
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.564465785865359e-07,
+      "loss": 0.9494,
+      "step": 1273
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.563745270509455e-07,
+      "loss": 1.035,
+      "step": 1274
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.563024216628889e-07,
+      "loss": 1.2957,
+      "step": 1275
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.56230262441182e-07,
+      "loss": 1.0929,
+      "step": 1276
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.561580494046542e-07,
+      "loss": 1.0285,
+      "step": 1277
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5608578257214916e-07,
+      "loss": 1.0813,
+      "step": 1278
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5601346196252465e-07,
+      "loss": 1.0815,
+      "step": 1279
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5594108759465235e-07,
+      "loss": 1.094,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.55868659487418e-07,
+      "loss": 1.0392,
+      "step": 1281
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5579617765972145e-07,
+      "loss": 0.9935,
+      "step": 1282
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5572364213047636e-07,
+      "loss": 0.9295,
+      "step": 1283
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.556510529186107e-07,
+      "loss": 0.9284,
+      "step": 1284
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.555784100430662e-07,
+      "loss": 0.8905,
+      "step": 1285
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5550571352279863e-07,
+      "loss": 1.115,
+      "step": 1286
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.554329633767779e-07,
+      "loss": 0.9387,
+      "step": 1287
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.553601596239877e-07,
+      "loss": 0.9228,
+      "step": 1288
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.55287302283426e-07,
+      "loss": 1.0889,
+      "step": 1289
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5521439137410444e-07,
+      "loss": 1.1619,
+      "step": 1290
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5514142691504887e-07,
+      "loss": 0.942,
+      "step": 1291
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.550684089252989e-07,
+      "loss": 0.9718,
+      "step": 1292
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.549953374239084e-07,
+      "loss": 1.0151,
+      "step": 1293
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.549222124299449e-07,
+      "loss": 1.0944,
+      "step": 1294
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5484903396249006e-07,
+      "loss": 1.1314,
+      "step": 1295
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.547758020406395e-07,
+      "loss": 1.0408,
+      "step": 1296
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.547025166835027e-07,
+      "loss": 1.1178,
+      "step": 1297
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5462917791020315e-07,
+      "loss": 1.0759,
+      "step": 1298
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.545557857398783e-07,
+      "loss": 1.1347,
+      "step": 1299
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.544823401916794e-07,
+      "loss": 0.9926,
+      "step": 1300
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.544088412847718e-07,
+      "loss": 1.0898,
+      "step": 1301
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5433528903833463e-07,
+      "loss": 0.9547,
+      "step": 1302
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5426168347156114e-07,
+      "loss": 1.0457,
+      "step": 1303
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5418802460365814e-07,
+      "loss": 1.1351,
+      "step": 1304
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5411431245384677e-07,
+      "loss": 1.126,
+      "step": 1305
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5404054704136176e-07,
+      "loss": 1.141,
+      "step": 1306
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5396672838545184e-07,
+      "loss": 1.0428,
+      "step": 1307
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.538928565053798e-07,
+      "loss": 1.032,
+      "step": 1308
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5381893142042195e-07,
+      "loss": 1.1437,
+      "step": 1309
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.537449531498687e-07,
+      "loss": 1.0657,
+      "step": 1310
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5367092171302457e-07,
+      "loss": 1.092,
+      "step": 1311
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5359683712920746e-07,
+      "loss": 1.0999,
+      "step": 1312
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5352269941774955e-07,
+      "loss": 0.9259,
+      "step": 1313
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5344850859799665e-07,
+      "loss": 0.9362,
+      "step": 1314
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.533742646893086e-07,
+      "loss": 0.9785,
+      "step": 1315
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.532999677110588e-07,
+      "loss": 0.9945,
+      "step": 1316
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.53225617682635e-07,
+      "loss": 1.0216,
+      "step": 1317
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.531512146234382e-07,
+      "loss": 1.0442,
+      "step": 1318
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5307675855288373e-07,
+      "loss": 0.984,
+      "step": 1319
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5300224949040047e-07,
+      "loss": 0.936,
+      "step": 1320
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5292768745543113e-07,
+      "loss": 0.9254,
+      "step": 1321
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5285307246743255e-07,
+      "loss": 1.2074,
+      "step": 1322
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.527784045458749e-07,
+      "loss": 1.1527,
+      "step": 1323
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5270368371024257e-07,
+      "loss": 1.1145,
+      "step": 1324
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5262890998003365e-07,
+      "loss": 1.0556,
+      "step": 1325
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.525540833747599e-07,
+      "loss": 1.1369,
+      "step": 1326
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.52479203913947e-07,
+      "loss": 0.9691,
+      "step": 1327
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.524042716171344e-07,
+      "loss": 1.0983,
+      "step": 1328
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5232928650387537e-07,
+      "loss": 1.1045,
+      "step": 1329
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5225424859373684e-07,
+      "loss": 1.0199,
+      "step": 1330
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5217915790629966e-07,
+      "loss": 0.9506,
+      "step": 1331
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5210401446115844e-07,
+      "loss": 0.8799,
+      "step": 1332
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.520288182779213e-07,
+      "loss": 1.2559,
+      "step": 1333
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5195356937621063e-07,
+      "loss": 0.9303,
+      "step": 1334
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5187826777566206e-07,
+      "loss": 1.0883,
+      "step": 1335
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5180291349592523e-07,
+      "loss": 1.0019,
+      "step": 1336
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5172750655666364e-07,
+      "loss": 1.0763,
+      "step": 1337
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.516520469775542e-07,
+      "loss": 0.9074,
+      "step": 1338
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.515765347782877e-07,
+      "loss": 1.0362,
+      "step": 1339
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.515009699785689e-07,
+      "loss": 0.8412,
+      "step": 1340
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5142535259811594e-07,
+      "loss": 0.9583,
+      "step": 1341
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.513496826566608e-07,
+      "loss": 0.926,
+      "step": 1342
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5127396017394924e-07,
+      "loss": 1.1108,
+      "step": 1343
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5119818516974074e-07,
+      "loss": 1.0724,
+      "step": 1344
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5112235766380834e-07,
+      "loss": 1.1165,
+      "step": 1345
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5104647767593895e-07,
+      "loss": 0.9893,
+      "step": 1346
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5097054522593305e-07,
+      "loss": 1.0162,
+      "step": 1347
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5089456033360483e-07,
+      "loss": 1.0843,
+      "step": 1348
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5081852301878227e-07,
+      "loss": 1.0054,
+      "step": 1349
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.507424333013069e-07,
+      "loss": 1.0947,
+      "step": 1350
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5066629120103394e-07,
+      "loss": 1.081,
+      "step": 1351
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5059009673783244e-07,
+      "loss": 1.0418,
+      "step": 1352
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.505138499315848e-07,
+      "loss": 1.0083,
+      "step": 1353
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.504375508021875e-07,
+      "loss": 1.0067,
+      "step": 1354
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5036119936955023e-07,
+      "loss": 1.076,
+      "step": 1355
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.502847956535967e-07,
+      "loss": 1.0445,
+      "step": 1356
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5020833967426397e-07,
+      "loss": 0.878,
+      "step": 1357
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.501318314515029e-07,
+      "loss": 0.967,
+      "step": 1358
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.50055271005278e-07,
+      "loss": 0.9324,
+      "step": 1359
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4997865835556734e-07,
+      "loss": 1.053,
+      "step": 1360
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.499019935223627e-07,
+      "loss": 0.9512,
+      "step": 1361
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4982527652566924e-07,
+      "loss": 1.1842,
+      "step": 1362
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.497485073855061e-07,
+      "loss": 1.0581,
+      "step": 1363
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.496716861219056e-07,
+      "loss": 1.0573,
+      "step": 1364
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.495948127549141e-07,
+      "loss": 1.1228,
+      "step": 1365
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.495178873045913e-07,
+      "loss": 1.0582,
+      "step": 1366
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.494409097910104e-07,
+      "loss": 0.9345,
+      "step": 1367
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.493638802342585e-07,
+      "loss": 1.0017,
+      "step": 1368
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4928679865443597e-07,
+      "loss": 1.1187,
+      "step": 1369
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.492096650716569e-07,
+      "loss": 0.9371,
+      "step": 1370
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4913247950604903e-07,
+      "loss": 0.9673,
+      "step": 1371
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4905524197775355e-07,
+      "loss": 1.0485,
+      "step": 1372
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4897795250692515e-07,
+      "loss": 1.0236,
+      "step": 1373
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.489006111137322e-07,
+      "loss": 1.0643,
+      "step": 1374
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4882321781835663e-07,
+      "loss": 1.0224,
+      "step": 1375
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.487457726409937e-07,
+      "loss": 1.095,
+      "step": 1376
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.486682756018526e-07,
+      "loss": 1.0162,
+      "step": 1377
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.485907267211556e-07,
+      "loss": 1.1433,
+      "step": 1378
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.485131260191388e-07,
+      "loss": 0.9747,
+      "step": 1379
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.484354735160518e-07,
+      "loss": 0.8382,
+      "step": 1380
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4835776923215763e-07,
+      "loss": 0.983,
+      "step": 1381
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4828001318773276e-07,
+      "loss": 0.9476,
+      "step": 1382
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4820220540306736e-07,
+      "loss": 0.9457,
+      "step": 1383
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.48124345898465e-07,
+      "loss": 1.0253,
+      "step": 1384
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4804643469424274e-07,
+      "loss": 0.8363,
+      "step": 1385
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.479684718107311e-07,
+      "loss": 0.846,
+      "step": 1386
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.478904572682742e-07,
+      "loss": 0.9271,
+      "step": 1387
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4781239108722955e-07,
+      "loss": 1.0804,
+      "step": 1388
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.477342732879682e-07,
+      "loss": 0.9033,
+      "step": 1389
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.476561038908745e-07,
+      "loss": 0.9682,
+      "step": 1390
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.475778829163465e-07,
+      "loss": 1.1425,
+      "step": 1391
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4749961038479544e-07,
+      "loss": 1.1361,
+      "step": 1392
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.474212863166464e-07,
+      "loss": 1.1194,
+      "step": 1393
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4734291073233754e-07,
+      "loss": 1.1782,
+      "step": 1394
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.472644836523206e-07,
+      "loss": 0.9984,
+      "step": 1395
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.471860050970608e-07,
+      "loss": 0.9513,
+      "step": 1396
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4710747508703675e-07,
+      "loss": 0.9327,
+      "step": 1397
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.470288936427404e-07,
+      "loss": 1.1993,
+      "step": 1398
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4695026078467735e-07,
+      "loss": 1.0537,
+      "step": 1399
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4687157653336635e-07,
+      "loss": 0.9293,
+      "step": 1400
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.467928409093398e-07,
+      "loss": 1.0838,
+      "step": 1401
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.467140539331433e-07,
+      "loss": 0.9661,
+      "step": 1402
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.466352156253361e-07,
+      "loss": 0.9743,
+      "step": 1403
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.465563260064904e-07,
+      "loss": 1.1651,
+      "step": 1404
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.464773850971924e-07,
+      "loss": 1.0266,
+      "step": 1405
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.463983929180411e-07,
+      "loss": 1.1174,
+      "step": 1406
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.463193494896492e-07,
+      "loss": 1.1905,
+      "step": 1407
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.462402548326429e-07,
+      "loss": 0.946,
+      "step": 1408
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4616110896766136e-07,
+      "loss": 1.0734,
+      "step": 1409
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4608191191535736e-07,
+      "loss": 0.9741,
+      "step": 1410
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4600266369639705e-07,
+      "loss": 1.0915,
+      "step": 1411
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.459233643314599e-07,
+      "loss": 0.8883,
+      "step": 1412
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.458440138412387e-07,
+      "loss": 0.8033,
+      "step": 1413
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4576461224643957e-07,
+      "loss": 0.9681,
+      "step": 1414
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.45685159567782e-07,
+      "loss": 1.0789,
+      "step": 1415
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.456056558259986e-07,
+      "loss": 1.1905,
+      "step": 1416
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4552610104183584e-07,
+      "loss": 0.9718,
+      "step": 1417
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4544649523605294e-07,
+      "loss": 1.0401,
+      "step": 1418
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.453668384294227e-07,
+      "loss": 1.0891,
+      "step": 1419
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.452871306427313e-07,
+      "loss": 0.9489,
+      "step": 1420
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.45207371896778e-07,
+      "loss": 0.9019,
+      "step": 1421
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4512756221237547e-07,
+      "loss": 1.0367,
+      "step": 1422
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.450477016103497e-07,
+      "loss": 0.9278,
+      "step": 1423
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4496779011154e-07,
+      "loss": 0.8788,
+      "step": 1424
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4488782773679877e-07,
+      "loss": 0.9663,
+      "step": 1425
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.44807814506992e-07,
+      "loss": 0.8876,
+      "step": 1426
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.447277504429985e-07,
+      "loss": 1.0696,
+      "step": 1427
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4464763556571086e-07,
+      "loss": 1.05,
+      "step": 1428
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.445674698960346e-07,
+      "loss": 1.0458,
+      "step": 1429
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.444872534548886e-07,
+      "loss": 1.0079,
+      "step": 1430
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.444069862632048e-07,
+      "loss": 1.0869,
+      "step": 1431
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4432666834192877e-07,
+      "loss": 0.9555,
+      "step": 1432
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4424629971201906e-07,
+      "loss": 0.9052,
+      "step": 1433
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4416588039444733e-07,
+      "loss": 1.0483,
+      "step": 1434
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4408541041019876e-07,
+      "loss": 0.9933,
+      "step": 1435
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.440048897802716e-07,
+      "loss": 1.0092,
+      "step": 1436
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4392431852567734e-07,
+      "loss": 0.9257,
+      "step": 1437
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4384369666744057e-07,
+      "loss": 1.0354,
+      "step": 1438
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.437630242265994e-07,
+      "loss": 1.1994,
+      "step": 1439
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4368230122420464e-07,
+      "loss": 1.1882,
+      "step": 1440
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4360152768132075e-07,
+      "loss": 0.9621,
+      "step": 1441
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.435207036190253e-07,
+      "loss": 0.8998,
+      "step": 1442
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.434398290584087e-07,
+      "loss": 1.0445,
+      "step": 1443
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.43358904020575e-07,
+      "loss": 0.8824,
+      "step": 1444
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.43277928526641e-07,
+      "loss": 0.9039,
+      "step": 1445
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4319690259773716e-07,
+      "loss": 1.1223,
+      "step": 1446
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4311582625500664e-07,
+      "loss": 1.0328,
+      "step": 1447
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.430346995196059e-07,
+      "loss": 1.1063,
+      "step": 1448
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.429535224127047e-07,
+      "loss": 0.8915,
+      "step": 1449
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4287229495548573e-07,
+      "loss": 1.1712,
+      "step": 1450
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.427910171691449e-07,
+      "loss": 1.1179,
+      "step": 1451
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.427096890748914e-07,
+      "loss": 1.1097,
+      "step": 1452
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.426283106939473e-07,
+      "loss": 1.1319,
+      "step": 1453
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.425468820475479e-07,
+      "loss": 0.9583,
+      "step": 1454
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4246540315694167e-07,
+      "loss": 1.0402,
+      "step": 1455
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.423838740433902e-07,
+      "loss": 1.107,
+      "step": 1456
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4230229472816807e-07,
+      "loss": 1.0167,
+      "step": 1457
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.42220665232563e-07,
+      "loss": 1.0027,
+      "step": 1458
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4213898557787577e-07,
+      "loss": 1.0866,
+      "step": 1459
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4205725578542043e-07,
+      "loss": 0.9894,
+      "step": 1460
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.419754758765238e-07,
+      "loss": 0.9806,
+      "step": 1461
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.418936458725263e-07,
+      "loss": 1.059,
+      "step": 1462
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.418117657947808e-07,
+      "loss": 1.036,
+      "step": 1463
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.417298356646536e-07,
+      "loss": 1.1368,
+      "step": 1464
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.41647855503524e-07,
+      "loss": 1.0275,
+      "step": 1465
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.415658253327843e-07,
+      "loss": 0.9378,
+      "step": 1466
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.414837451738399e-07,
+      "loss": 1.0102,
+      "step": 1467
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.414016150481093e-07,
+      "loss": 1.1281,
+      "step": 1468
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.413194349770239e-07,
+      "loss": 1.179,
+      "step": 1469
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.412372049820282e-07,
+      "loss": 1.1901,
+      "step": 1470
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.411549250845798e-07,
+      "loss": 0.9309,
+      "step": 1471
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.410725953061492e-07,
+      "loss": 1.0337,
+      "step": 1472
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.409902156682199e-07,
+      "loss": 0.8472,
+      "step": 1473
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4090778619228867e-07,
+      "loss": 1.008,
+      "step": 1474
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4082530689986496e-07,
+      "loss": 0.9987,
+      "step": 1475
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.407427778124714e-07,
+      "loss": 0.9994,
+      "step": 1476
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.406601989516435e-07,
+      "loss": 0.9994,
+      "step": 1477
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.405775703389298e-07,
+      "loss": 1.0763,
+      "step": 1478
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4049489199589206e-07,
+      "loss": 0.9546,
+      "step": 1479
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4041216394410464e-07,
+      "loss": 1.064,
+      "step": 1480
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.403293862051551e-07,
+      "loss": 0.9927,
+      "step": 1481
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.402465588006438e-07,
+      "loss": 0.9177,
+      "step": 1482
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4016368175218427e-07,
+      "loss": 0.9132,
+      "step": 1483
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.400807550814029e-07,
+      "loss": 1.1143,
+      "step": 1484
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3999777880993895e-07,
+      "loss": 0.9065,
+      "step": 1485
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3991475295944465e-07,
+      "loss": 0.8949,
+      "step": 1486
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3983167755158524e-07,
+      "loss": 1.245,
+      "step": 1487
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.39748552608039e-07,
+      "loss": 1.0231,
+      "step": 1488
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.396653781504968e-07,
+      "loss": 0.9946,
+      "step": 1489
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.395821542006627e-07,
+      "loss": 0.9825,
+      "step": 1490
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3949888078025366e-07,
+      "loss": 0.9219,
+      "step": 1491
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3941555791099937e-07,
+      "loss": 1.0145,
+      "step": 1492
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3933218561464267e-07,
+      "loss": 0.9054,
+      "step": 1493
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.392487639129391e-07,
+      "loss": 1.0284,
+      "step": 1494
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.391652928276571e-07,
+      "loss": 0.9854,
+      "step": 1495
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.390817723805782e-07,
+      "loss": 1.128,
+      "step": 1496
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3899820259349655e-07,
+      "loss": 1.0283,
+      "step": 1497
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.389145834882194e-07,
+      "loss": 1.1028,
+      "step": 1498
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3883091508656666e-07,
+      "loss": 1.0408,
+      "step": 1499
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3874719741037124e-07,
+      "loss": 1.0097,
+      "step": 1500
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.386634304814789e-07,
+      "loss": 1.0312,
+      "step": 1501
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3857961432174825e-07,
+      "loss": 0.9477,
+      "step": 1502
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3849574895305066e-07,
+      "loss": 1.0826,
+      "step": 1503
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3841183439727037e-07,
+      "loss": 1.0501,
+      "step": 1504
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.383278706763046e-07,
+      "loss": 1.0155,
+      "step": 1505
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.382438578120633e-07,
+      "loss": 1.1267,
+      "step": 1506
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.381597958264691e-07,
+      "loss": 0.8765,
+      "step": 1507
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3807568474145773e-07,
+      "loss": 1.0107,
+      "step": 1508
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.379915245789775e-07,
+      "loss": 1.0793,
+      "step": 1509
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.379073153609896e-07,
+      "loss": 0.9629,
+      "step": 1510
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.37823057109468e-07,
+      "loss": 0.9372,
+      "step": 1511
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.377387498463996e-07,
+      "loss": 1.07,
+      "step": 1512
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3765439359378397e-07,
+      "loss": 0.9907,
+      "step": 1513
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3756998837363336e-07,
+      "loss": 1.0099,
+      "step": 1514
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.374855342079731e-07,
+      "loss": 1.1852,
+      "step": 1515
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.374010311188409e-07,
+      "loss": 1.0699,
+      "step": 1516
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3731647912828763e-07,
+      "loss": 0.8555,
+      "step": 1517
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3723187825837664e-07,
+      "loss": 1.0071,
+      "step": 1518
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.371472285311841e-07,
+      "loss": 1.1565,
+      "step": 1519
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.370625299687991e-07,
+      "loss": 1.1859,
+      "step": 1520
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.369777825933232e-07,
+      "loss": 1.1022,
+      "step": 1521
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.368929864268709e-07,
+      "loss": 1.0207,
+      "step": 1522
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.368081414915693e-07,
+      "loss": 0.9165,
+      "step": 1523
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.367232478095584e-07,
+      "loss": 0.9654,
+      "step": 1524
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.366383054029906e-07,
+      "loss": 1.0201,
+      "step": 1525
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.365533142940315e-07,
+      "loss": 1.127,
+      "step": 1526
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.364682745048589e-07,
+      "loss": 0.9779,
+      "step": 1527
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.363831860576637e-07,
+      "loss": 1.0059,
+      "step": 1528
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.362980489746492e-07,
+      "loss": 0.9734,
+      "step": 1529
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3621286327803155e-07,
+      "loss": 1.1246,
+      "step": 1530
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.361276289900396e-07,
+      "loss": 0.9274,
+      "step": 1531
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.360423461329148e-07,
+      "loss": 1.024,
+      "step": 1532
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3595701472891147e-07,
+      "loss": 1.0621,
+      "step": 1533
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3587163480029615e-07,
+      "loss": 1.2278,
+      "step": 1534
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.357862063693485e-07,
+      "loss": 1.043,
+      "step": 1535
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.357007294583608e-07,
+      "loss": 1.12,
+      "step": 1536
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.356152040896376e-07,
+      "loss": 0.9503,
+      "step": 1537
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3552963028549644e-07,
+      "loss": 0.9688,
+      "step": 1538
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3544400806826743e-07,
+      "loss": 1.027,
+      "step": 1539
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3535833746029327e-07,
+      "loss": 1.1482,
+      "step": 1540
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.352726184839294e-07,
+      "loss": 0.9478,
+      "step": 1541
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.351868511615435e-07,
+      "loss": 0.9674,
+      "step": 1542
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3510103551551645e-07,
+      "loss": 0.9926,
+      "step": 1543
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3501517156824136e-07,
+      "loss": 1.0539,
+      "step": 1544
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3492925934212397e-07,
+      "loss": 1.1475,
+      "step": 1545
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.348432988595827e-07,
+      "loss": 0.9676,
+      "step": 1546
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3475729014304854e-07,
+      "loss": 1.104,
+      "step": 1547
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.346712332149651e-07,
+      "loss": 1.0585,
+      "step": 1548
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.345851280977884e-07,
+      "loss": 0.9693,
+      "step": 1549
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.344989748139873e-07,
+      "loss": 1.0339,
+      "step": 1550
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3441277338604303e-07,
+      "loss": 0.9049,
+      "step": 1551
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.343265238364495e-07,
+      "loss": 0.9602,
+      "step": 1552
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3424022618771306e-07,
+      "loss": 0.909,
+      "step": 1553
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.341538804623528e-07,
+      "loss": 0.9472,
+      "step": 1554
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3406748668290005e-07,
+      "loss": 0.8406,
+      "step": 1555
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3398104487189903e-07,
+      "loss": 1.1254,
+      "step": 1556
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.338945550519062e-07,
+      "loss": 0.9967,
+      "step": 1557
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.338080172454907e-07,
+      "loss": 1.0682,
+      "step": 1558
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.337214314752342e-07,
+      "loss": 0.94,
+      "step": 1559
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.336347977637309e-07,
+      "loss": 1.0292,
+      "step": 1560
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3354811613358743e-07,
+      "loss": 1.1367,
+      "step": 1561
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3346138660742285e-07,
+      "loss": 0.9148,
+      "step": 1562
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3337460920786894e-07,
+      "loss": 0.9999,
+      "step": 1563
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3328778395756984e-07,
+      "loss": 1.0852,
+      "step": 1564
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3320091087918214e-07,
+      "loss": 1.0422,
+      "step": 1565
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3311398999537495e-07,
+      "loss": 0.9064,
+      "step": 1566
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3302702132883005e-07,
+      "loss": 1.0381,
+      "step": 1567
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3294000490224126e-07,
+      "loss": 0.9348,
+      "step": 1568
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3285294073831525e-07,
+      "loss": 0.961,
+      "step": 1569
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.32765828859771e-07,
+      "loss": 0.9779,
+      "step": 1570
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.326786692893398e-07,
+      "loss": 1.0179,
+      "step": 1571
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.325914620497658e-07,
+      "loss": 1.0597,
+      "step": 1572
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.325042071638051e-07,
+      "loss": 0.9503,
+      "step": 1573
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.324169046542265e-07,
+      "loss": 1.2251,
+      "step": 1574
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.323295545438112e-07,
+      "loss": 0.9886,
+      "step": 1575
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3224215685535287e-07,
+      "loss": 0.9176,
+      "step": 1576
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3215471161165737e-07,
+      "loss": 1.1978,
+      "step": 1577
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.320672188355433e-07,
+      "loss": 0.9766,
+      "step": 1578
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3197967854984146e-07,
+      "loss": 1.078,
+      "step": 1579
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.31892090777395e-07,
+      "loss": 1.1804,
+      "step": 1580
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.318044555410595e-07,
+      "loss": 0.9494,
+      "step": 1581
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3171677286370313e-07,
+      "loss": 1.0092,
+      "step": 1582
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3162904276820613e-07,
+      "loss": 1.0828,
+      "step": 1583
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3154126527746125e-07,
+      "loss": 0.8743,
+      "step": 1584
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.314534404143737e-07,
+      "loss": 1.1607,
+      "step": 1585
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3136556820186097e-07,
+      "loss": 1.1214,
+      "step": 1586
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.312776486628528e-07,
+      "loss": 1.1978,
+      "step": 1587
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3118968182029153e-07,
+      "loss": 0.9313,
+      "step": 1588
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3110166769713157e-07,
+      "loss": 0.8927,
+      "step": 1589
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3101360631633986e-07,
+      "loss": 0.9477,
+      "step": 1590
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.309254977008956e-07,
+      "loss": 0.9189,
+      "step": 1591
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.308373418737903e-07,
+      "loss": 0.9359,
+      "step": 1592
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3074913885802776e-07,
+      "loss": 1.0495,
+      "step": 1593
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.306608886766242e-07,
+      "loss": 1.1095,
+      "step": 1594
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3057259135260817e-07,
+      "loss": 1.1069,
+      "step": 1595
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.304842469090203e-07,
+      "loss": 1.1066,
+      "step": 1596
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.303958553689137e-07,
+      "loss": 1.019,
+      "step": 1597
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3030741675535374e-07,
+      "loss": 0.9697,
+      "step": 1598
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3021893109141813e-07,
+      "loss": 0.8947,
+      "step": 1599
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.301303984001967e-07,
+      "loss": 1.0401,
+      "step": 1600
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3004181870479175e-07,
+      "loss": 1.1012,
+      "step": 1601
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.299531920283176e-07,
+      "loss": 1.0236,
+      "step": 1602
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2986451839390096e-07,
+      "loss": 0.8838,
+      "step": 1603
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2977579782468094e-07,
+      "loss": 0.9933,
+      "step": 1604
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2968703034380867e-07,
+      "loss": 0.949,
+      "step": 1605
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.295982159744476e-07,
+      "loss": 1.1169,
+      "step": 1606
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2950935473977345e-07,
+      "loss": 1.0256,
+      "step": 1607
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2942044666297414e-07,
+      "loss": 0.9275,
+      "step": 1608
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.293314917672498e-07,
+      "loss": 1.0386,
+      "step": 1609
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.292424900758128e-07,
+      "loss": 1.0308,
+      "step": 1610
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.291534416118877e-07,
+      "loss": 1.0893,
+      "step": 1611
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2906434639871137e-07,
+      "loss": 1.0935,
+      "step": 1612
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.289752044595326e-07,
+      "loss": 1.0513,
+      "step": 1613
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.288860158176128e-07,
+      "loss": 0.9997,
+      "step": 1614
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2879678049622514e-07,
+      "loss": 1.041,
+      "step": 1615
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.287074985186553e-07,
+      "loss": 0.9728,
+      "step": 1616
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.286181699082008e-07,
+      "loss": 1.1874,
+      "step": 1617
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.285287946881717e-07,
+      "loss": 1.0447,
+      "step": 1618
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2843937288189004e-07,
+      "loss": 1.0156,
+      "step": 1619
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2834990451268995e-07,
+      "loss": 0.915,
+      "step": 1620
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2826038960391774e-07,
+      "loss": 1.0355,
+      "step": 1621
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2817082817893205e-07,
+      "loss": 0.9023,
+      "step": 1622
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.280812202611033e-07,
+      "loss": 0.9963,
+      "step": 1623
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2799156587381446e-07,
+      "loss": 1.239,
+      "step": 1624
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.279018650404604e-07,
+      "loss": 1.0152,
+      "step": 1625
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2781211778444803e-07,
+      "loss": 1.0594,
+      "step": 1626
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.277223241291965e-07,
+      "loss": 1.0702,
+      "step": 1627
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.276324840981371e-07,
+      "loss": 0.9517,
+      "step": 1628
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2754259771471314e-07,
+      "loss": 1.0093,
+      "step": 1629
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2745266500238007e-07,
+      "loss": 0.9473,
+      "step": 1630
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.273626859846053e-07,
+      "loss": 0.9645,
+      "step": 1631
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.272726606848686e-07,
+      "loss": 0.8759,
+      "step": 1632
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.271825891266616e-07,
+      "loss": 0.9626,
+      "step": 1633
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2709247133348795e-07,
+      "loss": 1.0178,
+      "step": 1634
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2700230732886356e-07,
+      "loss": 1.144,
+      "step": 1635
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2691209713631633e-07,
+      "loss": 1.1257,
+      "step": 1636
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.268218407793861e-07,
+      "loss": 0.9956,
+      "step": 1637
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.267315382816249e-07,
+      "loss": 1.046,
+      "step": 1638
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.266411896665967e-07,
+      "loss": 0.9792,
+      "step": 1639
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2655079495787747e-07,
+      "loss": 1.071,
+      "step": 1640
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2646035417905555e-07,
+      "loss": 1.0127,
+      "step": 1641
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.263698673537308e-07,
+      "loss": 1.026,
+      "step": 1642
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2627933450551545e-07,
+      "loss": 1.1134,
+      "step": 1643
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.261887556580336e-07,
+      "loss": 0.9046,
+      "step": 1644
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.260981308349213e-07,
+      "loss": 1.0125,
+      "step": 1645
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2600746005982677e-07,
+      "loss": 1.0167,
+      "step": 1646
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2591674335641016e-07,
+      "loss": 0.9081,
+      "step": 1647
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2582598074834336e-07,
+      "loss": 1.0558,
+      "step": 1648
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.257351722593106e-07,
+      "loss": 1.0389,
+      "step": 1649
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.25644317913008e-07,
+      "loss": 1.1108,
+      "step": 1650
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.255534177331434e-07,
+      "loss": 1.0314,
+      "step": 1651
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.254624717434369e-07,
+      "loss": 1.0146,
+      "step": 1652
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.253714799676204e-07,
+      "loss": 0.9043,
+      "step": 1653
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.252804424294378e-07,
+      "loss": 1.0644,
+      "step": 1654
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.251893591526449e-07,
+      "loss": 0.9507,
+      "step": 1655
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.250982301610093e-07,
+      "loss": 1.0513,
+      "step": 1656
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2500705547831096e-07,
+      "loss": 0.9714,
+      "step": 1657
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.249158351283413e-07,
+      "loss": 1.0142,
+      "step": 1658
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2482456913490394e-07,
+      "loss": 1.0076,
+      "step": 1659
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.247332575218143e-07,
+      "loss": 1.1034,
+      "step": 1660
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.246419003128996e-07,
+      "loss": 1.154,
+      "step": 1661
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2455049753199924e-07,
+      "loss": 1.0988,
+      "step": 1662
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2445904920296426e-07,
+      "loss": 0.9142,
+      "step": 1663
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2436755534965764e-07,
+      "loss": 1.0991,
+      "step": 1664
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2427601599595434e-07,
+      "loss": 0.9807,
+      "step": 1665
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.241844311657411e-07,
+      "loss": 1.1606,
+      "step": 1666
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.240928008829166e-07,
+      "loss": 1.0493,
+      "step": 1667
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.240011251713913e-07,
+      "loss": 0.9931,
+      "step": 1668
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2390940405508747e-07,
+      "loss": 0.883,
+      "step": 1669
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.238176375579394e-07,
+      "loss": 1.0975,
+      "step": 1670
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.237258257038931e-07,
+      "loss": 1.1637,
+      "step": 1671
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.236339685169064e-07,
+      "loss": 1.07,
+      "step": 1672
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.235420660209491e-07,
+      "loss": 0.8263,
+      "step": 1673
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2345011824000266e-07,
+      "loss": 1.0798,
+      "step": 1674
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.233581251980604e-07,
+      "loss": 0.9701,
+      "step": 1675
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2326608691912757e-07,
+      "loss": 0.992,
+      "step": 1676
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.23174003427221e-07,
+      "loss": 1.1032,
+      "step": 1677
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2308187474636956e-07,
+      "loss": 1.0894,
+      "step": 1678
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.229897009006137e-07,
+      "loss": 1.0009,
+      "step": 1679
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2289748191400586e-07,
+      "loss": 0.9475,
+      "step": 1680
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.228052178106101e-07,
+      "loss": 0.9391,
+      "step": 1681
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.227129086145022e-07,
+      "loss": 1.2183,
+      "step": 1682
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2262055434977003e-07,
+      "loss": 1.232,
+      "step": 1683
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.225281550405128e-07,
+      "loss": 1.1146,
+      "step": 1684
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2243571071084177e-07,
+      "loss": 0.8326,
+      "step": 1685
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2234322138487985e-07,
+      "loss": 1.1713,
+      "step": 1686
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.222506870867617e-07,
+      "loss": 1.0216,
+      "step": 1687
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2215810784063375e-07,
+      "loss": 1.0802,
+      "step": 1688
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2206548367065405e-07,
+      "loss": 1.0719,
+      "step": 1689
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2197281460099243e-07,
+      "loss": 1.0437,
+      "step": 1690
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.218801006558305e-07,
+      "loss": 0.8959,
+      "step": 1691
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.217873418593616e-07,
+      "loss": 1.1106,
+      "step": 1692
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.216945382357905e-07,
+      "loss": 0.9374,
+      "step": 1693
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.21601689809334e-07,
+      "loss": 1.0191,
+      "step": 1694
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2150879660422057e-07,
+      "loss": 1.1973,
+      "step": 1695
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2141585864469004e-07,
+      "loss": 0.9928,
+      "step": 1696
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2132287595499425e-07,
+      "loss": 1.0908,
+      "step": 1697
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.212298485593966e-07,
+      "loss": 0.9237,
+      "step": 1698
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2113677648217216e-07,
+      "loss": 1.0788,
+      "step": 1699
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.210436597476076e-07,
+      "loss": 0.9444,
+      "step": 1700
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.209504983800013e-07,
+      "loss": 0.982,
+      "step": 1701
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.208572924036633e-07,
+      "loss": 1.0623,
+      "step": 1702
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2076404184291537e-07,
+      "loss": 0.929,
+      "step": 1703
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.206707467220906e-07,
+      "loss": 1.0612,
+      "step": 1704
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.205774070655341e-07,
+      "loss": 0.9759,
+      "step": 1705
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2048402289760224e-07,
+      "loss": 0.9341,
+      "step": 1706
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.203905942426633e-07,
+      "loss": 1.0712,
+      "step": 1707
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.202971211250971e-07,
+      "loss": 0.9866,
+      "step": 1708
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2020360356929493e-07,
+      "loss": 0.8197,
+      "step": 1709
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.201100415996597e-07,
+      "loss": 1.1769,
+      "step": 1710
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.20016435240606e-07,
+      "loss": 0.8816,
+      "step": 1711
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1992278451656014e-07,
+      "loss": 0.8265,
+      "step": 1712
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1982908945195963e-07,
+      "loss": 1.2318,
+      "step": 1713
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1973535007125383e-07,
+      "loss": 0.822,
+      "step": 1714
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1964156639890364e-07,
+      "loss": 0.9559,
+      "step": 1715
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.195477384593814e-07,
+      "loss": 1.1636,
+      "step": 1716
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1945386627717105e-07,
+      "loss": 1.1711,
+      "step": 1717
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1935994987676824e-07,
+      "loss": 1.0018,
+      "step": 1718
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1926598928267986e-07,
+      "loss": 0.934,
+      "step": 1719
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.191719845194245e-07,
+      "loss": 1.1243,
+      "step": 1720
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1907793561153237e-07,
+      "loss": 1.2419,
+      "step": 1721
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.18983842583545e-07,
+      "loss": 1.0475,
+      "step": 1722
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1888970546001555e-07,
+      "loss": 1.0328,
+      "step": 1723
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.187955242655087e-07,
+      "loss": 0.9202,
+      "step": 1724
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.187012990246005e-07,
+      "loss": 0.9938,
+      "step": 1725
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.186070297618787e-07,
+      "loss": 0.9986,
+      "step": 1726
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.185127165019423e-07,
+      "loss": 0.9678,
+      "step": 1727
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.18418359269402e-07,
+      "loss": 0.9871,
+      "step": 1728
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1832395808887987e-07,
+      "loss": 1.0378,
+      "step": 1729
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.182295129850094e-07,
+      "loss": 0.922,
+      "step": 1730
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.181350239824356e-07,
+      "loss": 1.0941,
+      "step": 1731
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1804049110581494e-07,
+      "loss": 0.989,
+      "step": 1732
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.179459143798155e-07,
+      "loss": 0.9918,
+      "step": 1733
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.178512938291163e-07,
+      "loss": 1.3242,
+      "step": 1734
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1775662947840845e-07,
+      "loss": 1.038,
+      "step": 1735
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1766192135239395e-07,
+      "loss": 1.1995,
+      "step": 1736
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.175671694757866e-07,
+      "loss": 1.1068,
+      "step": 1737
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.174723738733114e-07,
+      "loss": 0.9713,
+      "step": 1738
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.173775345697047e-07,
+      "loss": 1.1445,
+      "step": 1739
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.172826515897145e-07,
+      "loss": 0.9696,
+      "step": 1740
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.171877249581001e-07,
+      "loss": 1.0291,
+      "step": 1741
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1709275469963214e-07,
+      "loss": 0.9778,
+      "step": 1742
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.169977408390925e-07,
+      "loss": 1.1107,
+      "step": 1743
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.169026834012748e-07,
+      "loss": 1.0642,
+      "step": 1744
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.168075824109837e-07,
+      "loss": 1.168,
+      "step": 1745
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1671243789303544e-07,
+      "loss": 0.9229,
+      "step": 1746
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.166172498722576e-07,
+      "loss": 1.0571,
+      "step": 1747
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.165220183734889e-07,
+      "loss": 0.9206,
+      "step": 1748
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1642674342157965e-07,
+      "loss": 0.9678,
+      "step": 1749
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.163314250413913e-07,
+      "loss": 0.985,
+      "step": 1750
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.162360632577969e-07,
+      "loss": 0.9922,
+      "step": 1751
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.161406580956805e-07,
+      "loss": 1.0545,
+      "step": 1752
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1604520957993774e-07,
+      "loss": 1.0397,
+      "step": 1753
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1594971773547537e-07,
+      "loss": 0.9531,
+      "step": 1754
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.158541825872116e-07,
+      "loss": 1.1262,
+      "step": 1755
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1575860416007586e-07,
+      "loss": 0.9567,
+      "step": 1756
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.156629824790089e-07,
+      "loss": 1.023,
+      "step": 1757
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.155673175689628e-07,
+      "loss": 1.1514,
+      "step": 1758
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.154716094549008e-07,
+      "loss": 1.1796,
+      "step": 1759
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1537585816179744e-07,
+      "loss": 1.1728,
+      "step": 1760
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.152800637146387e-07,
+      "loss": 0.9014,
+      "step": 1761
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.151842261384216e-07,
+      "loss": 1.0288,
+      "step": 1762
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.150883454581546e-07,
+      "loss": 1.1462,
+      "step": 1763
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1499242169885717e-07,
+      "loss": 1.0527,
+      "step": 1764
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1489645488556023e-07,
+      "loss": 1.0495,
+      "step": 1765
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1480044504330594e-07,
+      "loss": 0.995,
+      "step": 1766
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.147043921971475e-07,
+      "loss": 0.9637,
+      "step": 1767
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.146082963721496e-07,
+      "loss": 1.0381,
+      "step": 1768
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.145121575933879e-07,
+      "loss": 1.016,
+      "step": 1769
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1441597588594933e-07,
+      "loss": 1.0,
+      "step": 1770
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1431975127493214e-07,
+      "loss": 1.0031,
+      "step": 1771
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.142234837854456e-07,
+      "loss": 0.9411,
+      "step": 1772
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1412717344261027e-07,
+      "loss": 1.0956,
+      "step": 1773
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.14030820271558e-07,
+      "loss": 0.9243,
+      "step": 1774
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1393442429743166e-07,
+      "loss": 1.1257,
+      "step": 1775
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.138379855453852e-07,
+      "loss": 1.1804,
+      "step": 1776
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1374150404058394e-07,
+      "loss": 1.1336,
+      "step": 1777
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.136449798082043e-07,
+      "loss": 1.2582,
+      "step": 1778
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.135484128734339e-07,
+      "loss": 0.9934,
+      "step": 1779
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1345180326147125e-07,
+      "loss": 1.0884,
+      "step": 1780
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.133551509975264e-07,
+      "loss": 0.9538,
+      "step": 1781
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1325845610682e-07,
+      "loss": 0.9908,
+      "step": 1782
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1316171861458436e-07,
+      "loss": 1.0447,
+      "step": 1783
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.130649385460627e-07,
+      "loss": 1.0822,
+      "step": 1784
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1296811592650916e-07,
+      "loss": 0.9994,
+      "step": 1785
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.128712507811892e-07,
+      "loss": 0.9944,
+      "step": 1786
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.127743431353795e-07,
+      "loss": 1.0746,
+      "step": 1787
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.126773930143674e-07,
+      "loss": 1.0944,
+      "step": 1788
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1258040044345163e-07,
+      "loss": 1.051,
+      "step": 1789
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.124833654479422e-07,
+      "loss": 1.1946,
+      "step": 1790
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.123862880531596e-07,
+      "loss": 1.1145,
+      "step": 1791
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1228916828443593e-07,
+      "loss": 0.8966,
+      "step": 1792
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.12192006167114e-07,
+      "loss": 1.0231,
+      "step": 1793
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.12094801726548e-07,
+      "loss": 0.9519,
+      "step": 1794
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1199755498810285e-07,
+      "loss": 1.1067,
+      "step": 1795
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.119002659771546e-07,
+      "loss": 0.9846,
+      "step": 1796
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.118029347190904e-07,
+      "loss": 1.0484,
+      "step": 1797
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1170556123930837e-07,
+      "loss": 1.1231,
+      "step": 1798
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.116081455632178e-07,
+      "loss": 1.0261,
+      "step": 1799
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.115106877162386e-07,
+      "loss": 1.0899,
+      "step": 1800
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1141318772380206e-07,
+      "loss": 1.049,
+      "step": 1801
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1131564561135046e-07,
+      "loss": 0.9915,
+      "step": 1802
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1121806140433685e-07,
+      "loss": 1.0526,
+      "step": 1803
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1112043512822536e-07,
+      "loss": 1.1849,
+      "step": 1804
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.110227668084911e-07,
+      "loss": 0.9696,
+      "step": 1805
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1092505647062024e-07,
+      "loss": 1.0182,
+      "step": 1806
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.108273041401097e-07,
+      "loss": 1.0552,
+      "step": 1807
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.107295098424677e-07,
+      "loss": 0.9297,
+      "step": 1808
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1063167360321303e-07,
+      "loss": 1.0488,
+      "step": 1809
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1053379544787557e-07,
+      "loss": 0.9737,
+      "step": 1810
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.104358754019964e-07,
+      "loss": 0.925,
+      "step": 1811
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1033791349112713e-07,
+      "loss": 1.0384,
+      "step": 1812
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1023990974083036e-07,
+      "loss": 0.9531,
+      "step": 1813
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.101418641766799e-07,
+      "loss": 1.0748,
+      "step": 1814
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1004377682426017e-07,
+      "loss": 0.8999,
+      "step": 1815
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.099456477091666e-07,
+      "loss": 1.0182,
+      "step": 1816
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0984747685700563e-07,
+      "loss": 1.036,
+      "step": 1817
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0974926429339443e-07,
+      "loss": 1.171,
+      "step": 1818
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.096510100439611e-07,
+      "loss": 1.0842,
+      "step": 1819
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0955271413434457e-07,
+      "loss": 0.9168,
+      "step": 1820
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0945437659019477e-07,
+      "loss": 0.9205,
+      "step": 1821
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0935599743717244e-07,
+      "loss": 1.1106,
+      "step": 1822
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.092575767009491e-07,
+      "loss": 1.041,
+      "step": 1823
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0915911440720713e-07,
+      "loss": 1.0849,
+      "step": 1824
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.090606105816399e-07,
+      "loss": 1.0257,
+      "step": 1825
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.089620652499516e-07,
+      "loss": 0.9648,
+      "step": 1826
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0886347843785694e-07,
+      "loss": 0.978,
+      "step": 1827
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0876485017108184e-07,
+      "loss": 1.0212,
+      "step": 1828
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0866618047536297e-07,
+      "loss": 0.9105,
+      "step": 1829
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.085674693764475e-07,
+      "loss": 1.1455,
+      "step": 1830
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0846871690009374e-07,
+      "loss": 0.9449,
+      "step": 1831
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0836992307207073e-07,
+      "loss": 0.9651,
+      "step": 1832
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.082710879181582e-07,
+      "loss": 0.9755,
+      "step": 1833
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.081722114641468e-07,
+      "loss": 1.015,
+      "step": 1834
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0807329373583775e-07,
+      "loss": 0.9166,
+      "step": 1835
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0797433475904324e-07,
+      "loss": 0.9668,
+      "step": 1836
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.078753345595862e-07,
+      "loss": 1.0517,
+      "step": 1837
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0777629316330013e-07,
+      "loss": 0.9708,
+      "step": 1838
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0767721059602956e-07,
+      "loss": 1.0337,
+      "step": 1839
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.075780868836295e-07,
+      "loss": 1.0268,
+      "step": 1840
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.074789220519659e-07,
+      "loss": 0.9745,
+      "step": 1841
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0737971612691535e-07,
+      "loss": 1.1461,
+      "step": 1842
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.072804691343652e-07,
+      "loss": 0.9793,
+      "step": 1843
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.071811811002134e-07,
+      "loss": 0.985,
+      "step": 1844
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.070818520503687e-07,
+      "loss": 0.9813,
+      "step": 1845
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.069824820107507e-07,
+      "loss": 1.0077,
+      "step": 1846
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.068830710072894e-07,
+      "loss": 0.9298,
+      "step": 1847
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.067836190659256e-07,
+      "loss": 1.0097,
+      "step": 1848
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.06684126212611e-07,
+      "loss": 1.0297,
+      "step": 1849
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.065845924733076e-07,
+      "loss": 1.0852,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.064850178739884e-07,
+      "loss": 1.1347,
+      "step": 1851
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.063854024406369e-07,
+      "loss": 0.9522,
+      "step": 1852
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.062857461992473e-07,
+      "loss": 0.8624,
+      "step": 1853
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0618604917582427e-07,
+      "loss": 0.9785,
+      "step": 1854
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.060863113963835e-07,
+      "loss": 1.0436,
+      "step": 1855
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.059865328869509e-07,
+      "loss": 0.9594,
+      "step": 1856
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0588671367356333e-07,
+      "loss": 0.9425,
+      "step": 1857
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.057868537822682e-07,
+      "loss": 1.0398,
+      "step": 1858
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0568695323912326e-07,
+      "loss": 0.8824,
+      "step": 1859
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.055870120701973e-07,
+      "loss": 0.8795,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0548703030156944e-07,
+      "loss": 1.2383,
+      "step": 1861
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0538700795932935e-07,
+      "loss": 1.0556,
+      "step": 1862
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0528694506957754e-07,
+      "loss": 1.0113,
+      "step": 1863
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.05186841658425e-07,
+      "loss": 1.0566,
+      "step": 1864
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.05086697751993e-07,
+      "loss": 1.1633,
+      "step": 1865
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0498651337641374e-07,
+      "loss": 0.8768,
+      "step": 1866
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0488628855783e-07,
+      "loss": 0.8479,
+      "step": 1867
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0478602332239476e-07,
+      "loss": 1.0443,
+      "step": 1868
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.046857176962719e-07,
+      "loss": 0.9484,
+      "step": 1869
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0458537170563577e-07,
+      "loss": 0.95,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0448498537667096e-07,
+      "loss": 0.9177,
+      "step": 1871
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.04384558735573e-07,
+      "loss": 1.1389,
+      "step": 1872
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0428409180854767e-07,
+      "loss": 1.1359,
+      "step": 1873
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.041835846218113e-07,
+      "loss": 0.9967,
+      "step": 1874
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.040830372015909e-07,
+      "loss": 0.9458,
+      "step": 1875
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0398244957412373e-07,
+      "loss": 0.909,
+      "step": 1876
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.038818217656578e-07,
+      "loss": 0.9826,
+      "step": 1877
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0378115380245136e-07,
+      "loss": 1.1472,
+      "step": 1878
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.036804457107732e-07,
+      "loss": 0.9662,
+      "step": 1879
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0357969751690273e-07,
+      "loss": 1.0078,
+      "step": 1880
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0347890924712967e-07,
+      "loss": 1.0092,
+      "step": 1881
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0337808092775426e-07,
+      "loss": 1.1003,
+      "step": 1882
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.032772125850871e-07,
+      "loss": 1.0436,
+      "step": 1883
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.031763042454494e-07,
+      "loss": 1.0231,
+      "step": 1884
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.030753559351728e-07,
+      "loss": 1.0928,
+      "step": 1885
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0297436768059905e-07,
+      "loss": 0.9113,
+      "step": 1886
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.028733395080808e-07,
+      "loss": 0.9244,
+      "step": 1887
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.027722714439807e-07,
+      "loss": 1.0103,
+      "step": 1888
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.026711635146721e-07,
+      "loss": 0.9323,
+      "step": 1889
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.025700157465386e-07,
+      "loss": 1.0118,
+      "step": 1890
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0246882816597427e-07,
+      "loss": 1.1563,
+      "step": 1891
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.023676007993835e-07,
+      "loss": 1.0611,
+      "step": 1892
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.022663336731811e-07,
+      "loss": 1.0759,
+      "step": 1893
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0216502681379227e-07,
+      "loss": 0.9191,
+      "step": 1894
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0206368024765247e-07,
+      "loss": 1.0495,
+      "step": 1895
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0196229400120775e-07,
+      "loss": 1.2147,
+      "step": 1896
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.018608681009143e-07,
+      "loss": 1.0888,
+      "step": 1897
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.017594025732387e-07,
+      "loss": 1.0035,
+      "step": 1898
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0165789744465805e-07,
+      "loss": 1.0058,
+      "step": 1899
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.015563527416595e-07,
+      "loss": 0.8925,
+      "step": 1900
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.014547684907407e-07,
+      "loss": 0.9675,
+      "step": 1901
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.013531447184095e-07,
+      "loss": 1.0923,
+      "step": 1902
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.012514814511844e-07,
+      "loss": 0.9736,
+      "step": 1903
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0114977871559377e-07,
+      "loss": 1.0689,
+      "step": 1904
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0104803653817634e-07,
+      "loss": 1.0117,
+      "step": 1905
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.009462549454815e-07,
+      "loss": 1.0349,
+      "step": 1906
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 4.008444339640685e-07,
+      "loss": 0.9907,
+      "step": 1907
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 4.007425736205072e-07,
+      "loss": 0.924,
+      "step": 1908
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0064067394137746e-07,
+      "loss": 1.0129,
+      "step": 1909
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0053873495326964e-07,
+      "loss": 1.0589,
+      "step": 1910
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 4.004367566827841e-07,
+      "loss": 0.9885,
+      "step": 1911
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 4.003347391565317e-07,
+      "loss": 0.9579,
+      "step": 1912
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0023268240113333e-07,
+      "loss": 0.9871,
+      "step": 1913
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0013058644322034e-07,
+      "loss": 1.1376,
+      "step": 1914
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0002845130943417e-07,
+      "loss": 1.0873,
+      "step": 1915
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9992627702642637e-07,
+      "loss": 0.9579,
+      "step": 1916
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.99824063620859e-07,
+      "loss": 0.9237,
+      "step": 1917
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9972181111940416e-07,
+      "loss": 1.0238,
+      "step": 1918
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.996195195487441e-07,
+      "loss": 0.8724,
+      "step": 1919
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9951718893557127e-07,
+      "loss": 0.9937,
+      "step": 1920
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9941481930658856e-07,
+      "loss": 0.8651,
+      "step": 1921
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.993124106885086e-07,
+      "loss": 1.0824,
+      "step": 1922
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.992099631080547e-07,
+      "loss": 0.9691,
+      "step": 1923
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9910747659195976e-07,
+      "loss": 1.0415,
+      "step": 1924
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.990049511669674e-07,
+      "loss": 1.0262,
+      "step": 1925
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.989023868598311e-07,
+      "loss": 1.1069,
+      "step": 1926
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.987997836973146e-07,
+      "loss": 1.0937,
+      "step": 1927
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9869714170619153e-07,
+      "loss": 0.9892,
+      "step": 1928
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.985944609132459e-07,
+      "loss": 1.0597,
+      "step": 1929
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.98491741345272e-07,
+      "loss": 1.0309,
+      "step": 1930
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9838898302907364e-07,
+      "loss": 1.1059,
+      "step": 1931
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.982861859914654e-07,
+      "loss": 1.0763,
+      "step": 1932
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.981833502592716e-07,
+      "loss": 1.1263,
+      "step": 1933
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.980804758593268e-07,
+      "loss": 1.0076,
+      "step": 1934
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.979775628184755e-07,
+      "loss": 1.0425,
+      "step": 1935
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9787461116357244e-07,
+      "loss": 1.1295,
+      "step": 1936
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.977716209214823e-07,
+      "loss": 0.9575,
+      "step": 1937
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9766859211908e-07,
+      "loss": 0.975,
+      "step": 1938
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.975655247832504e-07,
+      "loss": 0.8913,
+      "step": 1939
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.974624189408884e-07,
+      "loss": 1.0949,
+      "step": 1940
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9735927461889894e-07,
+      "loss": 1.0485,
+      "step": 1941
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.972560918441972e-07,
+      "loss": 1.1676,
+      "step": 1942
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9715287064370805e-07,
+      "loss": 1.2141,
+      "step": 1943
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.970496110443667e-07,
+      "loss": 1.1018,
+      "step": 1944
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9694631307311825e-07,
+      "loss": 1.0061,
+      "step": 1945
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.968429767569178e-07,
+      "loss": 0.9748,
+      "step": 1946
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.967396021227305e-07,
+      "loss": 0.993,
+      "step": 1947
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.966361891975315e-07,
+      "loss": 0.9524,
+      "step": 1948
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.965327380083059e-07,
+      "loss": 0.9675,
+      "step": 1949
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.964292485820487e-07,
+      "loss": 0.9584,
+      "step": 1950
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.963257209457651e-07,
+      "loss": 0.8954,
+      "step": 1951
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9622215512647014e-07,
+      "loss": 1.0049,
+      "step": 1952
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.961185511511888e-07,
+      "loss": 1.0672,
+      "step": 1953
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.96014909046956e-07,
+      "loss": 1.0197,
+      "step": 1954
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9591122884081674e-07,
+      "loss": 1.069,
+      "step": 1955
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9580751055982595e-07,
+      "loss": 0.9997,
+      "step": 1956
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.957037542310483e-07,
+      "loss": 0.9389,
+      "step": 1957
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9559995988155857e-07,
+      "loss": 1.1092,
+      "step": 1958
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.954961275384414e-07,
+      "loss": 0.9683,
+      "step": 1959
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.953922572287914e-07,
+      "loss": 1.1025,
+      "step": 1960
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9528834897971296e-07,
+      "loss": 0.9756,
+      "step": 1961
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9518440281832054e-07,
+      "loss": 0.9286,
+      "step": 1962
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9508041877173833e-07,
+      "loss": 1.1433,
+      "step": 1963
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9497639686710044e-07,
+      "loss": 0.9751,
+      "step": 1964
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.948723371315511e-07,
+      "loss": 1.0388,
+      "step": 1965
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.947682395922439e-07,
+      "loss": 1.0278,
+      "step": 1966
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.946641042763429e-07,
+      "loss": 1.0192,
+      "step": 1967
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.945599312110216e-07,
+      "loss": 1.1767,
+      "step": 1968
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.944557204234634e-07,
+      "loss": 0.9582,
+      "step": 1969
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.943514719408618e-07,
+      "loss": 1.1457,
+      "step": 1970
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9424718579041984e-07,
+      "loss": 1.0548,
+      "step": 1971
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9414286199935044e-07,
+      "loss": 1.2001,
+      "step": 1972
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9403850059487644e-07,
+      "loss": 0.9602,
+      "step": 1973
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9393410160423056e-07,
+      "loss": 0.9753,
+      "step": 1974
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9382966505465515e-07,
+      "loss": 1.0591,
+      "step": 1975
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.937251909734024e-07,
+      "loss": 0.9189,
+      "step": 1976
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9362067938773444e-07,
+      "loss": 0.9956,
+      "step": 1977
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9351613032492304e-07,
+      "loss": 1.0666,
+      "step": 1978
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.934115438122497e-07,
+      "loss": 0.9785,
+      "step": 1979
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.933069198770059e-07,
+      "loss": 1.009,
+      "step": 1980
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9320225854649267e-07,
+      "loss": 1.0372,
+      "step": 1981
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9309755984802094e-07,
+      "loss": 0.9997,
+      "step": 1982
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9299282380891143e-07,
+      "loss": 0.9167,
+      "step": 1983
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.928880504564943e-07,
+      "loss": 0.9713,
+      "step": 1984
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.927832398181099e-07,
+      "loss": 0.9808,
+      "step": 1985
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9267839192110797e-07,
+      "loss": 1.0155,
+      "step": 1986
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9257350679284797e-07,
+      "loss": 1.0004,
+      "step": 1987
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.924685844606994e-07,
+      "loss": 0.9807,
+      "step": 1988
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.923636249520411e-07,
+      "loss": 1.192,
+      "step": 1989
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9225862829426183e-07,
+      "loss": 0.9885,
+      "step": 1990
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9215359451475986e-07,
+      "loss": 0.9098,
+      "step": 1991
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9204852364094343e-07,
+      "loss": 1.1217,
+      "step": 1992
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.919434157002303e-07,
+      "loss": 1.0958,
+      "step": 1993
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9183827072004764e-07,
+      "loss": 1.0446,
+      "step": 1994
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.917330887278328e-07,
+      "loss": 1.1178,
+      "step": 1995
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.916278697510325e-07,
+      "loss": 1.0462,
+      "step": 1996
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9152261381710294e-07,
+      "loss": 1.0335,
+      "step": 1997
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9141732095351045e-07,
+      "loss": 1.0456,
+      "step": 1998
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.913119911877304e-07,
+      "loss": 1.0316,
+      "step": 1999
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9120662454724835e-07,
+      "loss": 0.973,
+      "step": 2000
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.911012210595591e-07,
+      "loss": 0.9544,
+      "step": 2001
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.909957807521673e-07,
+      "loss": 1.0346,
+      "step": 2002
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.90890303652587e-07,
+      "loss": 0.9517,
+      "step": 2003
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9078478978834205e-07,
+      "loss": 0.979,
+      "step": 2004
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.906792391869657e-07,
+      "loss": 0.9095,
+      "step": 2005
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.905736518760009e-07,
+      "loss": 1.0209,
+      "step": 2006
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9046802788300026e-07,
+      "loss": 1.1142,
+      "step": 2007
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9036236723552574e-07,
+      "loss": 1.046,
+      "step": 2008
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9025666996114903e-07,
+      "loss": 1.0143,
+      "step": 2009
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9015093608745143e-07,
+      "loss": 0.9384,
+      "step": 2010
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.900451656420236e-07,
+      "loss": 1.2166,
+      "step": 2011
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8993935865246584e-07,
+      "loss": 0.7629,
+      "step": 2012
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.898335151463881e-07,
+      "loss": 1.1286,
+      "step": 2013
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.897276351514096e-07,
+      "loss": 1.1696,
+      "step": 2014
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.896217186951593e-07,
+      "loss": 0.9494,
+      "step": 2015
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.895157658052756e-07,
+      "loss": 1.1469,
+      "step": 2016
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8940977650940643e-07,
+      "loss": 0.9944,
+      "step": 2017
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.893037508352092e-07,
+      "loss": 1.0105,
+      "step": 2018
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8919768881035075e-07,
+      "loss": 0.9878,
+      "step": 2019
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.890915904625075e-07,
+      "loss": 1.02,
+      "step": 2020
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.889854558193654e-07,
+      "loss": 1.0357,
+      "step": 2021
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.888792849086197e-07,
+      "loss": 1.1051,
+      "step": 2022
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8877307775797505e-07,
+      "loss": 1.0468,
+      "step": 2023
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.88666834395146e-07,
+      "loss": 1.0597,
+      "step": 2024
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.885605548478562e-07,
+      "loss": 1.1678,
+      "step": 2025
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.884542391438386e-07,
+      "loss": 0.8507,
+      "step": 2026
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.88347887310836e-07,
+      "loss": 1.2046,
+      "step": 2027
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8824149937660035e-07,
+      "loss": 1.012,
+      "step": 2028
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.88135075368893e-07,
+      "loss": 1.1103,
+      "step": 2029
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8802861531548485e-07,
+      "loss": 1.0122,
+      "step": 2030
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8792211924415625e-07,
+      "loss": 0.9852,
+      "step": 2031
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.878155871826967e-07,
+      "loss": 0.9964,
+      "step": 2032
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.877090191589053e-07,
+      "loss": 0.982,
+      "step": 2033
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.876024152005905e-07,
+      "loss": 1.0751,
+      "step": 2034
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8749577533557e-07,
+      "loss": 0.9914,
+      "step": 2035
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.873890995916712e-07,
+      "loss": 1.0517,
+      "step": 2036
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.872823879967304e-07,
+      "loss": 1.0355,
+      "step": 2037
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8717564057859366e-07,
+      "loss": 0.9922,
+      "step": 2038
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8706885736511606e-07,
+      "loss": 0.9607,
+      "step": 2039
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.869620383841623e-07,
+      "loss": 0.9928,
+      "step": 2040
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.868551836636063e-07,
+      "loss": 1.0016,
+      "step": 2041
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.867482932313312e-07,
+      "loss": 0.8558,
+      "step": 2042
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8664136711522963e-07,
+      "loss": 1.0772,
+      "step": 2043
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.865344053432034e-07,
+      "loss": 0.9679,
+      "step": 2044
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8642740794316374e-07,
+      "loss": 0.9233,
+      "step": 2045
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.863203749430312e-07,
+      "loss": 1.0555,
+      "step": 2046
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.862133063707353e-07,
+      "loss": 1.1181,
+      "step": 2047
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.861062022542153e-07,
+      "loss": 1.0159,
+      "step": 2048
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8599906262141945e-07,
+      "loss": 0.9892,
+      "step": 2049
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.858918875003053e-07,
+      "loss": 1.0701,
+      "step": 2050
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.857846769188397e-07,
+      "loss": 1.0678,
+      "step": 2051
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8567743090499884e-07,
+      "loss": 1.005,
+      "step": 2052
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.855701494867679e-07,
+      "loss": 1.254,
+      "step": 2053
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8546283269214166e-07,
+      "loss": 0.9077,
+      "step": 2054
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.853554805491238e-07,
+      "loss": 1.0531,
+      "step": 2055
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8524809308572747e-07,
+      "loss": 0.9717,
+      "step": 2056
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.851406703299747e-07,
+      "loss": 1.0497,
+      "step": 2057
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8503321230989723e-07,
+      "loss": 0.9458,
+      "step": 2058
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8492571905353553e-07,
+      "loss": 0.9286,
+      "step": 2059
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.848181905889396e-07,
+      "loss": 1.0844,
+      "step": 2060
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8471062694416847e-07,
+      "loss": 0.9501,
+      "step": 2061
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8460302814729015e-07,
+      "loss": 1.15,
+      "step": 2062
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.844953942263823e-07,
+      "loss": 1.1207,
+      "step": 2063
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8438772520953146e-07,
+      "loss": 0.841,
+      "step": 2064
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8428002112483325e-07,
+      "loss": 1.1459,
+      "step": 2065
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8417228200039254e-07,
+      "loss": 1.0121,
+      "step": 2066
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.840645078643235e-07,
+      "loss": 1.0157,
+      "step": 2067
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.839566987447491e-07,
+      "loss": 1.0212,
+      "step": 2068
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8384885466980174e-07,
+      "loss": 1.0827,
+      "step": 2069
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8374097566762286e-07,
+      "loss": 1.0132,
+      "step": 2070
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.836330617663629e-07,
+      "loss": 1.009,
+      "step": 2071
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8352511299418157e-07,
+      "loss": 0.9872,
+      "step": 2072
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.834171293792474e-07,
+      "loss": 0.9566,
+      "step": 2073
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.833091109497384e-07,
+      "loss": 1.0285,
+      "step": 2074
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8320105773384135e-07,
+      "loss": 0.999,
+      "step": 2075
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.830929697597525e-07,
+      "loss": 1.1555,
+      "step": 2076
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8298484705567644e-07,
+      "loss": 1.0523,
+      "step": 2077
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8287668964982764e-07,
+      "loss": 1.0535,
+      "step": 2078
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.827684975704293e-07,
+      "loss": 1.1429,
+      "step": 2079
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8266027084571336e-07,
+      "loss": 0.9031,
+      "step": 2080
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8255200950392126e-07,
+      "loss": 0.9771,
+      "step": 2081
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.824437135733032e-07,
+      "loss": 0.9773,
+      "step": 2082
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.823353830821186e-07,
+      "loss": 0.9504,
+      "step": 2083
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.822270180586358e-07,
+      "loss": 0.9222,
+      "step": 2084
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.82118618531132e-07,
+      "loss": 1.1183,
+      "step": 2085
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.820101845278937e-07,
+      "loss": 0.9713,
+      "step": 2086
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.819017160772161e-07,
+      "loss": 1.0413,
+      "step": 2087
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8179321320740363e-07,
+      "loss": 1.2601,
+      "step": 2088
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.816846759467696e-07,
+      "loss": 1.153,
+      "step": 2089
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8157610432363637e-07,
+      "loss": 1.0774,
+      "step": 2090
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8146749836633494e-07,
+      "loss": 0.9895,
+      "step": 2091
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8135885810320584e-07,
+      "loss": 0.8318,
+      "step": 2092
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 3.812501835625981e-07,
+      "loss": 1.0066,
+      "step": 2093
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8114147477286985e-07,
+      "loss": 1.0399,
+      "step": 2094
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.81032731762388e-07,
+      "loss": 0.9457,
+      "step": 2095
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8092395455952873e-07,
+      "loss": 1.1976,
+      "step": 2096
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8081514319267686e-07,
+      "loss": 1.0863,
+      "step": 2097
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.807062976902262e-07,
+      "loss": 1.0888,
+      "step": 2098
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.805974180805795e-07,
+      "loss": 1.1686,
+      "step": 2099
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.804885043921484e-07,
+      "loss": 0.8513,
+      "step": 2100
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.803795566533533e-07,
+      "loss": 1.0033,
+      "step": 2101
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.802705748926237e-07,
+      "loss": 1.0079,
+      "step": 2102
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.801615591383979e-07,
+      "loss": 1.0035,
+      "step": 2103
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8005250941912303e-07,
+      "loss": 0.9114,
+      "step": 2104
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7994342576325507e-07,
+      "loss": 1.1311,
+      "step": 2105
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.798343081992589e-07,
+      "loss": 1.0948,
+      "step": 2106
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.797251567556082e-07,
+      "loss": 0.9694,
+      "step": 2107
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.796159714607857e-07,
+      "loss": 1.0201,
+      "step": 2108
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7950675234328256e-07,
+      "loss": 0.916,
+      "step": 2109
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.79397499431599e-07,
+      "loss": 1.003,
+      "step": 2110
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.792882127542444e-07,
+      "loss": 0.9794,
+      "step": 2111
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.791788923397362e-07,
+      "loss": 0.9377,
+      "step": 2112
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7906953821660124e-07,
+      "loss": 1.0243,
+      "step": 2113
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.78960150413375e-07,
+      "loss": 0.9732,
+      "step": 2114
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.788507289586016e-07,
+      "loss": 1.083,
+      "step": 2115
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.787412738808341e-07,
+      "loss": 1.0387,
+      "step": 2116
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.786317852086344e-07,
+      "loss": 1.1116,
+      "step": 2117
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.785222629705729e-07,
+      "loss": 0.9005,
+      "step": 2118
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.784127071952289e-07,
+      "loss": 0.8371,
+      "step": 2119
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7830311791119063e-07,
+      "loss": 1.0851,
+      "step": 2120
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.781934951470547e-07,
+      "loss": 1.0175,
+      "step": 2121
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7808383893142687e-07,
+      "loss": 1.0512,
+      "step": 2122
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7797414929292126e-07,
+      "loss": 0.9784,
+      "step": 2123
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.778644262601609e-07,
+      "loss": 0.8618,
+      "step": 2124
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7775466986177757e-07,
+      "loss": 1.0535,
+      "step": 2125
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.776448801264116e-07,
+      "loss": 0.9874,
+      "step": 2126
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.775350570827122e-07,
+      "loss": 0.9141,
+      "step": 2127
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7742520075933705e-07,
+      "loss": 1.098,
+      "step": 2128
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7731531118495274e-07,
+      "loss": 0.878,
+      "step": 2129
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7720538838823445e-07,
+      "loss": 0.9914,
+      "step": 2130
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.770954323978659e-07,
+      "loss": 0.9928,
+      "step": 2131
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7698544324253976e-07,
+      "loss": 1.1301,
+      "step": 2132
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7687542095095715e-07,
+      "loss": 1.1253,
+      "step": 2133
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7676536555182764e-07,
+      "loss": 1.1778,
+      "step": 2134
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7665527707387e-07,
+      "loss": 0.9158,
+      "step": 2135
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.765451555458111e-07,
+      "loss": 1.1297,
+      "step": 2136
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.764350009963867e-07,
+      "loss": 0.9722,
+      "step": 2137
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.763248134543411e-07,
+      "loss": 1.0383,
+      "step": 2138
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.762145929484272e-07,
+      "loss": 0.8502,
+      "step": 2139
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.761043395074066e-07,
+      "loss": 1.0632,
+      "step": 2140
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.759940531600494e-07,
+      "loss": 0.9844,
+      "step": 2141
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.758837339351342e-07,
+      "loss": 1.0519,
+      "step": 2142
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.757733818614484e-07,
+      "loss": 0.878,
+      "step": 2143
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.756629969677878e-07,
+      "loss": 0.9355,
+      "step": 2144
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7555257928295684e-07,
+      "loss": 0.9819,
+      "step": 2145
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7544212883576854e-07,
+      "loss": 1.0217,
+      "step": 2146
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.753316456550444e-07,
+      "loss": 1.0781,
+      "step": 2147
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7522112976961436e-07,
+      "loss": 1.0204,
+      "step": 2148
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.751105812083172e-07,
+      "loss": 0.9774,
+      "step": 2149
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.75e-07,
+      "loss": 0.9254,
+      "step": 2150
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.748893861735184e-07,
+      "loss": 1.0073,
+      "step": 2151
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.747787397577365e-07,
+      "loss": 1.0165,
+      "step": 2152
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7466806078152703e-07,
+      "loss": 0.9868,
+      "step": 2153
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7455734927377117e-07,
+      "loss": 1.1417,
+      "step": 2154
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7444660526335847e-07,
+      "loss": 0.8319,
+      "step": 2155
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7433582877918726e-07,
+      "loss": 1.2568,
+      "step": 2156
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.742250198501639e-07,
+      "loss": 1.1987,
+      "step": 2157
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7411417850520356e-07,
+      "loss": 1.1487,
+      "step": 2158
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.740033047732298e-07,
+      "loss": 0.8787,
+      "step": 2159
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.738923986831746e-07,
+      "loss": 1.1162,
+      "step": 2160
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.737814602639784e-07,
+      "loss": 1.0893,
+      "step": 2161
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7367048954459e-07,
+      "loss": 1.1242,
+      "step": 2162
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.735594865539667e-07,
+      "loss": 1.0634,
+      "step": 2163
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7344845132107424e-07,
+      "loss": 0.923,
+      "step": 2164
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.733373838748868e-07,
+      "loss": 0.9858,
+      "step": 2165
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.732262842443867e-07,
+      "loss": 1.0108,
+      "step": 2166
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7311515245856504e-07,
+      "loss": 0.9233,
+      "step": 2167
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7300398854642125e-07,
+      "loss": 1.1379,
+      "step": 2168
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.728927925369628e-07,
+      "loss": 1.0435,
+      "step": 2169
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7278156445920577e-07,
+      "loss": 1.0159,
+      "step": 2170
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7267030434217485e-07,
+      "loss": 1.0346,
+      "step": 2171
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7255901221490254e-07,
+      "loss": 1.0428,
+      "step": 2172
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7244768810643023e-07,
+      "loss": 1.1282,
+      "step": 2173
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.723363320458074e-07,
+      "loss": 1.0041,
+      "step": 2174
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.722249440620917e-07,
+      "loss": 0.9314,
+      "step": 2175
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7211352418434956e-07,
+      "loss": 0.8915,
+      "step": 2176
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7200207244165526e-07,
+      "loss": 0.9902,
+      "step": 2177
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.718905888630918e-07,
+      "loss": 1.0227,
+      "step": 2178
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.717790734777501e-07,
+      "loss": 0.9439,
+      "step": 2179
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.716675263147298e-07,
+      "loss": 1.0499,
+      "step": 2180
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7155594740313833e-07,
+      "loss": 0.9849,
+      "step": 2181
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.71444336772092e-07,
+      "loss": 1.0051,
+      "step": 2182
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7133269445071487e-07,
+      "loss": 1.1267,
+      "step": 2183
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.712210204681395e-07,
+      "loss": 1.0699,
+      "step": 2184
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.711093148535067e-07,
+      "loss": 0.9719,
+      "step": 2185
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.709975776359656e-07,
+      "loss": 1.0399,
+      "step": 2186
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.708858088446735e-07,
+      "loss": 1.0394,
+      "step": 2187
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7077400850879587e-07,
+      "loss": 1.1307,
+      "step": 2188
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7066217665750654e-07,
+      "loss": 1.0553,
+      "step": 2189
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.705503133199874e-07,
+      "loss": 0.8117,
+      "step": 2190
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.704384185254288e-07,
+      "loss": 1.0669,
+      "step": 2191
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7032649230302915e-07,
+      "loss": 1.1692,
+      "step": 2192
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7021453468199494e-07,
+      "loss": 1.0086,
+      "step": 2193
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.701025456915411e-07,
+      "loss": 1.181,
+      "step": 2194
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6999052536089066e-07,
+      "loss": 1.1683,
+      "step": 2195
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.698784737192747e-07,
+      "loss": 0.907,
+      "step": 2196
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6976639079593266e-07,
+      "loss": 0.9126,
+      "step": 2197
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6965427662011206e-07,
+      "loss": 0.9527,
+      "step": 2198
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6954213122106847e-07,
+      "loss": 0.9745,
+      "step": 2199
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.694299546280657e-07,
+      "loss": 1.1155,
+      "step": 2200
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6931774687037586e-07,
+      "loss": 1.0485,
+      "step": 2201
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.692055079772789e-07,
+      "loss": 0.9938,
+      "step": 2202
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.690932379780631e-07,
+      "loss": 1.0404,
+      "step": 2203
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6898093690202483e-07,
+      "loss": 1.0624,
+      "step": 2204
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6886860477846834e-07,
+      "loss": 1.0312,
+      "step": 2205
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.687562416367063e-07,
+      "loss": 0.9022,
+      "step": 2206
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.686438475060595e-07,
+      "loss": 0.9367,
+      "step": 2207
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6853142241585644e-07,
+      "loss": 1.1561,
+      "step": 2208
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.684189663954339e-07,
+      "loss": 1.1484,
+      "step": 2209
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6830647947413694e-07,
+      "loss": 1.0091,
+      "step": 2210
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6819396168131835e-07,
+      "loss": 1.1354,
+      "step": 2211
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.680814130463392e-07,
+      "loss": 1.1033,
+      "step": 2212
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.679688335985685e-07,
+      "loss": 1.0123,
+      "step": 2213
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.678562233673833e-07,
+      "loss": 0.9952,
+      "step": 2214
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.677435823821687e-07,
+      "loss": 0.9738,
+      "step": 2215
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.67630910672318e-07,
+      "loss": 0.9788,
+      "step": 2216
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6751820826723213e-07,
+      "loss": 0.982,
+      "step": 2217
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 3.674054751963204e-07,
+      "loss": 1.104,
+      "step": 2218
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.672927114889999e-07,
+      "loss": 1.1639,
+      "step": 2219
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.671799171746958e-07,
+      "loss": 0.9873,
+      "step": 2220
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6706709228284135e-07,
+      "loss": 1.1272,
+      "step": 2221
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.669542368428775e-07,
+      "loss": 1.0,
+      "step": 2222
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.668413508842535e-07,
+      "loss": 1.0497,
+      "step": 2223
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6672843443642635e-07,
+      "loss": 0.9838,
+      "step": 2224
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.666154875288611e-07,
+      "loss": 1.0186,
+      "step": 2225
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.665025101910306e-07,
+      "loss": 1.0125,
+      "step": 2226
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.66389502452416e-07,
+      "loss": 1.0779,
+      "step": 2227
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6627646434250593e-07,
+      "loss": 0.8519,
+      "step": 2228
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.661633958907971e-07,
+      "loss": 0.8663,
+      "step": 2229
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6605029712679447e-07,
+      "loss": 0.8789,
+      "step": 2230
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.659371680800104e-07,
+      "loss": 0.9472,
+      "step": 2231
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.658240087799654e-07,
+      "loss": 1.0243,
+      "step": 2232
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6571081925618794e-07,
+      "loss": 0.9784,
+      "step": 2233
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6559759953821433e-07,
+      "loss": 1.0368,
+      "step": 2234
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6548434965558864e-07,
+      "loss": 1.1004,
+      "step": 2235
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.65371069637863e-07,
+      "loss": 1.0805,
+      "step": 2236
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.652577595145971e-07,
+      "loss": 0.9282,
+      "step": 2237
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.651444193153588e-07,
+      "loss": 0.9678,
+      "step": 2238
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6503104906972376e-07,
+      "loss": 1.0778,
+      "step": 2239
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.649176488072754e-07,
+      "loss": 0.9991,
+      "step": 2240
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6480421855760485e-07,
+      "loss": 1.0156,
+      "step": 2241
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.646907583503114e-07,
+      "loss": 1.0342,
+      "step": 2242
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.645772682150017e-07,
+      "loss": 0.8856,
+      "step": 2243
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.644637481812907e-07,
+      "loss": 0.9589,
+      "step": 2244
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.643501982788009e-07,
+      "loss": 1.0975,
+      "step": 2245
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6423661853716246e-07,
+      "loss": 1.0829,
+      "step": 2246
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6412300898601365e-07,
+      "loss": 0.9766,
+      "step": 2247
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.640093696550003e-07,
+      "loss": 0.9966,
+      "step": 2248
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.638957005737759e-07,
+      "loss": 1.004,
+      "step": 2249
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.637820017720022e-07,
+      "loss": 1.0673,
+      "step": 2250
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6366827327934814e-07,
+      "loss": 1.0326,
+      "step": 2251
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.635545151254906e-07,
+      "loss": 1.0229,
+      "step": 2252
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.634407273401143e-07,
+      "loss": 1.1061,
+      "step": 2253
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.633269099529117e-07,
+      "loss": 1.1802,
+      "step": 2254
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6321306299358285e-07,
+      "loss": 0.9738,
+      "step": 2255
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.630991864918356e-07,
+      "loss": 0.8429,
+      "step": 2256
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.629852804773854e-07,
+      "loss": 1.0607,
+      "step": 2257
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.628713449799556e-07,
+      "loss": 0.878,
+      "step": 2258
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.627573800292771e-07,
+      "loss": 1.0415,
+      "step": 2259
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.626433856550886e-07,
+      "loss": 1.0966,
+      "step": 2260
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.625293618871362e-07,
+      "loss": 0.8318,
+      "step": 2261
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6241530875517404e-07,
+      "loss": 1.0556,
+      "step": 2262
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6230122628896366e-07,
+      "loss": 1.0068,
+      "step": 2263
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.621871145182744e-07,
+      "loss": 1.1408,
+      "step": 2264
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6207297347288324e-07,
+      "loss": 0.9474,
+      "step": 2265
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.619588031825746e-07,
+      "loss": 0.9118,
+      "step": 2266
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.618446036771408e-07,
+      "loss": 1.0128,
+      "step": 2267
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.617303749863816e-07,
+      "loss": 1.2038,
+      "step": 2268
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.616161171401045e-07,
+      "loss": 0.9955,
+      "step": 2269
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.615018301681246e-07,
+      "loss": 0.9265,
+      "step": 2270
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6138751410026437e-07,
+      "loss": 1.0375,
+      "step": 2271
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6127316896635417e-07,
+      "loss": 0.9925,
+      "step": 2272
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6115879479623183e-07,
+      "loss": 1.1332,
+      "step": 2273
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.610443916197428e-07,
+      "loss": 1.1143,
+      "step": 2274
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6092995946673993e-07,
+      "loss": 1.0623,
+      "step": 2275
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.608154983670838e-07,
+      "loss": 1.0375,
+      "step": 2276
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.607010083506425e-07,
+      "loss": 1.0864,
+      "step": 2277
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6058648944729173e-07,
+      "loss": 0.9362,
+      "step": 2278
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6047194168691465e-07,
+      "loss": 1.0108,
+      "step": 2279
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6035736509940186e-07,
+      "loss": 1.0793,
+      "step": 2280
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6024275971465155e-07,
+      "loss": 1.0346,
+      "step": 2281
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6012812556256965e-07,
+      "loss": 0.9827,
+      "step": 2282
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6001346267306934e-07,
+      "loss": 0.9091,
+      "step": 2283
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.598987710760713e-07,
+      "loss": 0.9963,
+      "step": 2284
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5978405080150375e-07,
+      "loss": 0.8945,
+      "step": 2285
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5966930187930246e-07,
+      "loss": 0.9625,
+      "step": 2286
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5955452433941067e-07,
+      "loss": 0.9255,
+      "step": 2287
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5943971821177897e-07,
+      "loss": 1.2477,
+      "step": 2288
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.593248835263655e-07,
+      "loss": 1.1047,
+      "step": 2289
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.592100203131358e-07,
+      "loss": 1.2464,
+      "step": 2290
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.590951286020629e-07,
+      "loss": 0.9716,
+      "step": 2291
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5898020842312733e-07,
+      "loss": 1.0362,
+      "step": 2292
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.588652598063169e-07,
+      "loss": 1.0092,
+      "step": 2293
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5875028278162694e-07,
+      "loss": 0.9855,
+      "step": 2294
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.586352773790601e-07,
+      "loss": 1.0606,
+      "step": 2295
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5852024362862666e-07,
+      "loss": 1.1098,
+      "step": 2296
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5840518156034393e-07,
+      "loss": 1.0604,
+      "step": 2297
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.58290091204237e-07,
+      "loss": 1.0123,
+      "step": 2298
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.581749725903381e-07,
+      "loss": 1.038,
+      "step": 2299
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.580598257486867e-07,
+      "loss": 1.0141,
+      "step": 2300
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5794465070933013e-07,
+      "loss": 0.9175,
+      "step": 2301
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.578294475023227e-07,
+      "loss": 0.9536,
+      "step": 2302
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.57714216157726e-07,
+      "loss": 0.9434,
+      "step": 2303
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5759895670560925e-07,
+      "loss": 1.0294,
+      "step": 2304
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5748366917604883e-07,
+      "loss": 1.0704,
+      "step": 2305
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.573683535991284e-07,
+      "loss": 1.0935,
+      "step": 2306
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5725301000493917e-07,
+      "loss": 1.026,
+      "step": 2307
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5713763842357946e-07,
+      "loss": 1.0634,
+      "step": 2308
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5702223888515485e-07,
+      "loss": 0.8224,
+      "step": 2309
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5690681141977837e-07,
+      "loss": 0.9891,
+      "step": 2310
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5679135605757034e-07,
+      "loss": 0.9228,
+      "step": 2311
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.566758728286582e-07,
+      "loss": 0.9795,
+      "step": 2312
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.565603617631768e-07,
+      "loss": 0.9708,
+      "step": 2313
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5644482289126813e-07,
+      "loss": 1.0084,
+      "step": 2314
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.563292562430817e-07,
+      "loss": 1.0647,
+      "step": 2315
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.562136618487739e-07,
+      "loss": 1.0636,
+      "step": 2316
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.560980397385087e-07,
+      "loss": 0.9944,
+      "step": 2317
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.55982389942457e-07,
+      "loss": 0.9222,
+      "step": 2318
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5586671249079715e-07,
+      "loss": 0.9989,
+      "step": 2319
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5575100741371465e-07,
+      "loss": 1.0899,
+      "step": 2320
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.556352747414022e-07,
+      "loss": 1.0753,
+      "step": 2321
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.555195145040597e-07,
+      "loss": 0.9975,
+      "step": 2322
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5540372673189415e-07,
+      "loss": 0.9977,
+      "step": 2323
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5528791145511995e-07,
+      "loss": 1.0811,
+      "step": 2324
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.551720687039585e-07,
+      "loss": 1.1625,
+      "step": 2325
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5505619850863844e-07,
+      "loss": 1.1901,
+      "step": 2326
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.549403008993955e-07,
+      "loss": 1.012,
+      "step": 2327
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5482437590647273e-07,
+      "loss": 0.961,
+      "step": 2328
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5470842356012005e-07,
+      "loss": 1.0422,
+      "step": 2329
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.545924438905948e-07,
+      "loss": 1.1705,
+      "step": 2330
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.544764369281613e-07,
+      "loss": 1.1287,
+      "step": 2331
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.543604027030911e-07,
+      "loss": 0.9269,
+      "step": 2332
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.542443412456626e-07,
+      "loss": 1.0121,
+      "step": 2333
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.541282525861617e-07,
+      "loss": 1.0153,
+      "step": 2334
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5401213675488105e-07,
+      "loss": 0.9713,
+      "step": 2335
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5389599378212067e-07,
+      "loss": 0.9544,
+      "step": 2336
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5377982369818737e-07,
+      "loss": 1.0216,
+      "step": 2337
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.536636265333952e-07,
+      "loss": 1.1782,
+      "step": 2338
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.535474023180654e-07,
+      "loss": 0.9875,
+      "step": 2339
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.53431151082526e-07,
+      "loss": 1.003,
+      "step": 2340
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5331487285711234e-07,
+      "loss": 0.9896,
+      "step": 2341
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.531985676721666e-07,
+      "loss": 0.9612,
+      "step": 2342
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5308223555803805e-07,
+      "loss": 1.0305,
+      "step": 2343
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5296587654508316e-07,
+      "loss": 1.0299,
+      "step": 2344
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5284949066366497e-07,
+      "loss": 0.9385,
+      "step": 2345
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5273307794415417e-07,
+      "loss": 1.0568,
+      "step": 2346
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5261663841692785e-07,
+      "loss": 0.8936,
+      "step": 2347
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5250017211237046e-07,
+      "loss": 0.8708,
+      "step": 2348
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5238367906087344e-07,
+      "loss": 1.0562,
+      "step": 2349
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.52267159292835e-07,
+      "loss": 0.9835,
+      "step": 2350
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.521506128386604e-07,
+      "loss": 0.9213,
+      "step": 2351
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.520340397287619e-07,
+      "loss": 1.0736,
+      "step": 2352
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.519174399935587e-07,
+      "loss": 1.0405,
+      "step": 2353
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.518008136634771e-07,
+      "loss": 0.9192,
+      "step": 2354
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.516841607689501e-07,
+      "loss": 1.1552,
+      "step": 2355
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5156748134041766e-07,
+      "loss": 0.9899,
+      "step": 2356
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.514507754083267e-07,
+      "loss": 1.0547,
+      "step": 2357
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5133404300313123e-07,
+      "loss": 1.096,
+      "step": 2358
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.51217284155292e-07,
+      "loss": 0.9093,
+      "step": 2359
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.511004988952766e-07,
+      "loss": 1.0537,
+      "step": 2360
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5098368725355965e-07,
+      "loss": 1.0287,
+      "step": 2361
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.508668492606226e-07,
+      "loss": 0.8603,
+      "step": 2362
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5074998494695375e-07,
+      "loss": 1.0165,
+      "step": 2363
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.506330943430484e-07,
+      "loss": 1.0388,
+      "step": 2364
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.505161774794084e-07,
+      "loss": 1.1091,
+      "step": 2365
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.503992343865429e-07,
+      "loss": 0.9843,
+      "step": 2366
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.502822650949674e-07,
+      "loss": 0.9693,
+      "step": 2367
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5016526963520473e-07,
+      "loss": 1.0238,
+      "step": 2368
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5004824803778417e-07,
+      "loss": 0.9481,
+      "step": 2369
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4993120033324195e-07,
+      "loss": 0.8997,
+      "step": 2370
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.498141265521212e-07,
+      "loss": 0.9799,
+      "step": 2371
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4969702672497166e-07,
+      "loss": 0.9889,
+      "step": 2372
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4957990088235014e-07,
+      "loss": 1.2229,
+      "step": 2373
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4946274905481995e-07,
+      "loss": 0.8956,
+      "step": 2374
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4934557127295135e-07,
+      "loss": 1.0867,
+      "step": 2375
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4922836756732124e-07,
+      "loss": 1.0345,
+      "step": 2376
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.491111379685136e-07,
+      "loss": 0.9242,
+      "step": 2377
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.489938825071188e-07,
+      "loss": 1.171,
+      "step": 2378
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4887660121373414e-07,
+      "loss": 0.9755,
+      "step": 2379
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4875929411896354e-07,
+      "loss": 1.0922,
+      "step": 2380
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4864196125341775e-07,
+      "loss": 1.0652,
+      "step": 2381
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.485246026477144e-07,
+      "loss": 1.0857,
+      "step": 2382
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4840721833247753e-07,
+      "loss": 1.0492,
+      "step": 2383
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.48289808338338e-07,
+      "loss": 1.0828,
+      "step": 2384
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.481723726959335e-07,
+      "loss": 0.9753,
+      "step": 2385
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.480549114359082e-07,
+      "loss": 0.91,
+      "step": 2386
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4793742458891316e-07,
+      "loss": 0.8799,
+      "step": 2387
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4781991218560604e-07,
+      "loss": 1.0018,
+      "step": 2388
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.47702374256651e-07,
+      "loss": 0.9493,
+      "step": 2389
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4758481083271907e-07,
+      "loss": 1.0223,
+      "step": 2390
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.47467221944488e-07,
+      "loss": 1.0118,
+      "step": 2391
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.47349607622642e-07,
+      "loss": 0.9478,
+      "step": 2392
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.472319678978719e-07,
+      "loss": 0.9307,
+      "step": 2393
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4711430280087535e-07,
+      "loss": 1.0303,
+      "step": 2394
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4699661236235626e-07,
+      "loss": 1.0923,
+      "step": 2395
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.468788966130257e-07,
+      "loss": 1.0524,
+      "step": 2396
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4676115558360096e-07,
+      "loss": 0.9511,
+      "step": 2397
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4664338930480597e-07,
+      "loss": 1.0647,
+      "step": 2398
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4652559780737125e-07,
+      "loss": 0.9736,
+      "step": 2399
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4640778112203393e-07,
+      "loss": 1.0084,
+      "step": 2400
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.462899392795378e-07,
+      "loss": 1.1844,
+      "step": 2401
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.461720723106331e-07,
+      "loss": 0.8876,
+      "step": 2402
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.460541802460768e-07,
+      "loss": 1.0166,
+      "step": 2403
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.459362631166319e-07,
+      "loss": 1.0807,
+      "step": 2404
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4581832095306877e-07,
+      "loss": 0.9129,
+      "step": 2405
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4570035378616367e-07,
+      "loss": 1.0153,
+      "step": 2406
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4558236164669953e-07,
+      "loss": 0.9041,
+      "step": 2407
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.454643445654659e-07,
+      "loss": 1.0312,
+      "step": 2408
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.453463025732588e-07,
+      "loss": 1.0207,
+      "step": 2409
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4522823570088067e-07,
+      "loss": 1.0802,
+      "step": 2410
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4511014397914064e-07,
+      "loss": 0.9467,
+      "step": 2411
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4499202743885403e-07,
+      "loss": 1.2403,
+      "step": 2412
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.448738861108429e-07,
+      "loss": 0.9317,
+      "step": 2413
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4475572002593563e-07,
+      "loss": 1.0377,
+      "step": 2414
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4463752921496715e-07,
+      "loss": 1.0801,
+      "step": 2415
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4451931370877876e-07,
+      "loss": 0.9096,
+      "step": 2416
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4440107353821825e-07,
+      "loss": 1.0106,
+      "step": 2417
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4428280873413985e-07,
+      "loss": 1.0015,
+      "step": 2418
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.441645193274042e-07,
+      "loss": 1.0025,
+      "step": 2419
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.440462053488783e-07,
+      "loss": 1.0152,
+      "step": 2420
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4392786682943586e-07,
+      "loss": 1.1605,
+      "step": 2421
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.438095037999565e-07,
+      "loss": 0.9578,
+      "step": 2422
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4369111629132655e-07,
+      "loss": 1.2327,
+      "step": 2423
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.435727043344388e-07,
+      "loss": 1.0265,
+      "step": 2424
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4345426796019214e-07,
+      "loss": 1.087,
+      "step": 2425
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.433358071994922e-07,
+      "loss": 0.8441,
+      "step": 2426
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.432173220832505e-07,
+      "loss": 0.9447,
+      "step": 2427
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.430988126423853e-07,
+      "loss": 0.8568,
+      "step": 2428
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.429802789078212e-07,
+      "loss": 0.9807,
+      "step": 2429
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.428617209104888e-07,
+      "loss": 1.0182,
+      "step": 2430
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.427431386813254e-07,
+      "loss": 0.9399,
+      "step": 2431
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.426245322512744e-07,
+      "loss": 1.0278,
+      "step": 2432
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4250590165128567e-07,
+      "loss": 1.0937,
+      "step": 2433
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.423872469123153e-07,
+      "loss": 0.9342,
+      "step": 2434
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.422685680653256e-07,
+      "loss": 1.2278,
+      "step": 2435
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4214986514128543e-07,
+      "loss": 1.0156,
+      "step": 2436
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4203113817116953e-07,
+      "loss": 1.0415,
+      "step": 2437
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.419123871859592e-07,
+      "loss": 1.0253,
+      "step": 2438
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4179361221664214e-07,
+      "loss": 1.0582,
+      "step": 2439
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.41674813294212e-07,
+      "loss": 1.1242,
+      "step": 2440
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.415559904496688e-07,
+      "loss": 1.0037,
+      "step": 2441
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.414371437140188e-07,
+      "loss": 0.9417,
+      "step": 2442
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.413182731182744e-07,
+      "loss": 1.1205,
+      "step": 2443
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.411993786934546e-07,
+      "loss": 1.0389,
+      "step": 2444
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4108046047058414e-07,
+      "loss": 0.8817,
+      "step": 2445
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.409615184806941e-07,
+      "loss": 0.9783,
+      "step": 2446
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4084255275482205e-07,
+      "loss": 1.0091,
+      "step": 2447
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.407235633240114e-07,
+      "loss": 1.0937,
+      "step": 2448
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4060455021931187e-07,
+      "loss": 0.9263,
+      "step": 2449
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4048551347177943e-07,
+      "loss": 1.1305,
+      "step": 2450
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4036645311247626e-07,
+      "loss": 1.0977,
+      "step": 2451
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.402473691724704e-07,
+      "loss": 0.9923,
+      "step": 2452
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.401282616828363e-07,
+      "loss": 0.9222,
+      "step": 2453
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4000913067465465e-07,
+      "loss": 0.9303,
+      "step": 2454
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.398899761790119e-07,
+      "loss": 0.9163,
+      "step": 2455
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.397707982270011e-07,
+      "loss": 1.1051,
+      "step": 2456
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.39651596849721e-07,
+      "loss": 1.0849,
+      "step": 2457
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.395323720782767e-07,
+      "loss": 1.1633,
+      "step": 2458
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.394131239437793e-07,
+      "loss": 1.1322,
+      "step": 2459
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3929385247734617e-07,
+      "loss": 1.0252,
+      "step": 2460
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3917455771010044e-07,
+      "loss": 0.9736,
+      "step": 2461
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.390552396731717e-07,
+      "loss": 1.1097,
+      "step": 2462
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3893589839769545e-07,
+      "loss": 0.9224,
+      "step": 2463
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3881653391481304e-07,
+      "loss": 0.9525,
+      "step": 2464
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.386971462556722e-07,
+      "loss": 1.0116,
+      "step": 2465
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.385777354514266e-07,
+      "loss": 0.9696,
+      "step": 2466
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3845830153323583e-07,
+      "loss": 1.0789,
+      "step": 2467
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3833884453226575e-07,
+      "loss": 0.9231,
+      "step": 2468
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.382193644796881e-07,
+      "loss": 1.1036,
+      "step": 2469
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3809986140668044e-07,
+      "loss": 1.0327,
+      "step": 2470
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.379803353444267e-07,
+      "loss": 1.0937,
+      "step": 2471
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.378607863241167e-07,
+      "loss": 1.0664,
+      "step": 2472
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3774121437694604e-07,
+      "loss": 0.9663,
+      "step": 2473
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3762161953411655e-07,
+      "loss": 1.2265,
+      "step": 2474
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.375020018268359e-07,
+      "loss": 0.9936,
+      "step": 2475
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.373823612863178e-07,
+      "loss": 1.1059,
+      "step": 2476
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3726269794378193e-07,
+      "loss": 0.9684,
+      "step": 2477
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.371430118304538e-07,
+      "loss": 0.9551,
+      "step": 2478
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3702330297756496e-07,
+      "loss": 1.004,
+      "step": 2479
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3690357141635295e-07,
+      "loss": 0.8749,
+      "step": 2480
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3678381717806105e-07,
+      "loss": 0.9913,
+      "step": 2481
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3666404029393863e-07,
+      "loss": 0.9189,
+      "step": 2482
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.365442407952409e-07,
+      "loss": 1.0706,
+      "step": 2483
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3642441871322903e-07,
+      "loss": 1.0085,
+      "step": 2484
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3630457407916977e-07,
+      "loss": 0.9888,
+      "step": 2485
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.361847069243364e-07,
+      "loss": 1.0099,
+      "step": 2486
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.360648172800075e-07,
+      "loss": 0.991,
+      "step": 2487
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3594490517746773e-07,
+      "loss": 1.0939,
+      "step": 2488
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.358249706480075e-07,
+      "loss": 0.9988,
+      "step": 2489
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3570501372292324e-07,
+      "loss": 1.1027,
+      "step": 2490
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.355850344335173e-07,
+      "loss": 1.0223,
+      "step": 2491
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3546503281109747e-07,
+      "loss": 1.0002,
+      "step": 2492
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.353450088869777e-07,
+      "loss": 1.0461,
+      "step": 2493
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.352249626924777e-07,
+      "loss": 1.0948,
+      "step": 2494
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3510489425892293e-07,
+      "loss": 1.0448,
+      "step": 2495
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.349848036176447e-07,
+      "loss": 0.9319,
+      "step": 2496
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3486469079998004e-07,
+      "loss": 0.8994,
+      "step": 2497
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.34744555837272e-07,
+      "loss": 0.9819,
+      "step": 2498
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3462439876086903e-07,
+      "loss": 1.0317,
+      "step": 2499
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3450421960212564e-07,
+      "loss": 0.9895,
+      "step": 2500
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3438401839240217e-07,
+      "loss": 1.0868,
+      "step": 2501
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3426379516306425e-07,
+      "loss": 1.0609,
+      "step": 2502
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.341435499454838e-07,
+      "loss": 0.993,
+      "step": 2503
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3402328277103817e-07,
+      "loss": 1.0788,
+      "step": 2504
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.339029936711105e-07,
+      "loss": 0.9747,
+      "step": 2505
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3378268267708977e-07,
+      "loss": 0.9374,
+      "step": 2506
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.336623498203705e-07,
+      "loss": 1.113,
+      "step": 2507
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.335419951323529e-07,
+      "loss": 1.0797,
+      "step": 2508
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3342161864444307e-07,
+      "loss": 1.0108,
+      "step": 2509
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3330122038805277e-07,
+      "loss": 1.098,
+      "step": 2510
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.331808003945991e-07,
+      "loss": 0.9091,
+      "step": 2511
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.330603586955053e-07,
+      "loss": 0.9926,
+      "step": 2512
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.329398953222e-07,
+      "loss": 0.967,
+      "step": 2513
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.328194103061176e-07,
+      "loss": 1.0875,
+      "step": 2514
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3269890367869807e-07,
+      "loss": 0.8817,
+      "step": 2515
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.325783754713871e-07,
+      "loss": 1.0145,
+      "step": 2516
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.324578257156358e-07,
+      "loss": 1.0466,
+      "step": 2517
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.323372544429012e-07,
+      "loss": 0.9757,
+      "step": 2518
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.322166616846458e-07,
+      "loss": 1.1603,
+      "step": 2519
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3209604747233773e-07,
+      "loss": 0.9669,
+      "step": 2520
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.319754118374506e-07,
+      "loss": 0.9884,
+      "step": 2521
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3185475481146385e-07,
+      "loss": 0.98,
+      "step": 2522
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3173407642586226e-07,
+      "loss": 1.1481,
+      "step": 2523
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.316133767121363e-07,
+      "loss": 1.0014,
+      "step": 2524
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.314926557017821e-07,
+      "loss": 1.1501,
+      "step": 2525
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3137191342630113e-07,
+      "loss": 1.0475,
+      "step": 2526
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3125114991720056e-07,
+      "loss": 0.9523,
+      "step": 2527
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3113036520599304e-07,
+      "loss": 1.0857,
+      "step": 2528
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.310095593241968e-07,
+      "loss": 0.9966,
+      "step": 2529
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.308887323033356e-07,
+      "loss": 1.1456,
+      "step": 2530
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3076788417493866e-07,
+      "loss": 0.9529,
+      "step": 2531
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.306470149705406e-07,
+      "loss": 1.0193,
+      "step": 2532
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.305261247216819e-07,
+      "loss": 0.9518,
+      "step": 2533
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3040521345990824e-07,
+      "loss": 0.9183,
+      "step": 2534
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.302842812167707e-07,
+      "loss": 1.1124,
+      "step": 2535
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.301633280238261e-07,
+      "loss": 0.9828,
+      "step": 2536
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.300423539126367e-07,
+      "loss": 0.9486,
+      "step": 2537
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.299213589147699e-07,
+      "loss": 1.0822,
+      "step": 2538
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.298003430617989e-07,
+      "loss": 1.1402,
+      "step": 2539
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2967930638530237e-07,
+      "loss": 0.9439,
+      "step": 2540
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2955824891686403e-07,
+      "loss": 1.123,
+      "step": 2541
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2943717068807326e-07,
+      "loss": 0.9125,
+      "step": 2542
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.293160717305251e-07,
+      "loss": 0.9259,
+      "step": 2543
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.291949520758194e-07,
+      "loss": 1.0334,
+      "step": 2544
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2907381175556215e-07,
+      "loss": 1.031,
+      "step": 2545
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.28952650801364e-07,
+      "loss": 1.0428,
+      "step": 2546
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2883146924484165e-07,
+      "loss": 0.7934,
+      "step": 2547
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.287102671176166e-07,
+      "loss": 0.9843,
+      "step": 2548
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.285890444513162e-07,
+      "loss": 1.0762,
+      "step": 2549
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.284678012775727e-07,
+      "loss": 1.0006,
+      "step": 2550
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.283465376280241e-07,
+      "loss": 1.0519,
+      "step": 2551
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.282252535343135e-07,
+      "loss": 0.941,
+      "step": 2552
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2810394902808953e-07,
+      "loss": 1.0297,
+      "step": 2553
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2798262414100594e-07,
+      "loss": 1.0374,
+      "step": 2554
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.278612789047219e-07,
+      "loss": 0.9603,
+      "step": 2555
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2773991335090177e-07,
+      "loss": 1.0295,
+      "step": 2556
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.276185275112156e-07,
+      "loss": 0.899,
+      "step": 2557
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2749712141733824e-07,
+      "loss": 1.0753,
+      "step": 2558
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2737569510095014e-07,
+      "loss": 1.0665,
+      "step": 2559
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.272542485937368e-07,
+      "loss": 0.9168,
+      "step": 2560
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.271327819273893e-07,
+      "loss": 0.9366,
+      "step": 2561
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2701129513360367e-07,
+      "loss": 1.06,
+      "step": 2562
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.268897882440813e-07,
+      "loss": 1.0207,
+      "step": 2563
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.26768261290529e-07,
+      "loss": 1.075,
+      "step": 2564
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.266467143046585e-07,
+      "loss": 1.1753,
+      "step": 2565
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2652514731818695e-07,
+      "loss": 0.96,
+      "step": 2566
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.264035603628368e-07,
+      "loss": 1.1647,
+      "step": 2567
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.262819534703354e-07,
+      "loss": 0.9309,
+      "step": 2568
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.261603266724156e-07,
+      "loss": 1.0317,
+      "step": 2569
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2603868000081543e-07,
+      "loss": 1.1032,
+      "step": 2570
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2591701348727785e-07,
+      "loss": 0.8114,
+      "step": 2571
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2579532716355127e-07,
+      "loss": 1.0525,
+      "step": 2572
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.256736210613892e-07,
+      "loss": 1.0992,
+      "step": 2573
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.255518952125502e-07,
+      "loss": 1.1359,
+      "step": 2574
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.254301496487981e-07,
+      "loss": 1.1841,
+      "step": 2575
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.253083844019018e-07,
+      "loss": 0.9113,
+      "step": 2576
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.251865995036354e-07,
+      "loss": 1.2117,
+      "step": 2577
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2506479498577806e-07,
+      "loss": 0.9558,
+      "step": 2578
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.249429708801141e-07,
+      "loss": 0.9489,
+      "step": 2579
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2482112721843305e-07,
+      "loss": 0.9841,
+      "step": 2580
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2469926403252926e-07,
+      "loss": 1.0063,
+      "step": 2581
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2457738135420264e-07,
+      "loss": 1.0423,
+      "step": 2582
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2445547921525764e-07,
+      "loss": 1.1511,
+      "step": 2583
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2433355764750415e-07,
+      "loss": 1.0719,
+      "step": 2584
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.24211616682757e-07,
+      "loss": 0.9711,
+      "step": 2585
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2408965635283614e-07,
+      "loss": 1.0224,
+      "step": 2586
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2396767668956655e-07,
+      "loss": 1.0392,
+      "step": 2587
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.238456777247783e-07,
+      "loss": 0.9665,
+      "step": 2588
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.237236594903063e-07,
+      "loss": 0.9563,
+      "step": 2589
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.236016220179908e-07,
+      "loss": 1.0723,
+      "step": 2590
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2347956533967683e-07,
+      "loss": 1.0214,
+      "step": 2591
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.233574894872145e-07,
+      "loss": 1.012,
+      "step": 2592
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 3.23235394492459e-07,
+      "loss": 1.0092,
+      "step": 2593
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2311328038727037e-07,
+      "loss": 1.1293,
+      "step": 2594
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2299114720351373e-07,
+      "loss": 0.9411,
+      "step": 2595
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2286899497305915e-07,
+      "loss": 0.9288,
+      "step": 2596
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2274682372778174e-07,
+      "loss": 1.113,
+      "step": 2597
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.226246334995615e-07,
+      "loss": 0.8952,
+      "step": 2598
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.225024243202833e-07,
+      "loss": 0.8818,
+      "step": 2599
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2238019622183716e-07,
+      "loss": 1.0135,
+      "step": 2600
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.222579492361179e-07,
+      "loss": 0.9573,
+      "step": 2601
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.221356833950253e-07,
+      "loss": 0.9522,
+      "step": 2602
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2201339873046406e-07,
+      "loss": 0.8108,
+      "step": 2603
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.218910952743438e-07,
+      "loss": 1.1144,
+      "step": 2604
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2176877305857894e-07,
+      "loss": 1.1288,
+      "step": 2605
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2164643211508907e-07,
+      "loss": 1.0198,
+      "step": 2606
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2152407247579835e-07,
+      "loss": 1.0912,
+      "step": 2607
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2140169417263597e-07,
+      "loss": 1.1045,
+      "step": 2608
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.21279297237536e-07,
+      "loss": 1.0519,
+      "step": 2609
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2115688170243734e-07,
+      "loss": 1.1783,
+      "step": 2610
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.210344475992838e-07,
+      "loss": 1.1632,
+      "step": 2611
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2091199496002387e-07,
+      "loss": 1.022,
+      "step": 2612
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2078952381661104e-07,
+      "loss": 0.8775,
+      "step": 2613
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2066703420100374e-07,
+      "loss": 0.9908,
+      "step": 2614
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2054452614516476e-07,
+      "loss": 0.9383,
+      "step": 2615
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.204219996810622e-07,
+      "loss": 1.0616,
+      "step": 2616
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2029945484066877e-07,
+      "loss": 1.0251,
+      "step": 2617
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2017689165596195e-07,
+      "loss": 0.9874,
+      "step": 2618
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.20054310158924e-07,
+      "loss": 1.1243,
+      "step": 2619
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.19931710381542e-07,
+      "loss": 1.1404,
+      "step": 2620
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.198090923558078e-07,
+      "loss": 1.2072,
+      "step": 2621
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.196864561137179e-07,
+      "loss": 1.1481,
+      "step": 2622
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.195638016872738e-07,
+      "loss": 1.0487,
+      "step": 2623
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1944112910848155e-07,
+      "loss": 0.8954,
+      "step": 2624
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1931843840935194e-07,
+      "loss": 1.0473,
+      "step": 2625
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1919572962190066e-07,
+      "loss": 1.0019,
+      "step": 2626
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.190730027781477e-07,
+      "loss": 1.2319,
+      "step": 2627
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1895025791011844e-07,
+      "loss": 1.0656,
+      "step": 2628
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.188274950498424e-07,
+      "loss": 0.9756,
+      "step": 2629
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.187047142293539e-07,
+      "loss": 0.8995,
+      "step": 2630
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.185819154806921e-07,
+      "loss": 1.1492,
+      "step": 2631
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1845909883590073e-07,
+      "loss": 0.9497,
+      "step": 2632
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1833626432702823e-07,
+      "loss": 1.0122,
+      "step": 2633
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1821341198612764e-07,
+      "loss": 1.0324,
+      "step": 2634
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1809054184525684e-07,
+      "loss": 1.0641,
+      "step": 2635
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1796765393647813e-07,
+      "loss": 1.1693,
+      "step": 2636
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1784474829185845e-07,
+      "loss": 1.0761,
+      "step": 2637
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.177218249434696e-07,
+      "loss": 0.9103,
+      "step": 2638
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.175988839233878e-07,
+      "loss": 0.8612,
+      "step": 2639
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1747592526369395e-07,
+      "loss": 1.1042,
+      "step": 2640
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.173529489964734e-07,
+      "loss": 1.0212,
+      "step": 2641
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.172299551538164e-07,
+      "loss": 0.9563,
+      "step": 2642
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1710694376781757e-07,
+      "loss": 0.946,
+      "step": 2643
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.169839148705762e-07,
+      "loss": 0.9198,
+      "step": 2644
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.16860868494196e-07,
+      "loss": 0.9661,
+      "step": 2645
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1673780467078534e-07,
+      "loss": 1.0148,
+      "step": 2646
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1661472343245723e-07,
+      "loss": 1.0982,
+      "step": 2647
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.164916248113291e-07,
+      "loss": 1.0444,
+      "step": 2648
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.16368508839523e-07,
+      "loss": 0.9834,
+      "step": 2649
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.162453755491655e-07,
+      "loss": 1.004,
+      "step": 2650
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.161222249723874e-07,
+      "loss": 0.9365,
+      "step": 2651
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.159990571413246e-07,
+      "loss": 0.9721,
+      "step": 2652
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1587587208811705e-07,
+      "loss": 1.1517,
+      "step": 2653
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.157526698449092e-07,
+      "loss": 0.9794,
+      "step": 2654
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1562945044385023e-07,
+      "loss": 0.9662,
+      "step": 2655
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1550621391709366e-07,
+      "loss": 1.0269,
+      "step": 2656
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.153829602967974e-07,
+      "loss": 0.9322,
+      "step": 2657
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1525968961512397e-07,
+      "loss": 0.9676,
+      "step": 2658
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.151364019042403e-07,
+      "loss": 0.9189,
+      "step": 2659
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1501309719631764e-07,
+      "loss": 0.9427,
+      "step": 2660
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.148897755235318e-07,
+      "loss": 0.9596,
+      "step": 2661
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1476643691806317e-07,
+      "loss": 1.1035,
+      "step": 2662
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.146430814120961e-07,
+      "loss": 0.9592,
+      "step": 2663
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.145197090378198e-07,
+      "loss": 1.0959,
+      "step": 2664
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.143963198274278e-07,
+      "loss": 1.1758,
+      "step": 2665
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1427291381311764e-07,
+      "loss": 0.9971,
+      "step": 2666
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.141494910270918e-07,
+      "loss": 1.0159,
+      "step": 2667
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.140260515015569e-07,
+      "loss": 1.019,
+      "step": 2668
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.139025952687237e-07,
+      "loss": 1.0776,
+      "step": 2669
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.137791223608076e-07,
+      "loss": 1.2103,
+      "step": 2670
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1365563281002836e-07,
+      "loss": 1.0821,
+      "step": 2671
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1353212664860993e-07,
+      "loss": 1.2424,
+      "step": 2672
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.134086039087807e-07,
+      "loss": 0.9779,
+      "step": 2673
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1328506462277336e-07,
+      "loss": 1.08,
+      "step": 2674
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1316150882282485e-07,
+      "loss": 0.9525,
+      "step": 2675
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.130379365411765e-07,
+      "loss": 1.1492,
+      "step": 2676
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.12914347810074e-07,
+      "loss": 1.0617,
+      "step": 2677
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1279074266176707e-07,
+      "loss": 0.974,
+      "step": 2678
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1266712112851003e-07,
+      "loss": 1.1986,
+      "step": 2679
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.125434832425613e-07,
+      "loss": 0.9751,
+      "step": 2680
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.124198290361837e-07,
+      "loss": 1.0426,
+      "step": 2681
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1229615854164406e-07,
+      "loss": 0.8567,
+      "step": 2682
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.121724717912138e-07,
+      "loss": 0.8895,
+      "step": 2683
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1204876881716807e-07,
+      "loss": 1.0498,
+      "step": 2684
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.119250496517869e-07,
+      "loss": 0.9821,
+      "step": 2685
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.118013143273542e-07,
+      "loss": 1.0015,
+      "step": 2686
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.11677562876158e-07,
+      "loss": 0.9078,
+      "step": 2687
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.115537953304907e-07,
+      "loss": 0.901,
+      "step": 2688
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.114300117226489e-07,
+      "loss": 0.919,
+      "step": 2689
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1130621208493323e-07,
+      "loss": 1.0497,
+      "step": 2690
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1118239644964884e-07,
+      "loss": 0.9664,
+      "step": 2691
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.110585648491047e-07,
+      "loss": 1.0509,
+      "step": 2692
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1093471731561406e-07,
+      "loss": 1.0172,
+      "step": 2693
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1081085388149436e-07,
+      "loss": 0.9769,
+      "step": 2694
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.106869745790673e-07,
+      "loss": 0.9933,
+      "step": 2695
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1056307944065856e-07,
+      "loss": 1.1643,
+      "step": 2696
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.104391684985979e-07,
+      "loss": 0.9978,
+      "step": 2697
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.103152417852193e-07,
+      "loss": 0.9656,
+      "step": 2698
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.10191299332861e-07,
+      "loss": 1.176,
+      "step": 2699
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.100673411738651e-07,
+      "loss": 0.919,
+      "step": 2700
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0994336734057804e-07,
+      "loss": 0.9809,
+      "step": 2701
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0981937786534995e-07,
+      "loss": 1.0457,
+      "step": 2702
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.096953727805355e-07,
+      "loss": 1.1433,
+      "step": 2703
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.095713521184931e-07,
+      "loss": 1.0688,
+      "step": 2704
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.094473159115855e-07,
+      "loss": 1.0609,
+      "step": 2705
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.093232641921793e-07,
+      "loss": 1.1065,
+      "step": 2706
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.091991969926453e-07,
+      "loss": 0.9294,
+      "step": 2707
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0907511434535807e-07,
+      "loss": 1.0513,
+      "step": 2708
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.089510162826965e-07,
+      "loss": 1.0002,
+      "step": 2709
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.088269028370435e-07,
+      "loss": 0.9629,
+      "step": 2710
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.087027740407857e-07,
+      "loss": 1.1617,
+      "step": 2711
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0857862992631404e-07,
+      "loss": 1.029,
+      "step": 2712
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0845447052602334e-07,
+      "loss": 1.1079,
+      "step": 2713
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.083302958723124e-07,
+      "loss": 1.0149,
+      "step": 2714
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.08206105997584e-07,
+      "loss": 0.9661,
+      "step": 2715
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.08081900934245e-07,
+      "loss": 1.2308,
+      "step": 2716
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.07957680714706e-07,
+      "loss": 0.9461,
+      "step": 2717
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.078334453713817e-07,
+      "loss": 1.087,
+      "step": 2718
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.077091949366908e-07,
+      "loss": 1.1872,
+      "step": 2719
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.075849294430558e-07,
+      "loss": 0.9481,
+      "step": 2720
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.074606489229032e-07,
+      "loss": 1.1087,
+      "step": 2721
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0733635340866354e-07,
+      "loss": 0.9771,
+      "step": 2722
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0721204293277106e-07,
+      "loss": 0.861,
+      "step": 2723
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0708771752766395e-07,
+      "loss": 1.023,
+      "step": 2724
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.069633772257844e-07,
+      "loss": 1.1286,
+      "step": 2725
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0683902205957846e-07,
+      "loss": 1.0864,
+      "step": 2726
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0671465206149593e-07,
+      "loss": 1.0474,
+      "step": 2727
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.065902672639907e-07,
+      "loss": 1.1807,
+      "step": 2728
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0646586769952026e-07,
+      "loss": 0.9017,
+      "step": 2729
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0634145340054627e-07,
+      "loss": 1.0708,
+      "step": 2730
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.062170243995339e-07,
+      "loss": 1.01,
+      "step": 2731
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.060925807289523e-07,
+      "loss": 0.9534,
+      "step": 2732
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.059681224212747e-07,
+      "loss": 1.0009,
+      "step": 2733
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0584364950897764e-07,
+      "loss": 1.0934,
+      "step": 2734
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.057191620245419e-07,
+      "loss": 1.1708,
+      "step": 2735
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0559466000045183e-07,
+      "loss": 1.0975,
+      "step": 2736
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.054701434691957e-07,
+      "loss": 0.9601,
+      "step": 2737
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0534561246326544e-07,
+      "loss": 0.9041,
+      "step": 2738
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.05221067015157e-07,
+      "loss": 1.1863,
+      "step": 2739
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0509650715736976e-07,
+      "loss": 0.978,
+      "step": 2740
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.049719329224071e-07,
+      "loss": 1.3429,
+      "step": 2741
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0484734434277606e-07,
+      "loss": 1.1353,
+      "step": 2742
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0472274145098743e-07,
+      "loss": 1.0604,
+      "step": 2743
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.045981242795558e-07,
+      "loss": 1.0173,
+      "step": 2744
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0447349286099945e-07,
+      "loss": 0.8983,
+      "step": 2745
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.043488472278402e-07,
+      "loss": 0.8225,
+      "step": 2746
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0422418741260404e-07,
+      "loss": 0.9636,
+      "step": 2747
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0409951344782014e-07,
+      "loss": 1.0013,
+      "step": 2748
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0397482536602165e-07,
+      "loss": 1.0796,
+      "step": 2749
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0385012319974533e-07,
+      "loss": 1.2373,
+      "step": 2750
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0372540698153166e-07,
+      "loss": 0.8583,
+      "step": 2751
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.036006767439247e-07,
+      "loss": 1.0339,
+      "step": 2752
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0347593251947226e-07,
+      "loss": 0.9508,
+      "step": 2753
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.033511743407258e-07,
+      "loss": 0.9544,
+      "step": 2754
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.032264022402402e-07,
+      "loss": 0.9879,
+      "step": 2755
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0310161625057436e-07,
+      "loss": 1.0091,
+      "step": 2756
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.029768164042906e-07,
+      "loss": 1.0209,
+      "step": 2757
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0285200273395474e-07,
+      "loss": 1.2007,
+      "step": 2758
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.027271752721363e-07,
+      "loss": 0.8672,
+      "step": 2759
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.026023340514085e-07,
+      "loss": 1.074,
+      "step": 2760
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0247747910434806e-07,
+      "loss": 1.0276,
+      "step": 2761
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0235261046353534e-07,
+      "loss": 0.909,
+      "step": 2762
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0222772816155416e-07,
+      "loss": 1.1215,
+      "step": 2763
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.02102832230992e-07,
+      "loss": 0.9387,
+      "step": 2764
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0197792270443976e-07,
+      "loss": 0.903,
+      "step": 2765
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0185299961449227e-07,
+      "loss": 1.0017,
+      "step": 2766
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.017280629937473e-07,
+      "loss": 1.1114,
+      "step": 2767
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.016031128748067e-07,
+      "loss": 1.0645,
+      "step": 2768
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.014781492902755e-07,
+      "loss": 1.0015,
+      "step": 2769
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.013531722727624e-07,
+      "loss": 1.0637,
+      "step": 2770
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.012281818548796e-07,
+      "loss": 1.0027,
+      "step": 2771
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.011031780692428e-07,
+      "loss": 1.0765,
+      "step": 2772
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.00978160948471e-07,
+      "loss": 1.0998,
+      "step": 2773
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0085313052518697e-07,
+      "loss": 1.0354,
+      "step": 2774
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0072808683201665e-07,
+      "loss": 1.0813,
+      "step": 2775
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.006030299015898e-07,
+      "loss": 1.003,
+      "step": 2776
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.004779597665394e-07,
+      "loss": 0.9588,
+      "step": 2777
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0035287645950184e-07,
+      "loss": 0.9894,
+      "step": 2778
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.00227780013117e-07,
+      "loss": 1.2296,
+      "step": 2779
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0010267046002835e-07,
+      "loss": 0.9522,
+      "step": 2780
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 2.999775478328826e-07,
+      "loss": 0.8733,
+      "step": 2781
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.998524121643298e-07,
+      "loss": 0.9258,
+      "step": 2782
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.997272634870236e-07,
+      "loss": 1.0187,
+      "step": 2783
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.996021018336209e-07,
+      "loss": 0.9623,
+      "step": 2784
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9947692723678213e-07,
+      "loss": 1.069,
+      "step": 2785
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9935173972917106e-07,
+      "loss": 1.0121,
+      "step": 2786
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9922653934345466e-07,
+      "loss": 1.1705,
+      "step": 2787
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9910132611230346e-07,
+      "loss": 0.938,
+      "step": 2788
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.989761000683911e-07,
+      "loss": 0.844,
+      "step": 2789
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.98850861244395e-07,
+      "loss": 1.1389,
+      "step": 2790
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.987256096729955e-07,
+      "loss": 0.9765,
+      "step": 2791
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.986003453868764e-07,
+      "loss": 0.9229,
+      "step": 2792
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.984750684187248e-07,
+      "loss": 1.0979,
+      "step": 2793
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.983497788012313e-07,
+      "loss": 1.0037,
+      "step": 2794
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.982244765670894e-07,
+      "loss": 1.0532,
+      "step": 2795
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9809916174899643e-07,
+      "loss": 1.1101,
+      "step": 2796
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.979738343796524e-07,
+      "loss": 1.1297,
+      "step": 2797
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.978484944917611e-07,
+      "loss": 1.0289,
+      "step": 2798
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.977231421180293e-07,
+      "loss": 0.9665,
+      "step": 2799
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.975977772911671e-07,
+      "loss": 0.878,
+      "step": 2800
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.974724000438879e-07,
+      "loss": 0.9373,
+      "step": 2801
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9734701040890834e-07,
+      "loss": 1.0466,
+      "step": 2802
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.972216084189482e-07,
+      "loss": 1.0859,
+      "step": 2803
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9709619410673054e-07,
+      "loss": 1.0965,
+      "step": 2804
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9697076750498174e-07,
+      "loss": 0.9655,
+      "step": 2805
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.968453286464312e-07,
+      "loss": 1.1021,
+      "step": 2806
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.967198775638115e-07,
+      "loss": 1.1436,
+      "step": 2807
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.965944142898587e-07,
+      "loss": 1.0948,
+      "step": 2808
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9646893885731176e-07,
+      "loss": 1.1358,
+      "step": 2809
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9634345129891294e-07,
+      "loss": 0.93,
+      "step": 2810
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.962179516474076e-07,
+      "loss": 1.0321,
+      "step": 2811
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.960924399355443e-07,
+      "loss": 1.0864,
+      "step": 2812
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.959669161960747e-07,
+      "loss": 0.9247,
+      "step": 2813
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9584138046175367e-07,
+      "loss": 0.8778,
+      "step": 2814
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.957158327653392e-07,
+      "loss": 1.1114,
+      "step": 2815
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.955902731395923e-07,
+      "loss": 1.0167,
+      "step": 2816
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.954647016172772e-07,
+      "loss": 1.063,
+      "step": 2817
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.953391182311612e-07,
+      "loss": 0.9329,
+      "step": 2818
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.952135230140147e-07,
+      "loss": 0.9317,
+      "step": 2819
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.950879159986113e-07,
+      "loss": 1.051,
+      "step": 2820
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.949622972177273e-07,
+      "loss": 0.889,
+      "step": 2821
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9483666670414245e-07,
+      "loss": 1.098,
+      "step": 2822
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9471102449063963e-07,
+      "loss": 0.9794,
+      "step": 2823
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.945853706100043e-07,
+      "loss": 0.8948,
+      "step": 2824
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9445970509502544e-07,
+      "loss": 1.1102,
+      "step": 2825
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9433402797849475e-07,
+      "loss": 0.9591,
+      "step": 2826
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.942083392932072e-07,
+      "loss": 1.1113,
+      "step": 2827
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.940826390719606e-07,
+      "loss": 0.9294,
+      "step": 2828
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.939569273475559e-07,
+      "loss": 1.0547,
+      "step": 2829
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9383120415279697e-07,
+      "loss": 0.9347,
+      "step": 2830
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.937054695204905e-07,
+      "loss": 1.1088,
+      "step": 2831
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.935797234834466e-07,
+      "loss": 0.9718,
+      "step": 2832
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9345396607447803e-07,
+      "loss": 1.041,
+      "step": 2833
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9332819732640056e-07,
+      "loss": 0.9269,
+      "step": 2834
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9320241727203303e-07,
+      "loss": 1.027,
+      "step": 2835
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.93076625944197e-07,
+      "loss": 0.9114,
+      "step": 2836
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.929508233757173e-07,
+      "loss": 1.0264,
+      "step": 2837
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9282500959942146e-07,
+      "loss": 0.9448,
+      "step": 2838
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9269918464813995e-07,
+      "loss": 1.1453,
+      "step": 2839
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9257334855470626e-07,
+      "loss": 0.9708,
+      "step": 2840
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9244750135195674e-07,
+      "loss": 1.1631,
+      "step": 2841
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9232164307273053e-07,
+      "loss": 0.9123,
+      "step": 2842
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9219577374986985e-07,
+      "loss": 1.0612,
+      "step": 2843
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.920698934162198e-07,
+      "loss": 1.0873,
+      "step": 2844
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9194400210462807e-07,
+      "loss": 0.9778,
+      "step": 2845
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.918180998479454e-07,
+      "loss": 0.9363,
+      "step": 2846
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.916921866790256e-07,
+      "loss": 1.0286,
+      "step": 2847
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9156626263072493e-07,
+      "loss": 0.8563,
+      "step": 2848
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9144032773590274e-07,
+      "loss": 1.1345,
+      "step": 2849
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.913143820274212e-07,
+      "loss": 0.8722,
+      "step": 2850
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.911884255381452e-07,
+      "loss": 1.0507,
+      "step": 2851
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.910624583009426e-07,
+      "loss": 0.965,
+      "step": 2852
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.909364803486838e-07,
+      "loss": 1.0688,
+      "step": 2853
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.908104917142422e-07,
+      "loss": 1.0017,
+      "step": 2854
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.90684492430494e-07,
+      "loss": 0.9839,
+      "step": 2855
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.905584825303181e-07,
+      "loss": 1.0662,
+      "step": 2856
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.904324620465962e-07,
+      "loss": 0.8779,
+      "step": 2857
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.903064310122128e-07,
+      "loss": 1.1009,
+      "step": 2858
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9018038946005504e-07,
+      "loss": 1.0197,
+      "step": 2859
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9005433742301274e-07,
+      "loss": 1.09,
+      "step": 2860
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8992827493397886e-07,
+      "loss": 1.0276,
+      "step": 2861
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8980220202584865e-07,
+      "loss": 1.0002,
+      "step": 2862
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.896761187315203e-07,
+      "loss": 1.0661,
+      "step": 2863
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.895500250838946e-07,
+      "loss": 1.0241,
+      "step": 2864
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.894239211158752e-07,
+      "loss": 1.0623,
+      "step": 2865
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.892978068603683e-07,
+      "loss": 1.1005,
+      "step": 2866
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.891716823502827e-07,
+      "loss": 1.0103,
+      "step": 2867
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.890455476185302e-07,
+      "loss": 0.8884,
+      "step": 2868
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.889194026980249e-07,
+      "loss": 0.9274,
+      "step": 2869
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8879324762168376e-07,
+      "loss": 1.0398,
+      "step": 2870
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8866708242242645e-07,
+      "loss": 1.0211,
+      "step": 2871
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.885409071331751e-07,
+      "loss": 1.0946,
+      "step": 2872
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8841472178685464e-07,
+      "loss": 1.133,
+      "step": 2873
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.882885264163924e-07,
+      "loss": 1.1414,
+      "step": 2874
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.881623210547186e-07,
+      "loss": 0.9562,
+      "step": 2875
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8803610573476597e-07,
+      "loss": 0.9664,
+      "step": 2876
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8790988048946967e-07,
+      "loss": 1.047,
+      "step": 2877
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.877836453517677e-07,
+      "loss": 1.0281,
+      "step": 2878
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.876574003546004e-07,
+      "loss": 1.09,
+      "step": 2879
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8753114553091094e-07,
+      "loss": 0.9178,
+      "step": 2880
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.874048809136449e-07,
+      "loss": 1.0103,
+      "step": 2881
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8727860653575044e-07,
+      "loss": 0.9551,
+      "step": 2882
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8715232243017816e-07,
+      "loss": 1.0728,
+      "step": 2883
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.870260286298814e-07,
+      "loss": 1.0732,
+      "step": 2884
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8689972516781604e-07,
+      "loss": 1.0368,
+      "step": 2885
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.867734120769402e-07,
+      "loss": 1.0245,
+      "step": 2886
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8664708939021465e-07,
+      "loss": 0.9632,
+      "step": 2887
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.865207571406029e-07,
+      "loss": 0.9371,
+      "step": 2888
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8639441536107065e-07,
+      "loss": 0.9233,
+      "step": 2889
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.862680640845862e-07,
+      "loss": 1.0624,
+      "step": 2890
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.861417033441204e-07,
+      "loss": 0.9943,
+      "step": 2891
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8601533317264636e-07,
+      "loss": 0.9648,
+      "step": 2892
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.858889536031398e-07,
+      "loss": 0.9649,
+      "step": 2893
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8576256466857906e-07,
+      "loss": 1.0023,
+      "step": 2894
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8563616640194453e-07,
+      "loss": 0.9679,
+      "step": 2895
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.855097588362193e-07,
+      "loss": 0.969,
+      "step": 2896
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.853833420043889e-07,
+      "loss": 0.9287,
+      "step": 2897
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8525691593944115e-07,
+      "loss": 0.8418,
+      "step": 2898
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.851304806743664e-07,
+      "loss": 1.1124,
+      "step": 2899
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.850040362421573e-07,
+      "loss": 0.9942,
+      "step": 2900
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8487758267580885e-07,
+      "loss": 0.9422,
+      "step": 2901
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8475112000831865e-07,
+      "loss": 1.0178,
+      "step": 2902
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8462464827268647e-07,
+      "loss": 0.9679,
+      "step": 2903
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8449816750191446e-07,
+      "loss": 0.8989,
+      "step": 2904
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.843716777290073e-07,
+      "loss": 1.0001,
+      "step": 2905
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 2.84245178986972e-07,
+      "loss": 1.1531,
+      "step": 2906
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8411867130881744e-07,
+      "loss": 0.9461,
+      "step": 2907
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8399215472755553e-07,
+      "loss": 0.8778,
+      "step": 2908
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8386562927620017e-07,
+      "loss": 1.0169,
+      "step": 2909
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8373909498776744e-07,
+      "loss": 1.0451,
+      "step": 2910
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.836125518952759e-07,
+      "loss": 1.1239,
+      "step": 2911
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.834860000317464e-07,
+      "loss": 0.8829,
+      "step": 2912
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8335943943020215e-07,
+      "loss": 0.9443,
+      "step": 2913
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.832328701236684e-07,
+      "loss": 1.0044,
+      "step": 2914
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8310629214517297e-07,
+      "loss": 1.1295,
+      "step": 2915
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8297970552774557e-07,
+      "loss": 1.1738,
+      "step": 2916
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8285311030441853e-07,
+      "loss": 1.0017,
+      "step": 2917
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.827265065082263e-07,
+      "loss": 1.0405,
+      "step": 2918
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8259989417220555e-07,
+      "loss": 0.887,
+      "step": 2919
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.824732733293951e-07,
+      "loss": 1.0154,
+      "step": 2920
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.823466440128361e-07,
+      "loss": 1.03,
+      "step": 2921
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8222000625557183e-07,
+      "loss": 1.1027,
+      "step": 2922
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.82093360090648e-07,
+      "loss": 0.9414,
+      "step": 2923
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8196670555111213e-07,
+      "loss": 0.9329,
+      "step": 2924
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.818400426700142e-07,
+      "loss": 0.9512,
+      "step": 2925
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8171337148040633e-07,
+      "loss": 1.0692,
+      "step": 2926
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8158669201534267e-07,
+      "loss": 1.1325,
+      "step": 2927
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8146000430787974e-07,
+      "loss": 1.2362,
+      "step": 2928
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8133330839107604e-07,
+      "loss": 0.9854,
+      "step": 2929
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8120660429799236e-07,
+      "loss": 0.949,
+      "step": 2930
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.810798920616914e-07,
+      "loss": 1.0173,
+      "step": 2931
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.809531717152383e-07,
+      "loss": 0.9771,
+      "step": 2932
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.808264432917e-07,
+      "loss": 1.042,
+      "step": 2933
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.806997068241458e-07,
+      "loss": 0.9756,
+      "step": 2934
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8057296234564686e-07,
+      "loss": 1.1605,
+      "step": 2935
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8044620988927666e-07,
+      "loss": 1.003,
+      "step": 2936
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8031944948811054e-07,
+      "loss": 1.1485,
+      "step": 2937
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.801926811752262e-07,
+      "loss": 1.1674,
+      "step": 2938
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8006590498370315e-07,
+      "loss": 1.111,
+      "step": 2939
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7993912094662305e-07,
+      "loss": 1.1421,
+      "step": 2940
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.798123290970695e-07,
+      "loss": 1.0397,
+      "step": 2941
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7968552946812845e-07,
+      "loss": 1.0551,
+      "step": 2942
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.795587220928875e-07,
+      "loss": 1.0355,
+      "step": 2943
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.794319070044365e-07,
+      "loss": 1.0258,
+      "step": 2944
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.793050842358672e-07,
+      "loss": 1.0173,
+      "step": 2945
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.791782538202735e-07,
+      "loss": 1.0896,
+      "step": 2946
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.790514157907512e-07,
+      "loss": 1.087,
+      "step": 2947
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.789245701803981e-07,
+      "loss": 1.0259,
+      "step": 2948
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7879771702231377e-07,
+      "loss": 1.0988,
+      "step": 2949
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.786708563496001e-07,
+      "loss": 0.9106,
+      "step": 2950
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7854398819536086e-07,
+      "loss": 1.0194,
+      "step": 2951
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7841711259270154e-07,
+      "loss": 0.9376,
+      "step": 2952
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.782902295747299e-07,
+      "loss": 0.8637,
+      "step": 2953
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7816333917455537e-07,
+      "loss": 0.8265,
+      "step": 2954
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.780364414252894e-07,
+      "loss": 1.0263,
+      "step": 2955
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.779095363600453e-07,
+      "loss": 1.0585,
+      "step": 2956
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.777826240119386e-07,
+      "loss": 0.9235,
+      "step": 2957
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7765570441408624e-07,
+      "loss": 1.0336,
+      "step": 2958
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7752877759960735e-07,
+      "loss": 0.9785,
+      "step": 2959
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7740184360162294e-07,
+      "loss": 0.9685,
+      "step": 2960
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.772749024532558e-07,
+      "loss": 1.0038,
+      "step": 2961
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7714795418763067e-07,
+      "loss": 1.0193,
+      "step": 2962
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7702099883787403e-07,
+      "loss": 0.9752,
+      "step": 2963
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7689403643711433e-07,
+      "loss": 0.9981,
+      "step": 2964
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7676706701848185e-07,
+      "loss": 1.0419,
+      "step": 2965
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7664009061510863e-07,
+      "loss": 1.0222,
+      "step": 2966
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7651310726012856e-07,
+      "loss": 1.1289,
+      "step": 2967
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7638611698667733e-07,
+      "loss": 0.9283,
+      "step": 2968
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7625911982789254e-07,
+      "loss": 0.9768,
+      "step": 2969
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.761321158169134e-07,
+      "loss": 1.0134,
+      "step": 2970
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.76005104986881e-07,
+      "loss": 1.1555,
+      "step": 2971
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7587808737093844e-07,
+      "loss": 1.0077,
+      "step": 2972
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.757510630022301e-07,
+      "loss": 0.963,
+      "step": 2973
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7562403191390243e-07,
+      "loss": 0.9353,
+      "step": 2974
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7549699413910383e-07,
+      "loss": 1.1291,
+      "step": 2975
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.75369949710984e-07,
+      "loss": 0.9741,
+      "step": 2976
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7524289866269465e-07,
+      "loss": 0.9868,
+      "step": 2977
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.751158410273891e-07,
+      "loss": 1.0575,
+      "step": 2978
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.749887768382226e-07,
+      "loss": 1.1248,
+      "step": 2979
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7486170612835176e-07,
+      "loss": 0.9235,
+      "step": 2980
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7473462893093525e-07,
+      "loss": 0.9891,
+      "step": 2981
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7460754527913313e-07,
+      "loss": 0.8785,
+      "step": 2982
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.744804552061074e-07,
+      "loss": 1.074,
+      "step": 2983
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7435335874502146e-07,
+      "loss": 0.9152,
+      "step": 2984
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7422625592904073e-07,
+      "loss": 1.1032,
+      "step": 2985
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7409914679133203e-07,
+      "loss": 1.1362,
+      "step": 2986
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7397203136506386e-07,
+      "loss": 0.9834,
+      "step": 2987
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7384490968340634e-07,
+      "loss": 1.1819,
+      "step": 2988
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7371778177953145e-07,
+      "loss": 0.9905,
+      "step": 2989
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7359064768661245e-07,
+      "loss": 0.9699,
+      "step": 2990
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.734635074378245e-07,
+      "loss": 1.0341,
+      "step": 2991
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.733363610663441e-07,
+      "loss": 0.8972,
+      "step": 2992
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.732092086053497e-07,
+      "loss": 0.9669,
+      "step": 2993
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.73082050088021e-07,
+      "loss": 1.0137,
+      "step": 2994
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.729548855475395e-07,
+      "loss": 0.937,
+      "step": 2995
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7282771501708816e-07,
+      "loss": 1.0309,
+      "step": 2996
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7270053852985155e-07,
+      "loss": 1.0353,
+      "step": 2997
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7257335611901566e-07,
+      "loss": 0.9547,
+      "step": 2998
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.724461678177683e-07,
+      "loss": 0.9484,
+      "step": 2999
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7231897365929855e-07,
+      "loss": 1.1369,
+      "step": 3000
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.721917736767973e-07,
+      "loss": 0.9038,
+      "step": 3001
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.720645679034566e-07,
+      "loss": 0.8737,
+      "step": 3002
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.719373563724702e-07,
+      "loss": 1.0052,
+      "step": 3003
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7181013911703356e-07,
+      "loss": 1.0267,
+      "step": 3004
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7168291617034325e-07,
+      "loss": 0.987,
+      "step": 3005
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7155568756559747e-07,
+      "loss": 0.9832,
+      "step": 3006
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7142845333599604e-07,
+      "loss": 0.9687,
+      "step": 3007
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7130121351474013e-07,
+      "loss": 1.118,
+      "step": 3008
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.711739681350323e-07,
+      "loss": 1.0227,
+      "step": 3009
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.710467172300768e-07,
+      "loss": 1.1267,
+      "step": 3010
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.709194608330789e-07,
+      "loss": 0.8777,
+      "step": 3011
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.707921989772457e-07,
+      "loss": 0.9155,
+      "step": 3012
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7066493169578564e-07,
+      "loss": 1.0041,
+      "step": 3013
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7053765902190847e-07,
+      "loss": 1.0448,
+      "step": 3014
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7041038098882536e-07,
+      "loss": 1.1198,
+      "step": 3015
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7028309762974895e-07,
+      "loss": 1.0578,
+      "step": 3016
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7015580897789326e-07,
+      "loss": 1.0204,
+      "step": 3017
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.700285150664736e-07,
+      "loss": 1.0336,
+      "step": 3018
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6990121592870684e-07,
+      "loss": 1.1212,
+      "step": 3019
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6977391159781093e-07,
+      "loss": 0.9906,
+      "step": 3020
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6964660210700545e-07,
+      "loss": 1.0166,
+      "step": 3021
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.695192874895112e-07,
+      "loss": 0.9111,
+      "step": 3022
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6939196777855033e-07,
+      "loss": 0.9343,
+      "step": 3023
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.692646430073463e-07,
+      "loss": 0.8904,
+      "step": 3024
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.69137313209124e-07,
+      "loss": 1.0136,
+      "step": 3025
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6900997841710933e-07,
+      "loss": 0.9891,
+      "step": 3026
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6888263866452985e-07,
+      "loss": 1.0566,
+      "step": 3027
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6875529398461445e-07,
+      "loss": 0.9273,
+      "step": 3028
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6862794441059276e-07,
+      "loss": 0.9064,
+      "step": 3029
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6850058997569637e-07,
+      "loss": 0.8854,
+      "step": 3030
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 2.683732307131577e-07,
+      "loss": 1.0335,
+      "step": 3031
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.682458666562104e-07,
+      "loss": 1.055,
+      "step": 3032
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6811849783808976e-07,
+      "loss": 1.0323,
+      "step": 3033
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.679911242920321e-07,
+      "loss": 1.0242,
+      "step": 3034
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6786374605127466e-07,
+      "loss": 1.0936,
+      "step": 3035
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6773636314905644e-07,
+      "loss": 0.9543,
+      "step": 3036
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6760897561861737e-07,
+      "loss": 1.0196,
+      "step": 3037
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6748158349319863e-07,
+      "loss": 0.9656,
+      "step": 3038
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6735418680604255e-07,
+      "loss": 1.0384,
+      "step": 3039
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.672267855903927e-07,
+      "loss": 0.9994,
+      "step": 3040
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.670993798794938e-07,
+      "loss": 1.1458,
+      "step": 3041
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6697196970659193e-07,
+      "loss": 0.9479,
+      "step": 3042
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6684455510493407e-07,
+      "loss": 1.0213,
+      "step": 3043
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6671713610776846e-07,
+      "loss": 0.9755,
+      "step": 3044
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.665897127483444e-07,
+      "loss": 0.9694,
+      "step": 3045
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6646228505991265e-07,
+      "loss": 0.9123,
+      "step": 3046
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.663348530757247e-07,
+      "loss": 1.034,
+      "step": 3047
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.662074168290333e-07,
+      "loss": 1.1105,
+      "step": 3048
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6607997635309244e-07,
+      "loss": 1.0682,
+      "step": 3049
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6595253168115705e-07,
+      "loss": 0.9523,
+      "step": 3050
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.658250828464833e-07,
+      "loss": 1.12,
+      "step": 3051
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6569762988232837e-07,
+      "loss": 1.0602,
+      "step": 3052
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.655701728219504e-07,
+      "loss": 1.0309,
+      "step": 3053
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6544271169860875e-07,
+      "loss": 1.173,
+      "step": 3054
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6531524654556387e-07,
+      "loss": 0.8511,
+      "step": 3055
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.651877773960771e-07,
+      "loss": 0.9752,
+      "step": 3056
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.65060304283411e-07,
+      "loss": 1.0001,
+      "step": 3057
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.649328272408291e-07,
+      "loss": 0.9178,
+      "step": 3058
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.648053463015958e-07,
+      "loss": 1.0766,
+      "step": 3059
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6467786149897677e-07,
+      "loss": 1.0551,
+      "step": 3060
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.645503728662386e-07,
+      "loss": 0.9196,
+      "step": 3061
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.644228804366487e-07,
+      "loss": 0.9195,
+      "step": 3062
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.642953842434758e-07,
+      "loss": 0.9497,
+      "step": 3063
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.641678843199893e-07,
+      "loss": 0.9761,
+      "step": 3064
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6404038069945977e-07,
+      "loss": 0.9551,
+      "step": 3065
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6391287341515867e-07,
+      "loss": 0.8727,
+      "step": 3066
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.637853625003585e-07,
+      "loss": 0.9767,
+      "step": 3067
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6365784798833254e-07,
+      "loss": 1.1131,
+      "step": 3068
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.63530329912355e-07,
+      "loss": 0.9103,
+      "step": 3069
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.634028083057014e-07,
+      "loss": 1.1585,
+      "step": 3070
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6327528320164773e-07,
+      "loss": 0.9199,
+      "step": 3071
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6314775463347104e-07,
+      "loss": 0.8848,
+      "step": 3072
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6302022263444943e-07,
+      "loss": 1.0379,
+      "step": 3073
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.628926872378617e-07,
+      "loss": 1.1782,
+      "step": 3074
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.627651484769876e-07,
+      "loss": 1.0367,
+      "step": 3075
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.626376063851079e-07,
+      "loss": 1.0933,
+      "step": 3076
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.62510060995504e-07,
+      "loss": 0.9577,
+      "step": 3077
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6238251234145826e-07,
+      "loss": 1.045,
+      "step": 3078
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.622549604562539e-07,
+      "loss": 1.0563,
+      "step": 3079
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.62127405373175e-07,
+      "loss": 1.0076,
+      "step": 3080
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6199984712550653e-07,
+      "loss": 1.1519,
+      "step": 3081
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6187228574653427e-07,
+      "loss": 0.9338,
+      "step": 3082
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.617447212695445e-07,
+      "loss": 1.0331,
+      "step": 3083
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.616171537278248e-07,
+      "loss": 0.9679,
+      "step": 3084
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.614895831546633e-07,
+      "loss": 0.9577,
+      "step": 3085
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.613620095833488e-07,
+      "loss": 1.0102,
+      "step": 3086
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6123443304717116e-07,
+      "loss": 1.008,
+      "step": 3087
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6110685357942094e-07,
+      "loss": 0.9175,
+      "step": 3088
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.609792712133892e-07,
+      "loss": 1.2919,
+      "step": 3089
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.608516859823681e-07,
+      "loss": 1.0016,
+      "step": 3090
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6072409791965047e-07,
+      "loss": 1.0294,
+      "step": 3091
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6059650705852963e-07,
+      "loss": 0.9259,
+      "step": 3092
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6046891343229986e-07,
+      "loss": 1.0887,
+      "step": 3093
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6034131707425633e-07,
+      "loss": 0.9991,
+      "step": 3094
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.602137180176944e-07,
+      "loss": 1.0036,
+      "step": 3095
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.600861162959107e-07,
+      "loss": 1.0963,
+      "step": 3096
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.599585119422022e-07,
+      "loss": 0.8763,
+      "step": 3097
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5983090498986665e-07,
+      "loss": 1.0239,
+      "step": 3098
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5970329547220246e-07,
+      "loss": 1.0804,
+      "step": 3099
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5957568342250883e-07,
+      "loss": 0.9316,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.594480688740855e-07,
+      "loss": 0.9218,
+      "step": 3101
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.593204518602327e-07,
+      "loss": 1.0135,
+      "step": 3102
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.591928324142518e-07,
+      "loss": 0.9814,
+      "step": 3103
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.590652105694443e-07,
+      "loss": 0.9823,
+      "step": 3104
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.589375863591126e-07,
+      "loss": 0.887,
+      "step": 3105
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.588099598165596e-07,
+      "loss": 1.0679,
+      "step": 3106
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.586823309750888e-07,
+      "loss": 1.0567,
+      "step": 3107
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.585546998680044e-07,
+      "loss": 1.0718,
+      "step": 3108
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.584270665286113e-07,
+      "loss": 0.9265,
+      "step": 3109
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.582994309902146e-07,
+      "loss": 1.0499,
+      "step": 3110
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5817179328612014e-07,
+      "loss": 0.919,
+      "step": 3111
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.580441534496346e-07,
+      "loss": 1.1755,
+      "step": 3112
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.579165115140649e-07,
+      "loss": 1.0458,
+      "step": 3113
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.577888675127186e-07,
+      "loss": 1.1052,
+      "step": 3114
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5766122147890383e-07,
+      "loss": 1.2475,
+      "step": 3115
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.575335734459291e-07,
+      "loss": 1.0127,
+      "step": 3116
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.574059234471038e-07,
+      "loss": 0.9282,
+      "step": 3117
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5727827151573746e-07,
+      "loss": 1.0699,
+      "step": 3118
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5715061768514024e-07,
+      "loss": 0.9868,
+      "step": 3119
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.570229619886228e-07,
+      "loss": 1.0845,
+      "step": 3120
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5689530445949636e-07,
+      "loss": 1.1343,
+      "step": 3121
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.567676451310725e-07,
+      "loss": 1.0047,
+      "step": 3122
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5663998403666345e-07,
+      "loss": 0.9774,
+      "step": 3123
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5651232120958153e-07,
+      "loss": 1.0396,
+      "step": 3124
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5638465668314004e-07,
+      "loss": 0.8786,
+      "step": 3125
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.562569904906522e-07,
+      "loss": 0.8849,
+      "step": 3126
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.56129322665432e-07,
+      "loss": 0.9099,
+      "step": 3127
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5600165324079376e-07,
+      "loss": 1.0636,
+      "step": 3128
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.558739822500523e-07,
+      "loss": 1.0456,
+      "step": 3129
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.557463097265226e-07,
+      "loss": 0.9195,
+      "step": 3130
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5561863570352035e-07,
+      "loss": 0.9443,
+      "step": 3131
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5549096021436144e-07,
+      "loss": 1.0491,
+      "step": 3132
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.553632832923622e-07,
+      "loss": 1.0574,
+      "step": 3133
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5523560497083924e-07,
+      "loss": 1.0755,
+      "step": 3134
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5510792528310975e-07,
+      "loss": 0.9026,
+      "step": 3135
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.54980244262491e-07,
+      "loss": 1.0874,
+      "step": 3136
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.548525619423009e-07,
+      "loss": 0.9902,
+      "step": 3137
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.547248783558576e-07,
+      "loss": 0.9023,
+      "step": 3138
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5459719353647933e-07,
+      "loss": 1.1239,
+      "step": 3139
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5446950751748487e-07,
+      "loss": 1.0285,
+      "step": 3140
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.543418203321935e-07,
+      "loss": 1.1026,
+      "step": 3141
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.542141320139244e-07,
+      "loss": 1.0804,
+      "step": 3142
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5408644259599724e-07,
+      "loss": 1.0591,
+      "step": 3143
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5395875211173207e-07,
+      "loss": 1.071,
+      "step": 3144
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5383106059444907e-07,
+      "loss": 0.8962,
+      "step": 3145
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.537033680774688e-07,
+      "loss": 1.1057,
+      "step": 3146
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5357567459411193e-07,
+      "loss": 1.0775,
+      "step": 3147
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.534479801776996e-07,
+      "loss": 0.9532,
+      "step": 3148
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5332028486155286e-07,
+      "loss": 0.9578,
+      "step": 3149
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5319258867899344e-07,
+      "loss": 1.0865,
+      "step": 3150
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.530648916633429e-07,
+      "loss": 1.0004,
+      "step": 3151
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5293719384792336e-07,
+      "loss": 1.0282,
+      "step": 3152
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.528094952660568e-07,
+      "loss": 0.9541,
+      "step": 3153
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.526817959510655e-07,
+      "loss": 1.0915,
+      "step": 3154
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.525540959362722e-07,
+      "loss": 1.0499,
+      "step": 3155
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 2.524263952549995e-07,
+      "loss": 1.0208,
+      "step": 3156
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5229869394057035e-07,
+      "loss": 0.9342,
+      "step": 3157
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.521709920263078e-07,
+      "loss": 1.1141,
+      "step": 3158
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.52043289545535e-07,
+      "loss": 1.2225,
+      "step": 3159
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.519155865315754e-07,
+      "loss": 0.9048,
+      "step": 3160
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5178788301775244e-07,
+      "loss": 1.0795,
+      "step": 3161
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.516601790373899e-07,
+      "loss": 0.9217,
+      "step": 3162
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.515324746238113e-07,
+      "loss": 1.0784,
+      "step": 3163
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5140476981034057e-07,
+      "loss": 1.027,
+      "step": 3164
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5127706463030183e-07,
+      "loss": 1.0397,
+      "step": 3165
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5114935911701906e-07,
+      "loss": 1.0468,
+      "step": 3166
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.510216533038164e-07,
+      "loss": 1.0158,
+      "step": 3167
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.508939472240181e-07,
+      "loss": 0.9098,
+      "step": 3168
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5076624091094845e-07,
+      "loss": 1.0732,
+      "step": 3169
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.506385343979318e-07,
+      "loss": 1.0326,
+      "step": 3170
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.505108277182927e-07,
+      "loss": 1.1075,
+      "step": 3171
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.503831209053554e-07,
+      "loss": 1.0084,
+      "step": 3172
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.502554139924444e-07,
+      "loss": 0.9477,
+      "step": 3173
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5012770701288456e-07,
+      "loss": 1.1269,
+      "step": 3174
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5e-07,
+      "loss": 0.9038,
+      "step": 3175
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.498722929871155e-07,
+      "loss": 1.1256,
+      "step": 3176
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.497445860075555e-07,
+      "loss": 0.9051,
+      "step": 3177
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.496168790946446e-07,
+      "loss": 0.949,
+      "step": 3178
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.494891722817074e-07,
+      "loss": 1.145,
+      "step": 3179
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4936146560206813e-07,
+      "loss": 1.0838,
+      "step": 3180
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.492337590890516e-07,
+      "loss": 1.0749,
+      "step": 3181
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.491060527759819e-07,
+      "loss": 0.9666,
+      "step": 3182
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4897834669618356e-07,
+      "loss": 0.9993,
+      "step": 3183
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.488506408829809e-07,
+      "loss": 1.0064,
+      "step": 3184
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.487229353696982e-07,
+      "loss": 0.9216,
+      "step": 3185
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.485952301896594e-07,
+      "loss": 1.1433,
+      "step": 3186
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4846752537618874e-07,
+      "loss": 1.0689,
+      "step": 3187
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.483398209626102e-07,
+      "loss": 1.0348,
+      "step": 3188
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.482121169822475e-07,
+      "loss": 0.9312,
+      "step": 3189
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.480844134684246e-07,
+      "loss": 0.9432,
+      "step": 3190
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.47956710454465e-07,
+      "loss": 0.9905,
+      "step": 3191
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4782900797369223e-07,
+      "loss": 1.1104,
+      "step": 3192
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4770130605942963e-07,
+      "loss": 1.2338,
+      "step": 3193
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.475736047450005e-07,
+      "loss": 1.0391,
+      "step": 3194
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4744590406372777e-07,
+      "loss": 1.0057,
+      "step": 3195
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4731820404893454e-07,
+      "loss": 0.918,
+      "step": 3196
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.471905047339433e-07,
+      "loss": 0.9499,
+      "step": 3197
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.470628061520766e-07,
+      "loss": 1.025,
+      "step": 3198
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4693510833665706e-07,
+      "loss": 1.1327,
+      "step": 3199
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.468074113210066e-07,
+      "loss": 0.9412,
+      "step": 3200
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.466797151384471e-07,
+      "loss": 1.0645,
+      "step": 3201
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4655201982230044e-07,
+      "loss": 1.0937,
+      "step": 3202
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.464243254058881e-07,
+      "loss": 1.0031,
+      "step": 3203
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.462966319225312e-07,
+      "loss": 0.8863,
+      "step": 3204
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.461689394055509e-07,
+      "loss": 0.8792,
+      "step": 3205
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4604124788826796e-07,
+      "loss": 0.9459,
+      "step": 3206
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4591355740400274e-07,
+      "loss": 0.9384,
+      "step": 3207
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4578586798607565e-07,
+      "loss": 0.966,
+      "step": 3208
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.456581796678066e-07,
+      "loss": 1.0322,
+      "step": 3209
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.455304924825151e-07,
+      "loss": 1.042,
+      "step": 3210
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.454028064635207e-07,
+      "loss": 1.1272,
+      "step": 3211
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.452751216441425e-07,
+      "loss": 1.072,
+      "step": 3212
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.45147438057699e-07,
+      "loss": 0.9649,
+      "step": 3213
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4501975573750896e-07,
+      "loss": 0.9154,
+      "step": 3214
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.448920747168903e-07,
+      "loss": 0.9525,
+      "step": 3215
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.447643950291608e-07,
+      "loss": 1.0612,
+      "step": 3216
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4463671670763784e-07,
+      "loss": 1.0484,
+      "step": 3217
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 2.445090397856386e-07,
+      "loss": 0.8929,
+      "step": 3218
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4438136429647963e-07,
+      "loss": 1.0938,
+      "step": 3219
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.442536902734774e-07,
+      "loss": 0.966,
+      "step": 3220
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4412601774994773e-07,
+      "loss": 0.9803,
+      "step": 3221
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4399834675920616e-07,
+      "loss": 1.0436,
+      "step": 3222
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.43870677334568e-07,
+      "loss": 0.9959,
+      "step": 3223
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.437430095093479e-07,
+      "loss": 1.0408,
+      "step": 3224
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4361534331686e-07,
+      "loss": 1.0555,
+      "step": 3225
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4348767879041844e-07,
+      "loss": 0.9102,
+      "step": 3226
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4336001596333664e-07,
+      "loss": 0.9895,
+      "step": 3227
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4323235486892743e-07,
+      "loss": 1.0215,
+      "step": 3228
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.431046955405036e-07,
+      "loss": 1.1098,
+      "step": 3229
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.429770380113772e-07,
+      "loss": 1.0901,
+      "step": 3230
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4284938231485974e-07,
+      "loss": 1.1177,
+      "step": 3231
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.427217284842625e-07,
+      "loss": 1.0893,
+      "step": 3232
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.425940765528962e-07,
+      "loss": 1.1168,
+      "step": 3233
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.424664265540708e-07,
+      "loss": 1.0247,
+      "step": 3234
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4233877852109615e-07,
+      "loss": 1.0015,
+      "step": 3235
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.422111324872814e-07,
+      "loss": 0.812,
+      "step": 3236
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.420834884859351e-07,
+      "loss": 1.0431,
+      "step": 3237
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.419558465503654e-07,
+      "loss": 1.1082,
+      "step": 3238
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4182820671387984e-07,
+      "loss": 1.0326,
+      "step": 3239
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.417005690097854e-07,
+      "loss": 0.9463,
+      "step": 3240
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.415729334713887e-07,
+      "loss": 1.0877,
+      "step": 3241
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4144530013199556e-07,
+      "loss": 1.1217,
+      "step": 3242
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4131766902491114e-07,
+      "loss": 1.147,
+      "step": 3243
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.411900401834404e-07,
+      "loss": 1.0105,
+      "step": 3244
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.410624136408874e-07,
+      "loss": 1.0472,
+      "step": 3245
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4093478943055567e-07,
+      "loss": 1.1746,
+      "step": 3246
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4080716758574816e-07,
+      "loss": 1.2314,
+      "step": 3247
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.406795481397673e-07,
+      "loss": 1.0051,
+      "step": 3248
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4055193112591455e-07,
+      "loss": 1.053,
+      "step": 3249
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4042431657749115e-07,
+      "loss": 0.8953,
+      "step": 3250
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4029670452779757e-07,
+      "loss": 0.9489,
+      "step": 3251
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4016909501013333e-07,
+      "loss": 1.0952,
+      "step": 3252
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.400414880577978e-07,
+      "loss": 1.055,
+      "step": 3253
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.399138837040893e-07,
+      "loss": 1.1011,
+      "step": 3254
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.397862819823055e-07,
+      "loss": 1.222,
+      "step": 3255
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.396586829257437e-07,
+      "loss": 0.9382,
+      "step": 3256
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.395310865677001e-07,
+      "loss": 1.0523,
+      "step": 3257
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.394034929414704e-07,
+      "loss": 1.1099,
+      "step": 3258
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3927590208034957e-07,
+      "loss": 1.0997,
+      "step": 3259
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3914831401763194e-07,
+      "loss": 1.164,
+      "step": 3260
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.390207287866108e-07,
+      "loss": 1.0003,
+      "step": 3261
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3889314642057914e-07,
+      "loss": 1.0085,
+      "step": 3262
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.387655669528288e-07,
+      "loss": 1.0227,
+      "step": 3263
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3863799041665115e-07,
+      "loss": 0.9172,
+      "step": 3264
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.385104168453367e-07,
+      "loss": 0.9869,
+      "step": 3265
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3838284627217527e-07,
+      "loss": 1.0162,
+      "step": 3266
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3825527873045547e-07,
+      "loss": 1.1868,
+      "step": 3267
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3812771425346576e-07,
+      "loss": 1.1128,
+      "step": 3268
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3800015287449345e-07,
+      "loss": 1.1359,
+      "step": 3269
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.378725946268249e-07,
+      "loss": 0.9966,
+      "step": 3270
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.377450395437461e-07,
+      "loss": 0.9681,
+      "step": 3271
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3761748765854177e-07,
+      "loss": 1.0018,
+      "step": 3272
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.37489939004496e-07,
+      "loss": 0.9194,
+      "step": 3273
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3736239361489208e-07,
+      "loss": 1.0487,
+      "step": 3274
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3723485152301236e-07,
+      "loss": 1.1072,
+      "step": 3275
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3710731276213825e-07,
+      "loss": 0.986,
+      "step": 3276
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3697977736555057e-07,
+      "loss": 1.0318,
+      "step": 3277
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3685224536652894e-07,
+      "loss": 1.035,
+      "step": 3278
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3672471679835224e-07,
+      "loss": 0.9125,
+      "step": 3279
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3659719169429863e-07,
+      "loss": 0.9912,
+      "step": 3280
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 2.36469670087645e-07,
+      "loss": 0.9688,
+      "step": 3281
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3634215201166752e-07,
+      "loss": 1.0565,
+      "step": 3282
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3621463749964153e-07,
+      "loss": 0.8767,
+      "step": 3283
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3608712658484133e-07,
+      "loss": 1.1329,
+      "step": 3284
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3595961930054018e-07,
+      "loss": 0.9161,
+      "step": 3285
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.358321156800107e-07,
+      "loss": 1.1198,
+      "step": 3286
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3570461575652423e-07,
+      "loss": 0.8867,
+      "step": 3287
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.355771195633512e-07,
+      "loss": 1.0899,
+      "step": 3288
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3544962713376144e-07,
+      "loss": 1.023,
+      "step": 3289
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3532213850102326e-07,
+      "loss": 1.0293,
+      "step": 3290
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3519465369840418e-07,
+      "loss": 1.0564,
+      "step": 3291
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.350671727591709e-07,
+      "loss": 1.1333,
+      "step": 3292
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3493969571658902e-07,
+      "loss": 0.8016,
+      "step": 3293
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3481222260392285e-07,
+      "loss": 1.2297,
+      "step": 3294
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3468475345443616e-07,
+      "loss": 1.0635,
+      "step": 3295
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3455728830139128e-07,
+      "loss": 1.1795,
+      "step": 3296
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3442982717804963e-07,
+      "loss": 0.9777,
+      "step": 3297
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3430237011767164e-07,
+      "loss": 0.9031,
+      "step": 3298
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.341749171535167e-07,
+      "loss": 1.1116,
+      "step": 3299
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3404746831884288e-07,
+      "loss": 1.2142,
+      "step": 3300
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3392002364690757e-07,
+      "loss": 1.091,
+      "step": 3301
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.337925831709667e-07,
+      "loss": 1.0044,
+      "step": 3302
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3366514692427528e-07,
+      "loss": 0.9279,
+      "step": 3303
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3353771494008736e-07,
+      "loss": 1.0273,
+      "step": 3304
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3341028725165559e-07,
+      "loss": 1.1399,
+      "step": 3305
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3328286389223162e-07,
+      "loss": 1.1398,
+      "step": 3306
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.331554448950659e-07,
+      "loss": 0.9837,
+      "step": 3307
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3302803029340807e-07,
+      "loss": 1.0774,
+      "step": 3308
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.329006201205062e-07,
+      "loss": 0.9715,
+      "step": 3309
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3277321440960732e-07,
+      "loss": 1.1446,
+      "step": 3310
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3264581319395748e-07,
+      "loss": 1.2211,
+      "step": 3311
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3251841650680143e-07,
+      "loss": 1.1059,
+      "step": 3312
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3239102438138258e-07,
+      "loss": 1.0604,
+      "step": 3313
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.322636368509436e-07,
+      "loss": 0.9592,
+      "step": 3314
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3213625394872534e-07,
+      "loss": 1.0459,
+      "step": 3315
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3200887570796792e-07,
+      "loss": 0.996,
+      "step": 3316
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3188150216191022e-07,
+      "loss": 1.0328,
+      "step": 3317
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3175413334378966e-07,
+      "loss": 1.1853,
+      "step": 3318
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3162676928684238e-07,
+      "loss": 1.0501,
+      "step": 3319
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3149941002430366e-07,
+      "loss": 1.0778,
+      "step": 3320
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3137205558940724e-07,
+      "loss": 0.9976,
+      "step": 3321
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3124470601538558e-07,
+      "loss": 1.0468,
+      "step": 3322
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3111736133547013e-07,
+      "loss": 1.0754,
+      "step": 3323
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3099002158289073e-07,
+      "loss": 0.9181,
+      "step": 3324
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3086268679087606e-07,
+      "loss": 0.9458,
+      "step": 3325
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3073535699265374e-07,
+      "loss": 1.0073,
+      "step": 3326
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3060803222144976e-07,
+      "loss": 1.1361,
+      "step": 3327
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.304807125104888e-07,
+      "loss": 0.965,
+      "step": 3328
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3035339789299455e-07,
+      "loss": 0.9489,
+      "step": 3329
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3022608840218912e-07,
+      "loss": 0.9762,
+      "step": 3330
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.300987840712932e-07,
+      "loss": 1.1392,
+      "step": 3331
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2997148493352642e-07,
+      "loss": 0.9395,
+      "step": 3332
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2984419102210683e-07,
+      "loss": 0.9513,
+      "step": 3333
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2971690237025108e-07,
+      "loss": 0.9918,
+      "step": 3334
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2958961901117467e-07,
+      "loss": 1.2884,
+      "step": 3335
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.294623409780916e-07,
+      "loss": 0.9602,
+      "step": 3336
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2933506830421434e-07,
+      "loss": 1.1496,
+      "step": 3337
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2920780102275434e-07,
+      "loss": 1.0724,
+      "step": 3338
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2908053916692116e-07,
+      "loss": 0.9048,
+      "step": 3339
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2895328276992324e-07,
+      "loss": 0.9405,
+      "step": 3340
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2882603186496767e-07,
+      "loss": 0.8747,
+      "step": 3341
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2869878648525995e-07,
+      "loss": 0.91,
+      "step": 3342
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2857154666400396e-07,
+      "loss": 0.9373,
+      "step": 3343
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2844431243440254e-07,
+      "loss": 0.9692,
+      "step": 3344
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2831708382965686e-07,
+      "loss": 1.0325,
+      "step": 3345
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2818986088296647e-07,
+      "loss": 1.0229,
+      "step": 3346
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.280626436275298e-07,
+      "loss": 1.0676,
+      "step": 3347
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2793543209654348e-07,
+      "loss": 1.1075,
+      "step": 3348
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.278082263232027e-07,
+      "loss": 1.1211,
+      "step": 3349
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2768102634070143e-07,
+      "loss": 0.9034,
+      "step": 3350
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2755383218223175e-07,
+      "loss": 1.0601,
+      "step": 3351
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2742664388098435e-07,
+      "loss": 1.1786,
+      "step": 3352
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.272994614701485e-07,
+      "loss": 1.1156,
+      "step": 3353
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2717228498291187e-07,
+      "loss": 0.9113,
+      "step": 3354
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2704511445246046e-07,
+      "loss": 0.9831,
+      "step": 3355
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2691794991197898e-07,
+      "loss": 0.9338,
+      "step": 3356
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.267907913946503e-07,
+      "loss": 1.069,
+      "step": 3357
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2666363893365586e-07,
+      "loss": 1.0129,
+      "step": 3358
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2653649256217555e-07,
+      "loss": 1.0526,
+      "step": 3359
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2640935231338758e-07,
+      "loss": 1.0862,
+      "step": 3360
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2628221822046855e-07,
+      "loss": 1.047,
+      "step": 3361
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2615509031659366e-07,
+      "loss": 0.9797,
+      "step": 3362
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2602796863493617e-07,
+      "loss": 1.0895,
+      "step": 3363
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2590085320866794e-07,
+      "loss": 0.9652,
+      "step": 3364
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2577374407095925e-07,
+      "loss": 1.1181,
+      "step": 3365
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2564664125497857e-07,
+      "loss": 1.0999,
+      "step": 3366
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2551954479389265e-07,
+      "loss": 1.024,
+      "step": 3367
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2539245472086688e-07,
+      "loss": 0.879,
+      "step": 3368
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.252653710690648e-07,
+      "loss": 1.0214,
+      "step": 3369
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.251382938716482e-07,
+      "loss": 1.0304,
+      "step": 3370
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2501122316177746e-07,
+      "loss": 1.0211,
+      "step": 3371
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2488415897261088e-07,
+      "loss": 1.0281,
+      "step": 3372
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.247571013373053e-07,
+      "loss": 1.1101,
+      "step": 3373
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.24630050289016e-07,
+      "loss": 1.0734,
+      "step": 3374
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.245030058608962e-07,
+      "loss": 0.9538,
+      "step": 3375
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.243759680860975e-07,
+      "loss": 0.9486,
+      "step": 3376
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2424893699776992e-07,
+      "loss": 0.9489,
+      "step": 3377
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2412191262906167e-07,
+      "loss": 0.862,
+      "step": 3378
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2399489501311893e-07,
+      "loss": 0.9698,
+      "step": 3379
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2386788418308665e-07,
+      "loss": 0.9195,
+      "step": 3380
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2374088017210752e-07,
+      "loss": 0.8804,
+      "step": 3381
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2361388301332262e-07,
+      "loss": 1.0697,
+      "step": 3382
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2348689273987145e-07,
+      "loss": 1.0042,
+      "step": 3383
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2335990938489142e-07,
+      "loss": 1.1805,
+      "step": 3384
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2323293298151813e-07,
+      "loss": 0.8974,
+      "step": 3385
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.231059635628857e-07,
+      "loss": 0.9196,
+      "step": 3386
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.22979001162126e-07,
+      "loss": 0.8462,
+      "step": 3387
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2285204581236934e-07,
+      "loss": 1.1249,
+      "step": 3388
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2272509754674422e-07,
+      "loss": 0.9904,
+      "step": 3389
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2259815639837712e-07,
+      "loss": 1.0103,
+      "step": 3390
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2247122240039265e-07,
+      "loss": 0.976,
+      "step": 3391
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2234429558591377e-07,
+      "loss": 1.0927,
+      "step": 3392
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2221737598806144e-07,
+      "loss": 1.0938,
+      "step": 3393
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2209046363995463e-07,
+      "loss": 1.0675,
+      "step": 3394
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2196355857471066e-07,
+      "loss": 1.1048,
+      "step": 3395
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2183666082544466e-07,
+      "loss": 0.8881,
+      "step": 3396
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2170977042527005e-07,
+      "loss": 0.9798,
+      "step": 3397
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2158288740729843e-07,
+      "loss": 1.1492,
+      "step": 3398
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2145601180463922e-07,
+      "loss": 1.1578,
+      "step": 3399
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2132914365039989e-07,
+      "loss": 0.9604,
+      "step": 3400
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2120228297768626e-07,
+      "loss": 1.1181,
+      "step": 3401
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2107542981960203e-07,
+      "loss": 1.1518,
+      "step": 3402
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2094858420924877e-07,
+      "loss": 1.086,
+      "step": 3403
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.208217461797265e-07,
+      "loss": 1.1527,
+      "step": 3404
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.206949157641328e-07,
+      "loss": 0.9972,
+      "step": 3405
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2056809299556347e-07,
+      "loss": 1.0893,
+      "step": 3406
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2044127790711253e-07,
+      "loss": 0.9693,
+      "step": 3407
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2031447053187164e-07,
+      "loss": 1.1025,
+      "step": 3408
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2018767090293048e-07,
+      "loss": 1.058,
+      "step": 3409
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2006087905337698e-07,
+      "loss": 1.103,
+      "step": 3410
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.199340950162969e-07,
+      "loss": 0.9472,
+      "step": 3411
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1980731882477376e-07,
+      "loss": 1.2163,
+      "step": 3412
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1968055051188944e-07,
+      "loss": 1.0787,
+      "step": 3413
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.195537901107234e-07,
+      "loss": 0.9524,
+      "step": 3414
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1942703765435315e-07,
+      "loss": 1.1529,
+      "step": 3415
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1930029317585422e-07,
+      "loss": 0.9648,
+      "step": 3416
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1917355670830002e-07,
+      "loss": 1.0268,
+      "step": 3417
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1904682828476168e-07,
+      "loss": 1.0596,
+      "step": 3418
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1892010793830858e-07,
+      "loss": 0.9532,
+      "step": 3419
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1879339570200767e-07,
+      "loss": 1.0995,
+      "step": 3420
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1866669160892389e-07,
+      "loss": 1.1148,
+      "step": 3421
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.185399956921203e-07,
+      "loss": 1.0824,
+      "step": 3422
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.184133079846574e-07,
+      "loss": 1.0859,
+      "step": 3423
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1828662851959373e-07,
+      "loss": 1.0473,
+      "step": 3424
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1815995732998582e-07,
+      "loss": 0.9504,
+      "step": 3425
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1803329444888793e-07,
+      "loss": 0.9635,
+      "step": 3426
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1790663990935197e-07,
+      "loss": 1.1062,
+      "step": 3427
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1777999374442812e-07,
+      "loss": 1.0386,
+      "step": 3428
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1765335598716392e-07,
+      "loss": 0.994,
+      "step": 3429
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1752672667060485e-07,
+      "loss": 1.1935,
+      "step": 3430
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1740010582779443e-07,
+      "loss": 0.9371,
+      "step": 3431
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.172734934917737e-07,
+      "loss": 1.1764,
+      "step": 3432
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1714688969558142e-07,
+      "loss": 1.1104,
+      "step": 3433
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1702029447225443e-07,
+      "loss": 1.0426,
+      "step": 3434
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.168937078548271e-07,
+      "loss": 1.2086,
+      "step": 3435
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1676712987633157e-07,
+      "loss": 1.0341,
+      "step": 3436
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1664056056979785e-07,
+      "loss": 0.9755,
+      "step": 3437
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1651399996825355e-07,
+      "loss": 1.0136,
+      "step": 3438
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1638744810472408e-07,
+      "loss": 1.1867,
+      "step": 3439
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1626090501223256e-07,
+      "loss": 0.7895,
+      "step": 3440
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1613437072379988e-07,
+      "loss": 1.0757,
+      "step": 3441
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1600784527244442e-07,
+      "loss": 0.9273,
+      "step": 3442
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1588132869118256e-07,
+      "loss": 0.871,
+      "step": 3443
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.157548210130281e-07,
+      "loss": 1.1348,
+      "step": 3444
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.156283222709926e-07,
+      "loss": 0.9869,
+      "step": 3445
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1550183249808551e-07,
+      "loss": 1.0894,
+      "step": 3446
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1537535172731364e-07,
+      "loss": 1.1102,
+      "step": 3447
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1524887999168138e-07,
+      "loss": 1.0226,
+      "step": 3448
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1512241732419116e-07,
+      "loss": 1.1315,
+      "step": 3449
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1499596375784279e-07,
+      "loss": 1.0031,
+      "step": 3450
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1486951932563358e-07,
+      "loss": 0.9579,
+      "step": 3451
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1474308406055883e-07,
+      "loss": 0.9698,
+      "step": 3452
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1461665799561107e-07,
+      "loss": 0.8736,
+      "step": 3453
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.144902411637806e-07,
+      "loss": 1.0101,
+      "step": 3454
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1436383359805548e-07,
+      "loss": 1.0758,
+      "step": 3455
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.14237435331421e-07,
+      "loss": 1.1458,
+      "step": 3456
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1411104639686012e-07,
+      "loss": 1.0562,
+      "step": 3457
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1398466682735364e-07,
+      "loss": 1.1108,
+      "step": 3458
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1385829665587966e-07,
+      "loss": 1.0553,
+      "step": 3459
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1373193591541376e-07,
+      "loss": 1.0384,
+      "step": 3460
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1360558463892935e-07,
+      "loss": 0.8721,
+      "step": 3461
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1347924285939712e-07,
+      "loss": 0.9824,
+      "step": 3462
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1335291060978527e-07,
+      "loss": 0.8924,
+      "step": 3463
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1322658792305986e-07,
+      "loss": 0.9236,
+      "step": 3464
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1310027483218407e-07,
+      "loss": 0.9583,
+      "step": 3465
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1297397137011858e-07,
+      "loss": 1.1391,
+      "step": 3466
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1284767756982182e-07,
+      "loss": 1.0886,
+      "step": 3467
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1272139346424964e-07,
+      "loss": 1.1047,
+      "step": 3468
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1259511908635508e-07,
+      "loss": 0.9268,
+      "step": 3469
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1246885446908906e-07,
+      "loss": 1.0112,
+      "step": 3470
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1234259964539965e-07,
+      "loss": 1.1071,
+      "step": 3471
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1221635464823234e-07,
+      "loss": 1.0418,
+      "step": 3472
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.120901195105303e-07,
+      "loss": 1.118,
+      "step": 3473
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.119638942652341e-07,
+      "loss": 1.0719,
+      "step": 3474
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1183767894528135e-07,
+      "loss": 1.0304,
+      "step": 3475
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.117114735836076e-07,
+      "loss": 1.1045,
+      "step": 3476
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.115852782131454e-07,
+      "loss": 0.882,
+      "step": 3477
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1145909286682486e-07,
+      "loss": 0.8849,
+      "step": 3478
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1133291757757358e-07,
+      "loss": 0.9628,
+      "step": 3479
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1120675237831627e-07,
+      "loss": 0.9553,
+      "step": 3480
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1108059730197514e-07,
+      "loss": 1.0336,
+      "step": 3481
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1095445238146983e-07,
+      "loss": 1.0255,
+      "step": 3482
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1082831764971734e-07,
+      "loss": 0.87,
+      "step": 3483
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1070219313963173e-07,
+      "loss": 1.0961,
+      "step": 3484
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1057607888412482e-07,
+      "loss": 1.068,
+      "step": 3485
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1044997491610538e-07,
+      "loss": 1.0092,
+      "step": 3486
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1032388126847963e-07,
+      "loss": 0.8888,
+      "step": 3487
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1019779797415133e-07,
+      "loss": 1.1638,
+      "step": 3488
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.100717250660212e-07,
+      "loss": 0.9341,
+      "step": 3489
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.099456625769872e-07,
+      "loss": 1.0716,
+      "step": 3490
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0981961053994501e-07,
+      "loss": 1.0268,
+      "step": 3491
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0969356898778724e-07,
+      "loss": 1.1219,
+      "step": 3492
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0956753795340374e-07,
+      "loss": 0.8847,
+      "step": 3493
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0944151746968186e-07,
+      "loss": 1.0198,
+      "step": 3494
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0931550756950597e-07,
+      "loss": 1.0032,
+      "step": 3495
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0918950828575776e-07,
+      "loss": 1.0723,
+      "step": 3496
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.090635196513162e-07,
+      "loss": 1.102,
+      "step": 3497
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0893754169905745e-07,
+      "loss": 0.8986,
+      "step": 3498
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0881157446185471e-07,
+      "loss": 1.2119,
+      "step": 3499
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0868561797257877e-07,
+      "loss": 1.1943,
+      "step": 3500
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 6250,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "total_flos": 1.3290887619257303e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}