diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,14282 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9996315854107822,
+  "eval_steps": 1018,
+  "global_step": 2035,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.5,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 0.9016,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "eval_loss": 0.9116213917732239,
+      "eval_runtime": 568.7438,
+      "eval_samples_per_second": 25.416,
+      "eval_steps_per_second": 25.416,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.4375,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 0.8603,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.5625,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 0.929,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.375,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 0.9048,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.6875,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.9532,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 0.8978,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.9375,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 0.9101,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.34375,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 0.9418,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.59375,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 0.8974,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.46875,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.8577,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.28125,
+      "learning_rate": 2.2e-06,
+      "loss": 0.8509,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.1875,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.8675,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.6e-06,
+      "loss": 0.8244,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 0.7831,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.890625,
+      "learning_rate": 3e-06,
+      "loss": 0.8098,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.875,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 0.8588,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 0.8085,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 0.7503,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 0.7667,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.7651,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8984375,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 0.7608,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8203125,
+      "learning_rate": 4.4e-06,
+      "loss": 0.7527,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8125,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 0.7421,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8671875,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 0.7549,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.765625,
+      "learning_rate": 5e-06,
+      "loss": 0.7255,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.2e-06,
+      "loss": 0.8398,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.400000000000001e-06,
+      "loss": 0.761,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6015625,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 0.7299,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7578125,
+      "learning_rate": 5.8e-06,
+      "loss": 0.7857,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.734375,
+      "learning_rate": 6e-06,
+      "loss": 0.79,
+      "step": 30
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.200000000000001e-06,
+      "loss": 0.7288,
+      "step": 31
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 0.7145,
+      "step": 32
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8671875,
+      "learning_rate": 6.600000000000001e-06,
+      "loss": 0.7715,
+      "step": 33
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 0.7236,
+      "step": 34
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.59375,
+      "learning_rate": 7e-06,
+      "loss": 0.7415,
+      "step": 35
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 0.7333,
+      "step": 36
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.4e-06,
+      "loss": 0.7585,
+      "step": 37
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.600000000000001e-06,
+      "loss": 0.7366,
+      "step": 38
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.800000000000002e-06,
+      "loss": 0.7887,
+      "step": 39
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8046875,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.7498,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.2e-06,
+      "loss": 0.7751,
+      "step": 41
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.46875,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 0.7044,
+      "step": 42
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.6e-06,
+      "loss": 0.7989,
+      "step": 43
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.8e-06,
+      "loss": 0.6756,
+      "step": 44
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.59375,
+      "learning_rate": 9e-06,
+      "loss": 0.7055,
+      "step": 45
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.200000000000002e-06,
+      "loss": 0.6867,
+      "step": 46
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.4e-06,
+      "loss": 0.733,
+      "step": 47
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 0.6843,
+      "step": 48
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 0.6854,
+      "step": 49
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1e-05,
+      "loss": 0.7275,
+      "step": 50
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.02e-05,
+      "loss": 0.7706,
+      "step": 51
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.04e-05,
+      "loss": 0.7248,
+      "step": 52
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.0600000000000002e-05,
+      "loss": 0.7263,
+      "step": 53
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.625,
+      "learning_rate": 1.0800000000000002e-05,
+      "loss": 0.7337,
+      "step": 54
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 0.7208,
+      "step": 55
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 0.7383,
+      "step": 56
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.14e-05,
+      "loss": 0.7165,
+      "step": 57
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.16e-05,
+      "loss": 0.6962,
+      "step": 58
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.18e-05,
+      "loss": 0.7624,
+      "step": 59
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.2e-05,
+      "loss": 0.7225,
+      "step": 60
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.22e-05,
+      "loss": 0.6628,
+      "step": 61
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2400000000000002e-05,
+      "loss": 0.7263,
+      "step": 62
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.2600000000000001e-05,
+      "loss": 0.7618,
+      "step": 63
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 0.7206,
+      "step": 64
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 0.7221,
+      "step": 65
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.3200000000000002e-05,
+      "loss": 0.624,
+      "step": 66
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.3400000000000002e-05,
+      "loss": 0.7245,
+      "step": 67
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.3600000000000002e-05,
+      "loss": 0.7005,
+      "step": 68
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.38e-05,
+      "loss": 0.6807,
+      "step": 69
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.4e-05,
+      "loss": 0.7272,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.4200000000000001e-05,
+      "loss": 0.6904,
+      "step": 71
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.4400000000000001e-05,
+      "loss": 0.7841,
+      "step": 72
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.46e-05,
+      "loss": 0.7029,
+      "step": 73
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.48e-05,
+      "loss": 0.701,
+      "step": 74
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.7981,
+      "step": 75
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.5200000000000002e-05,
+      "loss": 0.7645,
+      "step": 76
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.54e-05,
+      "loss": 0.7372,
+      "step": 77
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.5600000000000003e-05,
+      "loss": 0.7021,
+      "step": 78
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.58e-05,
+      "loss": 0.6485,
+      "step": 79
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.7822,
+      "step": 80
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.62e-05,
+      "loss": 0.6821,
+      "step": 81
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.75,
+      "learning_rate": 1.64e-05,
+      "loss": 0.709,
+      "step": 82
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.66e-05,
+      "loss": 0.6858,
+      "step": 83
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 0.6923,
+      "step": 84
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.7e-05,
+      "loss": 0.6923,
+      "step": 85
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.72e-05,
+      "loss": 0.7059,
+      "step": 86
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.625,
+      "learning_rate": 1.7400000000000003e-05,
+      "loss": 0.704,
+      "step": 87
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.76e-05,
+      "loss": 0.7421,
+      "step": 88
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.7800000000000002e-05,
+      "loss": 0.6674,
+      "step": 89
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.8e-05,
+      "loss": 0.6883,
+      "step": 90
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8200000000000002e-05,
+      "loss": 0.7115,
+      "step": 91
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.8400000000000003e-05,
+      "loss": 0.7215,
+      "step": 92
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.86e-05,
+      "loss": 0.6849,
+      "step": 93
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.88e-05,
+      "loss": 0.7196,
+      "step": 94
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9e-05,
+      "loss": 0.751,
+      "step": 95
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 0.6719,
+      "step": 96
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.94e-05,
+      "loss": 0.6116,
+      "step": 97
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.9600000000000002e-05,
+      "loss": 0.7248,
+      "step": 98
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.98e-05,
+      "loss": 0.7034,
+      "step": 99
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.859375,
+      "learning_rate": 2e-05,
+      "loss": 0.7036,
+      "step": 100
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.99999968689593e-05,
+      "loss": 0.6927,
+      "step": 101
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9999987475839145e-05,
+      "loss": 0.7343,
+      "step": 102
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.999997182064543e-05,
+      "loss": 0.6965,
+      "step": 103
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.9999949903387952e-05,
+      "loss": 0.7497,
+      "step": 104
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.999992172408044e-05,
+      "loss": 0.6605,
+      "step": 105
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9999887282740532e-05,
+      "loss": 0.6982,
+      "step": 106
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9999846579389805e-05,
+      "loss": 0.6734,
+      "step": 107
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.999979961405374e-05,
+      "loss": 0.7405,
+      "step": 108
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.9999746386761752e-05,
+      "loss": 0.7565,
+      "step": 109
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.999968689754717e-05,
+      "loss": 0.6793,
+      "step": 110
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0,
+      "learning_rate": 1.999962114644725e-05,
+      "loss": 0.7327,
+      "step": 111
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.999954913350316e-05,
+      "loss": 0.7208,
+      "step": 112
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9999470858759998e-05,
+      "loss": 0.6889,
+      "step": 113
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9999386322266785e-05,
+      "loss": 0.8007,
+      "step": 114
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9999295524076455e-05,
+      "loss": 0.7843,
+      "step": 115
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.625,
+      "learning_rate": 1.999919846424586e-05,
+      "loss": 0.737,
+      "step": 116
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.999909514283579e-05,
+      "loss": 0.71,
+      "step": 117
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.875,
+      "learning_rate": 1.9998985559910938e-05,
+      "loss": 0.7246,
+      "step": 118
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.625,
+      "learning_rate": 1.999886971553993e-05,
+      "loss": 0.7178,
+      "step": 119
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.999874760979531e-05,
+      "loss": 0.7346,
+      "step": 120
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9998619242753534e-05,
+      "loss": 0.7084,
+      "step": 121
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9998484614494995e-05,
+      "loss": 0.7094,
+      "step": 122
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.9998343725103993e-05,
+      "loss": 0.7124,
+      "step": 123
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9998196574668758e-05,
+      "loss": 0.7114,
+      "step": 124
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.9998043163281435e-05,
+      "loss": 0.7354,
+      "step": 125
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9997883491038092e-05,
+      "loss": 0.6739,
+      "step": 126
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9997717558038712e-05,
+      "loss": 0.6864,
+      "step": 127
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.999754536438721e-05,
+      "loss": 0.7355,
+      "step": 128
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.9997366910191417e-05,
+      "loss": 0.7689,
+      "step": 129
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.999718219556307e-05,
+      "loss": 0.7374,
+      "step": 130
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9996991220617853e-05,
+      "loss": 0.7149,
+      "step": 131
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9996793985475352e-05,
+      "loss": 0.6466,
+      "step": 132
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9996590490259072e-05,
+      "loss": 0.6827,
+      "step": 133
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9996380735096448e-05,
+      "loss": 0.7492,
+      "step": 134
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.999616472011883e-05,
+      "loss": 0.6763,
+      "step": 135
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9995942445461486e-05,
+      "loss": 0.6216,
+      "step": 136
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.999571391126361e-05,
+      "loss": 0.6957,
+      "step": 137
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9995479117668307e-05,
+      "loss": 0.6901,
+      "step": 138
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.9995238064822613e-05,
+      "loss": 0.6763,
+      "step": 139
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9994990752877473e-05,
+      "loss": 0.6978,
+      "step": 140
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9994737181987754e-05,
+      "loss": 0.6291,
+      "step": 141
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9994477352312248e-05,
+      "loss": 0.6968,
+      "step": 142
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9994211264013664e-05,
+      "loss": 0.6941,
+      "step": 143
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9993938917258622e-05,
+      "loss": 0.6805,
+      "step": 144
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9993660312217674e-05,
+      "loss": 0.6886,
+      "step": 145
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9993375449065277e-05,
+      "loss": 0.6413,
+      "step": 146
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9993084327979827e-05,
+      "loss": 0.7364,
+      "step": 147
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0,
+      "learning_rate": 1.9992786949143614e-05,
+      "loss": 0.716,
+      "step": 148
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9992483312742866e-05,
+      "loss": 0.7244,
+      "step": 149
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.999217341896772e-05,
+      "loss": 0.7269,
+      "step": 150
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.999185726801224e-05,
+      "loss": 0.6845,
+      "step": 151
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.999153486007439e-05,
+      "loss": 0.7126,
+      "step": 152
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.999120619535608e-05,
+      "loss": 0.6498,
+      "step": 153
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.999087127406311e-05,
+      "loss": 0.7071,
+      "step": 154
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.999053009640521e-05,
+      "loss": 0.7132,
+      "step": 155
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9990182662596037e-05,
+      "loss": 0.7231,
+      "step": 156
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9989828972853152e-05,
+      "loss": 0.728,
+      "step": 157
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.998946902739804e-05,
+      "loss": 0.6585,
+      "step": 158
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.99891028264561e-05,
+      "loss": 0.7073,
+      "step": 159
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9988730370256654e-05,
+      "loss": 0.7574,
+      "step": 160
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.9988351659032933e-05,
+      "loss": 0.6619,
+      "step": 161
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.625,
+      "learning_rate": 1.998796669302209e-05,
+      "loss": 0.7161,
+      "step": 162
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9987575472465193e-05,
+      "loss": 0.7021,
+      "step": 163
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.9987177997607227e-05,
+      "loss": 0.6701,
+      "step": 164
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9986774268697098e-05,
+      "loss": 0.6676,
+      "step": 165
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.998636428598762e-05,
+      "loss": 0.6299,
+      "step": 166
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5,
+      "learning_rate": 1.998594804973553e-05,
+      "loss": 0.6866,
+      "step": 167
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.998552556020148e-05,
+      "loss": 0.7179,
+      "step": 168
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.998509681765003e-05,
+      "loss": 0.7257,
+      "step": 169
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.998466182234967e-05,
+      "loss": 0.7377,
+      "step": 170
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.998422057457279e-05,
+      "loss": 0.7212,
+      "step": 171
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9983773074595708e-05,
+      "loss": 0.6875,
+      "step": 172
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9983319322698652e-05,
+      "loss": 0.7754,
+      "step": 173
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.9982859319165764e-05,
+      "loss": 0.6669,
+      "step": 174
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.9982393064285106e-05,
+      "loss": 0.6754,
+      "step": 175
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.9981920558348642e-05,
+      "loss": 0.7274,
+      "step": 176
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.9981441801652263e-05,
+      "loss": 0.6401,
+      "step": 177
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9980956794495773e-05,
+      "loss": 0.723,
+      "step": 178
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.9980465537182888e-05,
+      "loss": 0.7025,
+      "step": 179
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.997996803002123e-05,
+      "loss": 0.763,
+      "step": 180
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.997946427332235e-05,
+      "loss": 0.6597,
+      "step": 181
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.99789542674017e-05,
+      "loss": 0.7367,
+      "step": 182
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.997843801257865e-05,
+      "loss": 0.7159,
+      "step": 183
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.9977915509176487e-05,
+      "loss": 0.714,
+      "step": 184
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.99773867575224e-05,
+      "loss": 0.8047,
+      "step": 185
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.9976851757947505e-05,
+      "loss": 0.689,
+      "step": 186
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.997631051078682e-05,
+      "loss": 0.7044,
+      "step": 187
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.997576301637928e-05,
+      "loss": 0.7352,
+      "step": 188
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.9975209275067725e-05,
+      "loss": 0.6391,
+      "step": 189
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.997464928719892e-05,
+      "loss": 0.6635,
+      "step": 190
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.9974083053123524e-05,
+      "loss": 0.7106,
+      "step": 191
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9973510573196125e-05,
+      "loss": 0.6913,
+      "step": 192
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9972931847775217e-05,
+      "loss": 0.7066,
+      "step": 193
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9972346877223195e-05,
+      "loss": 0.6997,
+      "step": 194
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9971755661906376e-05,
+      "loss": 0.7283,
+      "step": 195
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.9971158202194985e-05,
+      "loss": 0.6609,
+      "step": 196
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9970554498463154e-05,
+      "loss": 0.7705,
+      "step": 197
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.9969944551088926e-05,
+      "loss": 0.6865,
+      "step": 198
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.996932836045426e-05,
+      "loss": 0.6763,
+      "step": 199
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.9968705926945015e-05,
+      "loss": 0.639,
+      "step": 200
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.9968077250950966e-05,
+      "loss": 0.6853,
+      "step": 201
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9967442332865794e-05,
+      "loss": 0.742,
+      "step": 202
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9966801173087093e-05,
+      "loss": 0.6963,
+      "step": 203
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9966153772016355e-05,
+      "loss": 0.689,
+      "step": 204
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9965500130058996e-05,
+      "loss": 0.7414,
+      "step": 205
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5,
+      "learning_rate": 1.9964840247624327e-05,
+      "loss": 0.6952,
+      "step": 206
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9964174125125574e-05,
+      "loss": 0.6907,
+      "step": 207
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.9963501762979868e-05,
+      "loss": 0.6669,
+      "step": 208
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.9962823161608246e-05,
+      "loss": 0.6585,
+      "step": 209
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9962138321435658e-05,
+      "loss": 0.7127,
+      "step": 210
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9961447242890953e-05,
+      "loss": 0.6754,
+      "step": 211
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9960749926406884e-05,
+      "loss": 0.7361,
+      "step": 212
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.996004637242013e-05,
+      "loss": 0.7343,
+      "step": 213
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.9959336581371254e-05,
+      "loss": 0.699,
+      "step": 214
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.9958620553704737e-05,
+      "loss": 0.6439,
+      "step": 215
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.995789828986895e-05,
+      "loss": 0.6831,
+      "step": 216
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.995716979031619e-05,
+      "loss": 0.6588,
+      "step": 217
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9956435055502657e-05,
+      "loss": 0.7266,
+      "step": 218
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9955694085888432e-05,
+      "loss": 0.708,
+      "step": 219
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.9954946881937525e-05,
+      "loss": 0.7696,
+      "step": 220
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.9954193444117837e-05,
+      "loss": 0.6375,
+      "step": 221
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5,
+      "learning_rate": 1.9953433772901184e-05,
+      "loss": 0.7116,
+      "step": 222
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.995266786876327e-05,
+      "loss": 0.7028,
+      "step": 223
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9951895732183714e-05,
+      "loss": 0.7348,
+      "step": 224
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9951117363646036e-05,
+      "loss": 0.6273,
+      "step": 225
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.9950332763637656e-05,
+      "loss": 0.7082,
+      "step": 226
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9949541932649894e-05,
+      "loss": 0.6711,
+      "step": 227
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.9948744871177975e-05,
+      "loss": 0.681,
+      "step": 228
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.994794157972103e-05,
+      "loss": 0.713,
+      "step": 229
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.9947132058782082e-05,
+      "loss": 0.6694,
+      "step": 230
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9946316308868067e-05,
+      "loss": 0.6946,
+      "step": 231
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.99454943304898e-05,
+      "loss": 0.7825,
+      "step": 232
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.9944666124162026e-05,
+      "loss": 0.7364,
+      "step": 233
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9943831690403364e-05,
+      "loss": 0.6986,
+      "step": 234
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.9942991029736346e-05,
+      "loss": 0.6729,
+      "step": 235
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9942144142687403e-05,
+      "loss": 0.7088,
+      "step": 236
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9941291029786863e-05,
+      "loss": 0.6854,
+      "step": 237
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9940431691568948e-05,
+      "loss": 0.7062,
+      "step": 238
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9939566128571783e-05,
+      "loss": 0.6979,
+      "step": 239
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9938694341337398e-05,
+      "loss": 0.7285,
+      "step": 240
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9937816330411704e-05,
+      "loss": 0.6982,
+      "step": 241
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5,
+      "learning_rate": 1.9936932096344522e-05,
+      "loss": 0.6554,
+      "step": 242
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.9936041639689567e-05,
+      "loss": 0.6972,
+      "step": 243
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.993514496100445e-05,
+      "loss": 0.6546,
+      "step": 244
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.993424206085068e-05,
+      "loss": 0.6449,
+      "step": 245
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.993333293979366e-05,
+      "loss": 0.7411,
+      "step": 246
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9932417598402684e-05,
+      "loss": 0.7642,
+      "step": 247
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.9931496037250953e-05,
+      "loss": 0.6688,
+      "step": 248
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.9930568256915553e-05,
+      "loss": 0.7231,
+      "step": 249
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.9929634257977467e-05,
+      "loss": 0.7129,
+      "step": 250
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9928694041021573e-05,
+      "loss": 0.7211,
+      "step": 251
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.9927747606636646e-05,
+      "loss": 0.6315,
+      "step": 252
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9926794955415347e-05,
+      "loss": 0.7308,
+      "step": 253
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9925836087954234e-05,
+      "loss": 0.6769,
+      "step": 254
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.992487100485376e-05,
+      "loss": 0.744,
+      "step": 255
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9923899706718265e-05,
+      "loss": 0.7107,
+      "step": 256
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.9922922194155986e-05,
+      "loss": 0.6815,
+      "step": 257
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.992193846777905e-05,
+      "loss": 0.6577,
+      "step": 258
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5,
+      "learning_rate": 1.9920948528203473e-05,
+      "loss": 0.6999,
+      "step": 259
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.991995237604916e-05,
+      "loss": 0.7146,
+      "step": 260
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9918950011939915e-05,
+      "loss": 0.8865,
+      "step": 261
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.9917941436503427e-05,
+      "loss": 0.6906,
+      "step": 262
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.991692665037127e-05,
+      "loss": 0.7068,
+      "step": 263
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.9915905654178914e-05,
+      "loss": 0.7812,
+      "step": 264
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9914878448565712e-05,
+      "loss": 0.6431,
+      "step": 265
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.9913845034174912e-05,
+      "loss": 0.6738,
+      "step": 266
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9912805411653646e-05,
+      "loss": 0.7191,
+      "step": 267
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9911759581652926e-05,
+      "loss": 0.7017,
+      "step": 268
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.9910707544827673e-05,
+      "loss": 0.6836,
+      "step": 269
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9909649301836675e-05,
+      "loss": 0.7877,
+      "step": 270
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.990858485334261e-05,
+      "loss": 0.7325,
+      "step": 271
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.990751420001205e-05,
+      "loss": 0.6577,
+      "step": 272
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9906437342515435e-05,
+      "loss": 0.7281,
+      "step": 273
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9905354281527116e-05,
+      "loss": 0.7742,
+      "step": 274
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.990426501772531e-05,
+      "loss": 0.7435,
+      "step": 275
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.990316955179212e-05,
+      "loss": 0.7028,
+      "step": 276
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.9902067884413535e-05,
+      "loss": 0.67,
+      "step": 277
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.990096001627943e-05,
+      "loss": 0.6653,
+      "step": 278
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9899845948083563e-05,
+      "loss": 0.7369,
+      "step": 279
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.989872568052357e-05,
+      "loss": 0.7235,
+      "step": 280
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.989759921430097e-05,
+      "loss": 0.6322,
+      "step": 281
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.9896466550121172e-05,
+      "loss": 0.6737,
+      "step": 282
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.9895327688693453e-05,
+      "loss": 0.7007,
+      "step": 283
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9894182630730983e-05,
+      "loss": 0.6779,
+      "step": 284
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.98930313769508e-05,
+      "loss": 0.6909,
+      "step": 285
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.9891873928073832e-05,
+      "loss": 0.6977,
+      "step": 286
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.9890710284824884e-05,
+      "loss": 0.7279,
+      "step": 287
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9889540447932635e-05,
+      "loss": 0.6983,
+      "step": 288
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.988836441812965e-05,
+      "loss": 0.6987,
+      "step": 289
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.988718219615237e-05,
+      "loss": 0.6706,
+      "step": 290
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.9885993782741107e-05,
+      "loss": 0.6829,
+      "step": 291
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.988479917864006e-05,
+      "loss": 0.7316,
+      "step": 292
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5,
+      "learning_rate": 1.9883598384597293e-05,
+      "loss": 0.7212,
+      "step": 293
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.988239140136476e-05,
+      "loss": 0.642,
+      "step": 294
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.9881178229698278e-05,
+      "loss": 0.652,
+      "step": 295
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9879958870357552e-05,
+      "loss": 0.6774,
+      "step": 296
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.9878733324106146e-05,
+      "loss": 0.6415,
+      "step": 297
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.9877501591711518e-05,
+      "loss": 0.7066,
+      "step": 298
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.987626367394498e-05,
+      "loss": 0.6716,
+      "step": 299
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.987501957158173e-05,
+      "loss": 0.6473,
+      "step": 300
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.987376928540083e-05,
+      "loss": 0.7348,
+      "step": 301
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.9872512816185226e-05,
+      "loss": 0.6406,
+      "step": 302
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.987125016472173e-05,
+      "loss": 0.6514,
+      "step": 303
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9869981331801016e-05,
+      "loss": 0.7304,
+      "step": 304
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.9868706318217645e-05,
+      "loss": 0.6186,
+      "step": 305
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.9867425124770037e-05,
+      "loss": 0.6952,
+      "step": 306
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.9866137752260488e-05,
+      "loss": 0.7245,
+      "step": 307
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9864844201495162e-05,
+      "loss": 0.7067,
+      "step": 308
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.9863544473284085e-05,
+      "loss": 0.7254,
+      "step": 309
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.9862238568441166e-05,
+      "loss": 0.6367,
+      "step": 310
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.9860926487784166e-05,
+      "loss": 0.6912,
+      "step": 311
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.985960823213473e-05,
+      "loss": 0.6509,
+      "step": 312
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5,
+      "learning_rate": 1.985828380231835e-05,
+      "loss": 0.6857,
+      "step": 313
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9856953199164396e-05,
+      "loss": 0.7121,
+      "step": 314
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.9855616423506106e-05,
+      "loss": 0.658,
+      "step": 315
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9854273476180585e-05,
+      "loss": 0.7979,
+      "step": 316
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.9852924358028785e-05,
+      "loss": 0.7108,
+      "step": 317
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.9851569069895546e-05,
+      "loss": 0.677,
+      "step": 318
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5,
+      "learning_rate": 1.9850207612629557e-05,
+      "loss": 0.6907,
+      "step": 319
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.9848839987083366e-05,
+      "loss": 0.6697,
+      "step": 320
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.9847466194113402e-05,
+      "loss": 0.6575,
+      "step": 321
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9846086234579942e-05,
+      "loss": 0.6673,
+      "step": 322
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9844700109347125e-05,
+      "loss": 0.6532,
+      "step": 323
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.984330781928296e-05,
+      "loss": 0.7025,
+      "step": 324
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.98419093652593e-05,
+      "loss": 0.6979,
+      "step": 325
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5,
+      "learning_rate": 1.984050474815188e-05,
+      "loss": 0.7119,
+      "step": 326
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.9839093968840275e-05,
+      "loss": 0.6672,
+      "step": 327
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.983767702820793e-05,
+      "loss": 0.7348,
+      "step": 328
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.9836253927142142e-05,
+      "loss": 0.6897,
+      "step": 329
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.9834824666534073e-05,
+      "loss": 0.737,
+      "step": 330
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.9833389247278728e-05,
+      "loss": 0.7639,
+      "step": 331
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.983194767027499e-05,
+      "loss": 0.7069,
+      "step": 332
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.983049993642558e-05,
+      "loss": 0.6926,
+      "step": 333
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.982904604663708e-05,
+      "loss": 0.7287,
+      "step": 334
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.9827586001819933e-05,
+      "loss": 0.6506,
+      "step": 335
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9826119802888428e-05,
+      "loss": 0.6558,
+      "step": 336
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.9824647450760707e-05,
+      "loss": 0.7364,
+      "step": 337
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9823168946358774e-05,
+      "loss": 0.6573,
+      "step": 338
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9821684290608476e-05,
+      "loss": 0.698,
+      "step": 339
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.9820193484439523e-05,
+      "loss": 0.6914,
+      "step": 340
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.981869652878546e-05,
+      "loss": 0.7224,
+      "step": 341
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.9817193424583706e-05,
+      "loss": 0.7037,
+      "step": 342
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.9815684172775507e-05,
+      "loss": 0.6767,
+      "step": 343
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.981416877430597e-05,
+      "loss": 0.7367,
+      "step": 344
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.981264723012405e-05,
+      "loss": 0.7263,
+      "step": 345
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.9811119541182556e-05,
+      "loss": 0.6538,
+      "step": 346
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.9809585708438137e-05,
+      "loss": 0.7329,
+      "step": 347
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.375,
+      "learning_rate": 1.9808045732851284e-05,
+      "loss": 0.6787,
+      "step": 348
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.980649961538635e-05,
+      "loss": 0.6859,
+      "step": 349
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.9804947357011525e-05,
+      "loss": 0.6553,
+      "step": 350
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.9803388958698844e-05,
+      "loss": 0.6965,
+      "step": 351
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.980182442142419e-05,
+      "loss": 0.6725,
+      "step": 352
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.9800253746167287e-05,
+      "loss": 0.6633,
+      "step": 353
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.9798676933911712e-05,
+      "loss": 0.7147,
+      "step": 354
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.9797093985644866e-05,
+      "loss": 0.6615,
+      "step": 355
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.9795504902358017e-05,
+      "loss": 0.6669,
+      "step": 356
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.979390968504625e-05,
+      "loss": 0.644,
+      "step": 357
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.9792308334708507e-05,
+      "loss": 0.6195,
+      "step": 358
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.375,
+      "learning_rate": 1.9790700852347567e-05,
+      "loss": 0.6916,
+      "step": 359
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.9789087238970056e-05,
+      "loss": 0.6822,
+      "step": 360
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.978746749558642e-05,
+      "loss": 0.6762,
+      "step": 361
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.978584162321096e-05,
+      "loss": 0.6657,
+      "step": 362
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9784209622861813e-05,
+      "loss": 0.6916,
+      "step": 363
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.9782571495560952e-05,
+      "loss": 0.6575,
+      "step": 364
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.978092724233418e-05,
+      "loss": 0.6985,
+      "step": 365
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.9779276864211145e-05,
+      "loss": 0.6784,
+      "step": 366
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.9777620362225325e-05,
+      "loss": 0.6708,
+      "step": 367
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.9775957737414037e-05,
+      "loss": 0.685,
+      "step": 368
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5,
+      "learning_rate": 1.9774288990818432e-05,
+      "loss": 0.6952,
+      "step": 369
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5,
+      "learning_rate": 1.9772614123483488e-05,
+      "loss": 0.6519,
+      "step": 370
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.9770933136458025e-05,
+      "loss": 0.6941,
+      "step": 371
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.9769246030794685e-05,
+      "loss": 0.6812,
+      "step": 372
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.9767552807549955e-05,
+      "loss": 0.6913,
+      "step": 373
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.976585346778414e-05,
+      "loss": 0.6795,
+      "step": 374
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.9764148012561384e-05,
+      "loss": 0.6272,
+      "step": 375
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.9762436442949653e-05,
+      "loss": 0.6938,
+      "step": 376
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9760718760020747e-05,
+      "loss": 0.7009,
+      "step": 377
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5,
+      "learning_rate": 1.9758994964850296e-05,
+      "loss": 0.6893,
+      "step": 378
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9757265058517748e-05,
+      "loss": 0.7572,
+      "step": 379
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.9755529042106394e-05,
+      "loss": 0.7131,
+      "step": 380
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.9753786916703334e-05,
+      "loss": 0.6768,
+      "step": 381
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9752038683399498e-05,
+      "loss": 0.6796,
+      "step": 382
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.9750284343289656e-05,
+      "loss": 0.6901,
+      "step": 383
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5,
+      "learning_rate": 1.974852389747238e-05,
+      "loss": 0.6774,
+      "step": 384
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.974675734705008e-05,
+      "loss": 0.6813,
+      "step": 385
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.9744984693128986e-05,
+      "loss": 0.6017,
+      "step": 386
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.9743205936819142e-05,
+      "loss": 0.6907,
+      "step": 387
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.974142107923442e-05,
+      "loss": 0.7277,
+      "step": 388
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.9739630121492523e-05,
+      "loss": 0.6211,
+      "step": 389
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.973783306471495e-05,
+      "loss": 0.633,
+      "step": 390
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.973602991002704e-05,
+      "loss": 0.6992,
+      "step": 391
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.973422065855794e-05,
+      "loss": 0.6569,
+      "step": 392
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9732405311440623e-05,
+      "loss": 0.6578,
+      "step": 393
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.973058386981187e-05,
+      "loss": 0.6809,
+      "step": 394
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.972875633481228e-05,
+      "loss": 0.7267,
+      "step": 395
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.972692270758627e-05,
+      "loss": 0.6958,
+      "step": 396
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.9725082989282084e-05,
+      "loss": 0.6788,
+      "step": 397
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.9723237181051757e-05,
+      "loss": 0.6431,
+      "step": 398
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.972138528405115e-05,
+      "loss": 0.6955,
+      "step": 399
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5,
+      "learning_rate": 1.9719527299439944e-05,
+      "loss": 0.6377,
+      "step": 400
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.9717663228381612e-05,
+      "loss": 0.6587,
+      "step": 401
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.9715793072043462e-05,
+      "loss": 0.6866,
+      "step": 402
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.9713916831596593e-05,
+      "loss": 0.6393,
+      "step": 403
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.971203450821593e-05,
+      "loss": 0.7157,
+      "step": 404
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.971014610308019e-05,
+      "loss": 0.6559,
+      "step": 405
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.9708251617371915e-05,
+      "loss": 0.7192,
+      "step": 406
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.970635105227744e-05,
+      "loss": 0.6766,
+      "step": 407
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.9704444408986922e-05,
+      "loss": 0.6985,
+      "step": 408
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.970253168869431e-05,
+      "loss": 0.7023,
+      "step": 409
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9700612892597376e-05,
+      "loss": 0.7332,
+      "step": 410
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.969868802189767e-05,
+      "loss": 0.6261,
+      "step": 411
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.9696757077800576e-05,
+      "loss": 0.6726,
+      "step": 412
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9694820061515254e-05,
+      "loss": 0.683,
+      "step": 413
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.969287697425469e-05,
+      "loss": 0.684,
+      "step": 414
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.969092781723566e-05,
+      "loss": 0.6934,
+      "step": 415
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5,
+      "learning_rate": 1.9688972591678735e-05,
+      "loss": 0.7163,
+      "step": 416
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.9687011298808297e-05,
+      "loss": 0.6999,
+      "step": 417
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.9685043939852525e-05,
+      "loss": 0.692,
+      "step": 418
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.9683070516043394e-05,
+      "loss": 0.6399,
+      "step": 419
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.968109102861668e-05,
+      "loss": 0.6722,
+      "step": 420
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.967910547881195e-05,
+      "loss": 0.6815,
+      "step": 421
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9677113867872574e-05,
+      "loss": 0.6587,
+      "step": 422
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5,
+      "learning_rate": 1.9675116197045713e-05,
+      "loss": 0.7171,
+      "step": 423
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.967311246758233e-05,
+      "loss": 0.7014,
+      "step": 424
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.9671102680737172e-05,
+      "loss": 0.6458,
+      "step": 425
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.966908683776878e-05,
+      "loss": 0.6667,
+      "step": 426
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.96670649399395e-05,
+      "loss": 0.7001,
+      "step": 427
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.966503698851546e-05,
+      "loss": 0.6653,
+      "step": 428
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.9663002984766576e-05,
+      "loss": 0.6875,
+      "step": 429
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.9660962929966553e-05,
+      "loss": 0.7169,
+      "step": 430
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.96589168253929e-05,
+      "loss": 0.6653,
+      "step": 431
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.96568646723269e-05,
+      "loss": 0.7382,
+      "step": 432
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9654806472053627e-05,
+      "loss": 0.7223,
+      "step": 433
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.9652742225861946e-05,
+      "loss": 0.6542,
+      "step": 434
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.9650671935044497e-05,
+      "loss": 0.6436,
+      "step": 435
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.9648595600897722e-05,
+      "loss": 0.7841,
+      "step": 436
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.9646513224721833e-05,
+      "loss": 0.6503,
+      "step": 437
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9644424807820834e-05,
+      "loss": 0.6383,
+      "step": 438
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9642330351502507e-05,
+      "loss": 0.6654,
+      "step": 439
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9640229857078417e-05,
+      "loss": 0.6755,
+      "step": 440
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.9638123325863912e-05,
+      "loss": 0.6443,
+      "step": 441
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.9636010759178118e-05,
+      "loss": 0.6974,
+      "step": 442
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9633892158343943e-05,
+      "loss": 0.7253,
+      "step": 443
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9631767524688068e-05,
+      "loss": 0.6914,
+      "step": 444
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.9629636859540962e-05,
+      "loss": 0.6979,
+      "step": 445
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9627500164236857e-05,
+      "loss": 0.7394,
+      "step": 446
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.962535744011378e-05,
+      "loss": 0.6669,
+      "step": 447
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.9623208688513517e-05,
+      "loss": 0.6217,
+      "step": 448
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.9621053910781625e-05,
+      "loss": 0.5962,
+      "step": 449
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.9618893108267457e-05,
+      "loss": 0.7294,
+      "step": 450
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.9616726282324117e-05,
+      "loss": 0.7148,
+      "step": 451
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.961455343430849e-05,
+      "loss": 0.6368,
+      "step": 452
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9612374565581237e-05,
+      "loss": 0.7121,
+      "step": 453
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.9610189677506778e-05,
+      "loss": 0.6519,
+      "step": 454
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.9607998771453306e-05,
+      "loss": 0.7652,
+      "step": 455
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9605801848792788e-05,
+      "loss": 0.6815,
+      "step": 456
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.960359891090095e-05,
+      "loss": 0.6772,
+      "step": 457
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.375,
+      "learning_rate": 1.96013899591573e-05,
+      "loss": 0.6695,
+      "step": 458
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.959917499494509e-05,
+      "loss": 0.6973,
+      "step": 459
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9596954019651354e-05,
+      "loss": 0.7176,
+      "step": 460
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.9594727034666885e-05,
+      "loss": 0.7314,
+      "step": 461
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.959249404138624e-05,
+      "loss": 0.7324,
+      "step": 462
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.9590255041207734e-05,
+      "loss": 0.6373,
+      "step": 463
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.9588010035533446e-05,
+      "loss": 0.6645,
+      "step": 464
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.375,
+      "learning_rate": 1.9585759025769225e-05,
+      "loss": 0.6839,
+      "step": 465
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.9583502013324664e-05,
+      "loss": 0.7078,
+      "step": 466
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.9581238999613124e-05,
+      "loss": 0.695,
+      "step": 467
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.9578969986051727e-05,
+      "loss": 0.7439,
+      "step": 468
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.9576694974061342e-05,
+      "loss": 0.6497,
+      "step": 469
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.95744139650666e-05,
+      "loss": 0.701,
+      "step": 470
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.9572126960495893e-05,
+      "loss": 0.659,
+      "step": 471
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.956983396178136e-05,
+      "loss": 0.7553,
+      "step": 472
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9567534970358893e-05,
+      "loss": 0.7188,
+      "step": 473
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5,
+      "learning_rate": 1.956522998766814e-05,
+      "loss": 0.7454,
+      "step": 474
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.95629190151525e-05,
+      "loss": 0.7067,
+      "step": 475
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.9560602054259124e-05,
+      "loss": 0.7083,
+      "step": 476
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.955827910643891e-05,
+      "loss": 0.6644,
+      "step": 477
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.9555950173146505e-05,
+      "loss": 0.7333,
+      "step": 478
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.955361525584031e-05,
+      "loss": 0.7111,
+      "step": 479
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.375,
+      "learning_rate": 1.955127435598247e-05,
+      "loss": 0.675,
+      "step": 480
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5,
+      "learning_rate": 1.9548927475038875e-05,
+      "loss": 0.6496,
+      "step": 481
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.375,
+      "learning_rate": 1.9546574614479156e-05,
+      "loss": 0.6657,
+      "step": 482
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.95442157757767e-05,
+      "loss": 0.6306,
+      "step": 483
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5,
+      "learning_rate": 1.9541850960408622e-05,
+      "loss": 0.71,
+      "step": 484
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.95394801698558e-05,
+      "loss": 0.71,
+      "step": 485
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9537103405602836e-05,
+      "loss": 0.7118,
+      "step": 486
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.953472066913808e-05,
+      "loss": 0.6306,
+      "step": 487
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.9532331961953618e-05,
+      "loss": 0.7172,
+      "step": 488
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.375,
+      "learning_rate": 1.9529937285545285e-05,
+      "loss": 0.655,
+      "step": 489
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9527536641412637e-05,
+      "loss": 0.6988,
+      "step": 490
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.952513003105899e-05,
+      "loss": 0.7134,
+      "step": 491
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.952271745599137e-05,
+      "loss": 0.6504,
+      "step": 492
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.952029891772056e-05,
+      "loss": 0.7485,
+      "step": 493
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.951787441776106e-05,
+      "loss": 0.6564,
+      "step": 494
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.951544395763112e-05,
+      "loss": 0.7286,
+      "step": 495
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.951300753885271e-05,
+      "loss": 0.7277,
+      "step": 496
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.9510565162951538e-05,
+      "loss": 0.6459,
+      "step": 497
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.9508116831457035e-05,
+      "loss": 0.7214,
+      "step": 498
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.9505662545902368e-05,
+      "loss": 0.651,
+      "step": 499
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.9503202307824433e-05,
+      "loss": 0.7035,
+      "step": 500
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.950073611876385e-05,
+      "loss": 0.7042,
+      "step": 501
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9498263980264967e-05,
+      "loss": 0.6864,
+      "step": 502
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.9495785893875854e-05,
+      "loss": 0.6474,
+      "step": 503
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.9493301861148312e-05,
+      "loss": 0.7491,
+      "step": 504
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.949081188363786e-05,
+      "loss": 0.7347,
+      "step": 505
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9488315962903743e-05,
+      "loss": 0.6692,
+      "step": 506
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.948581410050893e-05,
+      "loss": 0.6904,
+      "step": 507
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.9483306298020106e-05,
+      "loss": 0.6724,
+      "step": 508
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.9480792557007677e-05,
+      "loss": 0.6707,
+      "step": 509
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9478272879045764e-05,
+      "loss": 0.6914,
+      "step": 510
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.9475747265712215e-05,
+      "loss": 0.6941,
+      "step": 511
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.947321571858859e-05,
+      "loss": 0.7098,
+      "step": 512
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.947067823926016e-05,
+      "loss": 0.6816,
+      "step": 513
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.9468134829315922e-05,
+      "loss": 0.7058,
+      "step": 514
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.9465585490348574e-05,
+      "loss": 0.674,
+      "step": 515
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.946303022395453e-05,
+      "loss": 0.707,
+      "step": 516
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.946046903173393e-05,
+      "loss": 0.6547,
+      "step": 517
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.375,
+      "learning_rate": 1.9457901915290604e-05,
+      "loss": 0.6898,
+      "step": 518
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5,
+      "learning_rate": 1.9455328876232104e-05,
+      "loss": 0.654,
+      "step": 519
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.945274991616969e-05,
+      "loss": 0.7088,
+      "step": 520
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.9450165036718322e-05,
+      "loss": 0.688,
+      "step": 521
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.9447574239496677e-05,
+      "loss": 0.6653,
+      "step": 522
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.9444977526127134e-05,
+      "loss": 0.6293,
+      "step": 523
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9442374898235775e-05,
+      "loss": 0.7333,
+      "step": 524
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.9439766357452386e-05,
+      "loss": 0.6583,
+      "step": 525
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.9437151905410455e-05,
+      "loss": 0.6927,
+      "step": 526
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.9434531543747176e-05,
+      "loss": 0.6873,
+      "step": 527
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.943190527410344e-05,
+      "loss": 0.7638,
+      "step": 528
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.942927309812384e-05,
+      "loss": 0.5866,
+      "step": 529
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.942663501745666e-05,
+      "loss": 0.6677,
+      "step": 530
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.9423991033753898e-05,
+      "loss": 0.7336,
+      "step": 531
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.942134114867123e-05,
+      "loss": 0.7227,
+      "step": 532
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9418685363868034e-05,
+      "loss": 0.6863,
+      "step": 533
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.941602368100739e-05,
+      "loss": 0.6726,
+      "step": 534
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.9413356101756063e-05,
+      "loss": 0.6982,
+      "step": 535
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.9410682627784514e-05,
+      "loss": 0.6818,
+      "step": 536
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.940800326076689e-05,
+      "loss": 0.6395,
+      "step": 537
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.9405318002381038e-05,
+      "loss": 0.6592,
+      "step": 538
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.940262685430849e-05,
+      "loss": 0.7127,
+      "step": 539
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.9399929818234457e-05,
+      "loss": 0.6752,
+      "step": 540
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.9397226895847848e-05,
+      "loss": 0.6898,
+      "step": 541
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9394518088841255e-05,
+      "loss": 0.6891,
+      "step": 542
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.9391803398910964e-05,
+      "loss": 0.6751,
+      "step": 543
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.938908282775692e-05,
+      "loss": 0.6874,
+      "step": 544
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.9386356377082776e-05,
+      "loss": 0.6536,
+      "step": 545
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.938362404859586e-05,
+      "loss": 0.7083,
+      "step": 546
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.938088584400717e-05,
+      "loss": 0.6711,
+      "step": 547
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.9378141765031402e-05,
+      "loss": 0.7024,
+      "step": 548
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.937539181338691e-05,
+      "loss": 0.7186,
+      "step": 549
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.9372635990795744e-05,
+      "loss": 0.7206,
+      "step": 550
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.936987429898362e-05,
+      "loss": 0.6724,
+      "step": 551
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.9367106739679933e-05,
+      "loss": 0.6858,
+      "step": 552
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.9364333314617748e-05,
+      "loss": 0.7173,
+      "step": 553
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.936155402553381e-05,
+      "loss": 0.7367,
+      "step": 554
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.935876887416853e-05,
+      "loss": 0.7017,
+      "step": 555
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.935597786226599e-05,
+      "loss": 0.6848,
+      "step": 556
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.9353180991573954e-05,
+      "loss": 0.6531,
+      "step": 557
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.9350378263843838e-05,
+      "loss": 0.7253,
+      "step": 558
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.9347569680830727e-05,
+      "loss": 0.6515,
+      "step": 559
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.934475524429339e-05,
+      "loss": 0.7284,
+      "step": 560
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.9341934955994247e-05,
+      "loss": 0.6727,
+      "step": 561
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.933910881769938e-05,
+      "loss": 0.67,
+      "step": 562
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.9336276831178548e-05,
+      "loss": 0.6957,
+      "step": 563
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.9333438998205158e-05,
+      "loss": 0.6332,
+      "step": 564
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.9330595320556286e-05,
+      "loss": 0.7029,
+      "step": 565
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.9327745800012665e-05,
+      "loss": 0.6773,
+      "step": 566
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.9324890438358688e-05,
+      "loss": 0.6665,
+      "step": 567
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.9322029237382408e-05,
+      "loss": 0.6625,
+      "step": 568
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.9319162198875532e-05,
+      "loss": 0.6904,
+      "step": 569
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.931628932463342e-05,
+      "loss": 0.6848,
+      "step": 570
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.931341061645509e-05,
+      "loss": 0.6759,
+      "step": 571
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.9310526076143216e-05,
+      "loss": 0.6596,
+      "step": 572
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.9307635705504117e-05,
+      "loss": 0.6592,
+      "step": 573
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.375,
+      "learning_rate": 1.930473950634777e-05,
+      "loss": 0.7081,
+      "step": 574
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.9301837480487794e-05,
+      "loss": 0.7487,
+      "step": 575
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.9298929629741466e-05,
+      "loss": 0.6778,
+      "step": 576
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.92960159559297e-05,
+      "loss": 0.705,
+      "step": 577
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.9293096460877066e-05,
+      "loss": 0.688,
+      "step": 578
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.375,
+      "learning_rate": 1.9290171146411777e-05,
+      "loss": 0.6173,
+      "step": 579
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.9287240014365682e-05,
+      "loss": 0.6741,
+      "step": 580
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9284303066574292e-05,
+      "loss": 0.9069,
+      "step": 581
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5,
+      "learning_rate": 1.9281360304876735e-05,
+      "loss": 0.6825,
+      "step": 582
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.92784117311158e-05,
+      "loss": 0.6501,
+      "step": 583
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.92754573471379e-05,
+      "loss": 0.6902,
+      "step": 584
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.9272497154793107e-05,
+      "loss": 0.6801,
+      "step": 585
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.9269531155935107e-05,
+      "loss": 0.6372,
+      "step": 586
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.9266559352421238e-05,
+      "loss": 0.6883,
+      "step": 587
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.9263581746112466e-05,
+      "loss": 0.677,
+      "step": 588
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.9260598338873387e-05,
+      "loss": 0.6592,
+      "step": 589
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.925760913257224e-05,
+      "loss": 0.7599,
+      "step": 590
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.9254614129080894e-05,
+      "loss": 0.6364,
+      "step": 591
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.9251613330274842e-05,
+      "loss": 0.6479,
+      "step": 592
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.375,
+      "learning_rate": 1.92486067380332e-05,
+      "loss": 0.7162,
+      "step": 593
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.375,
+      "learning_rate": 1.9245594354238734e-05,
+      "loss": 0.6958,
+      "step": 594
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.9242576180777816e-05,
+      "loss": 0.6779,
+      "step": 595
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.923955221954045e-05,
+      "loss": 0.7306,
+      "step": 596
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.9236522472420268e-05,
+      "loss": 0.677,
+      "step": 597
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.923348694131452e-05,
+      "loss": 0.6969,
+      "step": 598
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.9230445628124083e-05,
+      "loss": 0.6461,
+      "step": 599
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.922739853475345e-05,
+      "loss": 0.6443,
+      "step": 600
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.922434566311074e-05,
+      "loss": 0.6536,
+      "step": 601
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.9221287015107676e-05,
+      "loss": 0.7272,
+      "step": 602
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.9218222592659617e-05,
+      "loss": 0.6685,
+      "step": 603
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.9215152397685527e-05,
+      "loss": 0.7015,
+      "step": 604
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.921207643210799e-05,
+      "loss": 0.669,
+      "step": 605
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.920899469785319e-05,
+      "loss": 0.714,
+      "step": 606
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.920590719685095e-05,
+      "loss": 0.6384,
+      "step": 607
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5,
+      "learning_rate": 1.920281393103468e-05,
+      "loss": 0.7039,
+      "step": 608
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.9199714902341404e-05,
+      "loss": 0.7376,
+      "step": 609
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.9196610112711763e-05,
+      "loss": 0.7802,
+      "step": 610
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.375,
+      "learning_rate": 1.919349956409e-05,
+      "loss": 0.7129,
+      "step": 611
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.9190383258423974e-05,
+      "loss": 0.6516,
+      "step": 612
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.918726119766513e-05,
+      "loss": 0.6376,
+      "step": 613
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.9184133383768536e-05,
+      "loss": 0.6673,
+      "step": 614
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.918099981869285e-05,
+      "loss": 0.6809,
+      "step": 615
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.9177860504400336e-05,
+      "loss": 0.682,
+      "step": 616
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.917471544285686e-05,
+      "loss": 0.691,
+      "step": 617
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.9171564636031883e-05,
+      "loss": 0.698,
+      "step": 618
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.9168408085898468e-05,
+      "loss": 0.6837,
+      "step": 619
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9165245794433272e-05,
+      "loss": 0.8016,
+      "step": 620
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.9162077763616547e-05,
+      "loss": 0.6693,
+      "step": 621
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.915890399543214e-05,
+      "loss": 0.6047,
+      "step": 622
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.9155724491867488e-05,
+      "loss": 0.6795,
+      "step": 623
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.9152539254913627e-05,
+      "loss": 0.7559,
+      "step": 624
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.9149348286565176e-05,
+      "loss": 0.6602,
+      "step": 625
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.9146151588820343e-05,
+      "loss": 0.7103,
+      "step": 626
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.914294916368093e-05,
+      "loss": 0.6732,
+      "step": 627
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.913974101315232e-05,
+      "loss": 0.6626,
+      "step": 628
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.9136527139243485e-05,
+      "loss": 0.6681,
+      "step": 629
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.9133307543966976e-05,
+      "loss": 0.6784,
+      "step": 630
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9130082229338927e-05,
+      "loss": 0.6769,
+      "step": 631
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.912685119737906e-05,
+      "loss": 0.7364,
+      "step": 632
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.9123614450110678e-05,
+      "loss": 0.7017,
+      "step": 633
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.9120371989560652e-05,
+      "loss": 0.6405,
+      "step": 634
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.9117123817759438e-05,
+      "loss": 0.7171,
+      "step": 635
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.911386993674107e-05,
+      "loss": 0.6709,
+      "step": 636
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.9110610348543152e-05,
+      "loss": 0.6238,
+      "step": 637
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.910734505520687e-05,
+      "loss": 0.6441,
+      "step": 638
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.910407405877697e-05,
+      "loss": 0.6726,
+      "step": 639
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.910079736130178e-05,
+      "loss": 0.689,
+      "step": 640
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5,
+      "learning_rate": 1.9097514964833193e-05,
+      "loss": 0.7076,
+      "step": 641
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.9094226871426678e-05,
+      "loss": 0.6684,
+      "step": 642
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.909093308314126e-05,
+      "loss": 0.6564,
+      "step": 643
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.9087633602039543e-05,
+      "loss": 0.688,
+      "step": 644
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.9084328430187677e-05,
+      "loss": 0.6637,
+      "step": 645
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.90810175696554e-05,
+      "loss": 0.6144,
+      "step": 646
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.907770102251599e-05,
+      "loss": 0.6679,
+      "step": 647
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.9074378790846306e-05,
+      "loss": 0.7067,
+      "step": 648
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.9071050876726742e-05,
+      "loss": 0.6936,
+      "step": 649
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.906771728224128e-05,
+      "loss": 0.6143,
+      "step": 650
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.9064378009477434e-05,
+      "loss": 0.7354,
+      "step": 651
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5,
+      "learning_rate": 1.906103306052629e-05,
+      "loss": 0.7079,
+      "step": 652
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.375,
+      "learning_rate": 1.9057682437482478e-05,
+      "loss": 0.7084,
+      "step": 653
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.905432614244419e-05,
+      "loss": 0.7023,
+      "step": 654
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.9050964177513158e-05,
+      "loss": 0.6889,
+      "step": 655
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.9047596544794678e-05,
+      "loss": 0.742,
+      "step": 656
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.9044223246397587e-05,
+      "loss": 0.7448,
+      "step": 657
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.9040844284434273e-05,
+      "loss": 0.6538,
+      "step": 658
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.9037459661020666e-05,
+      "loss": 0.6788,
+      "step": 659
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.903406937827625e-05,
+      "loss": 0.6525,
+      "step": 660
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.9030673438324046e-05,
+      "loss": 0.638,
+      "step": 661
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.9027271843290614e-05,
+      "loss": 0.6232,
+      "step": 662
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.902386459530607e-05,
+      "loss": 0.6323,
+      "step": 663
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.9020451696504047e-05,
+      "loss": 0.7019,
+      "step": 664
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.9017033149021747e-05,
+      "loss": 0.6337,
+      "step": 665
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.9013608954999875e-05,
+      "loss": 0.6227,
+      "step": 666
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.9010179116582703e-05,
+      "loss": 0.6871,
+      "step": 667
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.9006743635918018e-05,
+      "loss": 0.6916,
+      "step": 668
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.9003302515157145e-05,
+      "loss": 0.6523,
+      "step": 669
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.8999855756454945e-05,
+      "loss": 0.6294,
+      "step": 670
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.89964033619698e-05,
+      "loss": 0.686,
+      "step": 671
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8992945333863636e-05,
+      "loss": 0.7315,
+      "step": 672
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.8989481674301892e-05,
+      "loss": 0.6741,
+      "step": 673
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.8986012385453544e-05,
+      "loss": 0.7291,
+      "step": 674
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.898253746949109e-05,
+      "loss": 0.7723,
+      "step": 675
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.8979056928590543e-05,
+      "loss": 0.626,
+      "step": 676
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.8975570764931453e-05,
+      "loss": 0.692,
+      "step": 677
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.8972078980696878e-05,
+      "loss": 0.6224,
+      "step": 678
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.896858157807341e-05,
+      "loss": 0.7064,
+      "step": 679
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.8965078559251144e-05,
+      "loss": 0.6836,
+      "step": 680
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.8961569926423702e-05,
+      "loss": 0.7078,
+      "step": 681
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.895805568178822e-05,
+      "loss": 0.65,
+      "step": 682
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.895453582754534e-05,
+      "loss": 0.7279,
+      "step": 683
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.8951010365899232e-05,
+      "loss": 0.6511,
+      "step": 684
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.8947479299057562e-05,
+      "loss": 0.7021,
+      "step": 685
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.375,
+      "learning_rate": 1.8943942629231518e-05,
+      "loss": 0.6927,
+      "step": 686
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.894040035863579e-05,
+      "loss": 0.6743,
+      "step": 687
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.893685248948857e-05,
+      "loss": 0.6512,
+      "step": 688
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.8933299024011572e-05,
+      "loss": 0.6958,
+      "step": 689
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.8929739964430002e-05,
+      "loss": 0.629,
+      "step": 690
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.892617531297257e-05,
+      "loss": 0.7434,
+      "step": 691
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.892260507187149e-05,
+      "loss": 0.6381,
+      "step": 692
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.891902924336248e-05,
+      "loss": 0.7009,
+      "step": 693
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.8915447829684744e-05,
+      "loss": 0.6487,
+      "step": 694
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.8911860833081e-05,
+      "loss": 0.6745,
+      "step": 695
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.8908268255797457e-05,
+      "loss": 0.6366,
+      "step": 696
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.375,
+      "learning_rate": 1.890467010008381e-05,
+      "loss": 0.7627,
+      "step": 697
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.8901066368193253e-05,
+      "loss": 0.6587,
+      "step": 698
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.8897457062382475e-05,
+      "loss": 0.7136,
+      "step": 699
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.8893842184911656e-05,
+      "loss": 0.6332,
+      "step": 700
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.8890221738044457e-05,
+      "loss": 0.6233,
+      "step": 701
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.888659572404803e-05,
+      "loss": 0.6567,
+      "step": 702
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.8882964145193018e-05,
+      "loss": 0.5958,
+      "step": 703
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.8879327003753546e-05,
+      "loss": 0.6635,
+      "step": 704
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8875684302007215e-05,
+      "loss": 0.7063,
+      "step": 705
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.8872036042235123e-05,
+      "loss": 0.6718,
+      "step": 706
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.8868382226721834e-05,
+      "loss": 0.6325,
+      "step": 707
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.8864722857755398e-05,
+      "loss": 0.6422,
+      "step": 708
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.8861057937627346e-05,
+      "loss": 0.7121,
+      "step": 709
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.8857387468632675e-05,
+      "loss": 0.6168,
+      "step": 710
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.885371145306987e-05,
+      "loss": 0.7611,
+      "step": 711
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.8850029893240868e-05,
+      "loss": 0.6164,
+      "step": 712
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.8846342791451107e-05,
+      "loss": 0.6576,
+      "step": 713
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.8842650150009472e-05,
+      "loss": 0.6822,
+      "step": 714
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.8838951971228326e-05,
+      "loss": 0.7326,
+      "step": 715
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.375,
+      "learning_rate": 1.88352482574235e-05,
+      "loss": 0.6495,
+      "step": 716
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.883153901091429e-05,
+      "loss": 0.6389,
+      "step": 717
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0,
+      "learning_rate": 1.882782423402346e-05,
+      "loss": 0.6748,
+      "step": 718
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.8824103929077222e-05,
+      "loss": 0.6719,
+      "step": 719
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.8820378098405272e-05,
+      "loss": 0.7211,
+      "step": 720
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.881664674434075e-05,
+      "loss": 0.731,
+      "step": 721
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.881290986922027e-05,
+      "loss": 0.6661,
+      "step": 722
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.8809167475383878e-05,
+      "loss": 0.7045,
+      "step": 723
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.88054195651751e-05,
+      "loss": 0.636,
+      "step": 724
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.880166614094091e-05,
+      "loss": 0.6409,
+      "step": 725
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.8797907205031725e-05,
+      "loss": 0.6186,
+      "step": 726
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.8794142759801434e-05,
+      "loss": 0.6688,
+      "step": 727
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.879037280760735e-05,
+      "loss": 0.683,
+      "step": 728
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.8786597350810255e-05,
+      "loss": 0.661,
+      "step": 729
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.878281639177437e-05,
+      "loss": 0.6922,
+      "step": 730
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.877902993286736e-05,
+      "loss": 0.6334,
+      "step": 731
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.8775237976460338e-05,
+      "loss": 0.7089,
+      "step": 732
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.877144052492786e-05,
+      "loss": 0.6506,
+      "step": 733
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.876763758064792e-05,
+      "loss": 0.6641,
+      "step": 734
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.876382914600195e-05,
+      "loss": 0.6545,
+      "step": 735
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.876001522337482e-05,
+      "loss": 0.6409,
+      "step": 736
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.8756195815154847e-05,
+      "loss": 0.681,
+      "step": 737
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.8752370923733774e-05,
+      "loss": 0.6736,
+      "step": 738
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.8748540551506775e-05,
+      "loss": 0.667,
+      "step": 739
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.8744704700872463e-05,
+      "loss": 0.6428,
+      "step": 740
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.874086337423288e-05,
+      "loss": 0.7186,
+      "step": 741
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.873701657399349e-05,
+      "loss": 0.6783,
+      "step": 742
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.87331643025632e-05,
+      "loss": 0.6737,
+      "step": 743
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.872930656235433e-05,
+      "loss": 0.6906,
+      "step": 744
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.8725443355782623e-05,
+      "loss": 0.6674,
+      "step": 745
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.8721574685267255e-05,
+      "loss": 0.6711,
+      "step": 746
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.8717700553230823e-05,
+      "loss": 0.6358,
+      "step": 747
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.8713820962099332e-05,
+      "loss": 0.6432,
+      "step": 748
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.8709935914302218e-05,
+      "loss": 0.7078,
+      "step": 749
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.870604541227233e-05,
+      "loss": 0.6535,
+      "step": 750
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.870214945844593e-05,
+      "loss": 0.6899,
+      "step": 751
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.8698248055262697e-05,
+      "loss": 0.6763,
+      "step": 752
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.869434120516572e-05,
+      "loss": 0.6405,
+      "step": 753
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.8690428910601503e-05,
+      "loss": 0.635,
+      "step": 754
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.8686511174019956e-05,
+      "loss": 0.6483,
+      "step": 755
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.8682587997874395e-05,
+      "loss": 0.682,
+      "step": 756
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8678659384621546e-05,
+      "loss": 0.692,
+      "step": 757
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.867472533672154e-05,
+      "loss": 0.6639,
+      "step": 758
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.25,
+      "learning_rate": 1.867078585663791e-05,
+      "loss": 0.6257,
+      "step": 759
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.866684094683759e-05,
+      "loss": 0.6916,
+      "step": 760
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.8662890609790914e-05,
+      "loss": 0.6968,
+      "step": 761
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.8658934847971614e-05,
+      "loss": 0.6773,
+      "step": 762
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.8654973663856824e-05,
+      "loss": 0.6365,
+      "step": 763
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.8651007059927063e-05,
+      "loss": 0.6662,
+      "step": 764
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.864703503866626e-05,
+      "loss": 0.671,
+      "step": 765
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.864305760256172e-05,
+      "loss": 0.6275,
+      "step": 766
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.863907475410415e-05,
+      "loss": 0.6182,
+      "step": 767
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.8635086495787638e-05,
+      "loss": 0.7348,
+      "step": 768
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.8631092830109666e-05,
+      "loss": 0.6397,
+      "step": 769
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.86270937595711e-05,
+      "loss": 0.7229,
+      "step": 770
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.862308928667619e-05,
+      "loss": 0.6945,
+      "step": 771
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.8619079413932572e-05,
+      "loss": 0.7198,
+      "step": 772
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.861506414385126e-05,
+      "loss": 0.681,
+      "step": 773
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.861104347894664e-05,
+      "loss": 0.7122,
+      "step": 774
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.86070174217365e-05,
+      "loss": 0.6319,
+      "step": 775
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.8602985974741982e-05,
+      "loss": 0.6205,
+      "step": 776
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.859894914048761e-05,
+      "loss": 0.7082,
+      "step": 777
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.8594906921501286e-05,
+      "loss": 0.7431,
+      "step": 778
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.8590859320314277e-05,
+      "loss": 0.6885,
+      "step": 779
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8586806339461226e-05,
+      "loss": 0.7745,
+      "step": 780
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.8582747981480142e-05,
+      "loss": 0.6615,
+      "step": 781
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.85786842489124e-05,
+      "loss": 0.6157,
+      "step": 782
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.8574615144302746e-05,
+      "loss": 0.6814,
+      "step": 783
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.8570540670199282e-05,
+      "loss": 0.6595,
+      "step": 784
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.8566460829153484e-05,
+      "loss": 0.6032,
+      "step": 785
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.8562375623720173e-05,
+      "loss": 0.6647,
+      "step": 786
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.375,
+      "learning_rate": 1.8558285056457545e-05,
+      "loss": 0.6909,
+      "step": 787
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.855418912992714e-05,
+      "loss": 0.6754,
+      "step": 788
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.8550087846693868e-05,
+      "loss": 0.707,
+      "step": 789
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.8545981209325975e-05,
+      "loss": 0.754,
+      "step": 790
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.8541869220395084e-05,
+      "loss": 0.6882,
+      "step": 791
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.853775188247615e-05,
+      "loss": 0.6214,
+      "step": 792
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.8533629198147482e-05,
+      "loss": 0.73,
+      "step": 793
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.852950116999074e-05,
+      "loss": 0.6478,
+      "step": 794
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.8525367800590927e-05,
+      "loss": 0.6617,
+      "step": 795
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.8521229092536395e-05,
+      "loss": 0.6496,
+      "step": 796
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.8517085048418834e-05,
+      "loss": 0.6174,
+      "step": 797
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.851293567083328e-05,
+      "loss": 0.7401,
+      "step": 798
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.850878096237811e-05,
+      "loss": 0.7122,
+      "step": 799
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.8504620925655034e-05,
+      "loss": 0.6366,
+      "step": 800
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.8500455563269097e-05,
+      "loss": 0.6669,
+      "step": 801
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.8496284877828684e-05,
+      "loss": 0.6758,
+      "step": 802
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.849210887194552e-05,
+      "loss": 0.7141,
+      "step": 803
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.8487927548234642e-05,
+      "loss": 0.6676,
+      "step": 804
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5,
+      "learning_rate": 1.848374090931444e-05,
+      "loss": 0.7168,
+      "step": 805
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.375,
+      "learning_rate": 1.8479548957806615e-05,
+      "loss": 0.7071,
+      "step": 806
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.84753516963362e-05,
+      "loss": 0.6478,
+      "step": 807
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.8471149127531558e-05,
+      "loss": 0.6565,
+      "step": 808
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.8466941254024374e-05,
+      "loss": 0.6809,
+      "step": 809
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.8462728078449642e-05,
+      "loss": 0.6574,
+      "step": 810
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.84585096034457e-05,
+      "loss": 0.7106,
+      "step": 811
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.8454285831654183e-05,
+      "loss": 0.6703,
+      "step": 812
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.8450056765720055e-05,
+      "loss": 0.6653,
+      "step": 813
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.844582240829159e-05,
+      "loss": 0.6862,
+      "step": 814
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.8441582762020374e-05,
+      "loss": 0.7085,
+      "step": 815
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.8437337829561316e-05,
+      "loss": 0.6421,
+      "step": 816
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.843308761357262e-05,
+      "loss": 0.7253,
+      "step": 817
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.842883211671581e-05,
+      "loss": 0.6889,
+      "step": 818
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.8424571341655713e-05,
+      "loss": 0.6572,
+      "step": 819
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.8420305291060457e-05,
+      "loss": 0.625,
+      "step": 820
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.8416033967601475e-05,
+      "loss": 0.5988,
+      "step": 821
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.8411757373953518e-05,
+      "loss": 0.6983,
+      "step": 822
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.840747551279461e-05,
+      "loss": 0.6656,
+      "step": 823
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.8403188386806094e-05,
+      "loss": 0.6731,
+      "step": 824
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.83988959986726e-05,
+      "loss": 0.6832,
+      "step": 825
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.839459835108206e-05,
+      "loss": 0.7544,
+      "step": 826
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.8390295446725693e-05,
+      "loss": 0.6281,
+      "step": 827
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.8385987288298015e-05,
+      "loss": 0.6207,
+      "step": 828
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.8381673878496826e-05,
+      "loss": 0.6604,
+      "step": 829
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.8377355220023223e-05,
+      "loss": 0.7329,
+      "step": 830
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.8373031315581582e-05,
+      "loss": 0.7013,
+      "step": 831
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.836870216787957e-05,
+      "loss": 0.65,
+      "step": 832
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.836436777962813e-05,
+      "loss": 0.7227,
+      "step": 833
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.8360028153541497e-05,
+      "loss": 0.7065,
+      "step": 834
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.8355683292337174e-05,
+      "loss": 0.6423,
+      "step": 835
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.835133319873595e-05,
+      "loss": 0.5816,
+      "step": 836
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.8346977875461897e-05,
+      "loss": 0.6854,
+      "step": 837
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.8342617325242345e-05,
+      "loss": 0.6197,
+      "step": 838
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.8338251550807907e-05,
+      "loss": 0.6245,
+      "step": 839
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.8333880554892466e-05,
+      "loss": 0.6981,
+      "step": 840
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.832950434023318e-05,
+      "loss": 0.656,
+      "step": 841
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.832512290957047e-05,
+      "loss": 0.6668,
+      "step": 842
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5,
+      "learning_rate": 1.8320736265648015e-05,
+      "loss": 0.6572,
+      "step": 843
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.8316344411212778e-05,
+      "loss": 0.7103,
+      "step": 844
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.8311947349014968e-05,
+      "loss": 0.6388,
+      "step": 845
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.8307545081808062e-05,
+      "loss": 0.6642,
+      "step": 846
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.8303137612348794e-05,
+      "loss": 0.6577,
+      "step": 847
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.829872494339716e-05,
+      "loss": 0.6297,
+      "step": 848
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.829430707771641e-05,
+      "loss": 0.6746,
+      "step": 849
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.8289884018073042e-05,
+      "loss": 0.6153,
+      "step": 850
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.828545576723682e-05,
+      "loss": 0.7378,
+      "step": 851
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.8281022327980746e-05,
+      "loss": 0.6824,
+      "step": 852
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.8276583703081074e-05,
+      "loss": 0.6733,
+      "step": 853
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.8272139895317312e-05,
+      "loss": 0.6475,
+      "step": 854
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.82676909074722e-05,
+      "loss": 0.7045,
+      "step": 855
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.8263236742331743e-05,
+      "loss": 0.7122,
+      "step": 856
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.8258777402685164e-05,
+      "loss": 0.7111,
+      "step": 857
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.825431289132495e-05,
+      "loss": 0.6626,
+      "step": 858
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.82498432110468e-05,
+      "loss": 0.6753,
+      "step": 859
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.8245368364649675e-05,
+      "loss": 0.6951,
+      "step": 860
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.824088835493575e-05,
+      "loss": 0.6842,
+      "step": 861
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.8236403184710454e-05,
+      "loss": 0.656,
+      "step": 862
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.25,
+      "learning_rate": 1.8231912856782436e-05,
+      "loss": 0.6384,
+      "step": 863
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.8227417373963566e-05,
+      "loss": 0.6767,
+      "step": 864
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.8222916739068964e-05,
+      "loss": 0.6779,
+      "step": 865
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.8218410954916958e-05,
+      "loss": 0.5751,
+      "step": 866
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.821390002432911e-05,
+      "loss": 0.6552,
+      "step": 867
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.820938395013019e-05,
+      "loss": 0.621,
+      "step": 868
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.8204862735148218e-05,
+      "loss": 0.7557,
+      "step": 869
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.8200336382214406e-05,
+      "loss": 0.6355,
+      "step": 870
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.8195804894163196e-05,
+      "loss": 0.6854,
+      "step": 871
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.819126827383224e-05,
+      "loss": 0.6686,
+      "step": 872
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.8186726524062406e-05,
+      "loss": 0.6922,
+      "step": 873
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.375,
+      "learning_rate": 1.8182179647697776e-05,
+      "loss": 0.6595,
+      "step": 874
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.8177627647585644e-05,
+      "loss": 0.6878,
+      "step": 875
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.8173070526576502e-05,
+      "loss": 0.6353,
+      "step": 876
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.8168508287524066e-05,
+      "loss": 0.6105,
+      "step": 877
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.816394093328524e-05,
+      "loss": 0.6372,
+      "step": 878
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.375,
+      "learning_rate": 1.8159368466720138e-05,
+      "loss": 0.7103,
+      "step": 879
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.25,
+      "learning_rate": 1.8154790890692082e-05,
+      "loss": 0.5986,
+      "step": 880
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.8150208208067582e-05,
+      "loss": 0.6496,
+      "step": 881
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.8145620421716355e-05,
+      "loss": 0.6398,
+      "step": 882
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.8141027534511307e-05,
+      "loss": 0.6809,
+      "step": 883
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.8136429549328545e-05,
+      "loss": 0.6503,
+      "step": 884
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.813182646904736e-05,
+      "loss": 0.6661,
+      "step": 885
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.375,
+      "learning_rate": 1.812721829655024e-05,
+      "loss": 0.7096,
+      "step": 886
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.812260503472286e-05,
+      "loss": 0.6596,
+      "step": 887
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.375,
+      "learning_rate": 1.8117986686454084e-05,
+      "loss": 0.7117,
+      "step": 888
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.811336325463596e-05,
+      "loss": 0.7073,
+      "step": 889
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.8108734742163717e-05,
+      "loss": 0.6234,
+      "step": 890
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.8104101151935764e-05,
+      "loss": 0.6809,
+      "step": 891
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.80994624868537e-05,
+      "loss": 0.6208,
+      "step": 892
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.8094818749822287e-05,
+      "loss": 0.6418,
+      "step": 893
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.8090169943749477e-05,
+      "loss": 0.6708,
+      "step": 894
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.8085516071546385e-05,
+      "loss": 0.7215,
+      "step": 895
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.808085713612731e-05,
+      "loss": 0.7019,
+      "step": 896
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.807619314040971e-05,
+      "loss": 0.6903,
+      "step": 897
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.807152408731422e-05,
+      "loss": 0.7761,
+      "step": 898
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.8066849979764635e-05,
+      "loss": 0.6796,
+      "step": 899
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.8062170820687925e-05,
+      "loss": 0.6512,
+      "step": 900
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.8057486613014217e-05,
+      "loss": 0.6431,
+      "step": 901
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.8052797359676792e-05,
+      "loss": 0.6387,
+      "step": 902
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.804810306361211e-05,
+      "loss": 0.6367,
+      "step": 903
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.8043403727759767e-05,
+      "loss": 0.628,
+      "step": 904
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.803869935506253e-05,
+      "loss": 0.6559,
+      "step": 905
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.803398994846632e-05,
+      "loss": 0.6911,
+      "step": 906
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.80292755109202e-05,
+      "loss": 0.6205,
+      "step": 907
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.375,
+      "learning_rate": 1.8024556045376388e-05,
+      "loss": 0.67,
+      "step": 908
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.8019831554790258e-05,
+      "loss": 0.6781,
+      "step": 909
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.8015102042120314e-05,
+      "loss": 0.699,
+      "step": 910
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.801036751032823e-05,
+      "loss": 0.6396,
+      "step": 911
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.8005627962378797e-05,
+      "loss": 0.639,
+      "step": 912
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.800088340123996e-05,
+      "loss": 0.6635,
+      "step": 913
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.7996133829882806e-05,
+      "loss": 0.6905,
+      "step": 914
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.799137925128155e-05,
+      "loss": 0.6831,
+      "step": 915
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.7986619668413553e-05,
+      "loss": 0.6689,
+      "step": 916
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.7981855084259302e-05,
+      "loss": 0.6804,
+      "step": 917
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.797708550180242e-05,
+      "loss": 0.6556,
+      "step": 918
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.7972310924029653e-05,
+      "loss": 0.6301,
+      "step": 919
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.7967531353930893e-05,
+      "loss": 0.6533,
+      "step": 920
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.796274679449913e-05,
+      "loss": 0.7503,
+      "step": 921
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.7957957248730502e-05,
+      "loss": 0.6729,
+      "step": 922
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.7953162719624263e-05,
+      "loss": 0.7021,
+      "step": 923
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.794836321018278e-05,
+      "loss": 0.6915,
+      "step": 924
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.7943558723411555e-05,
+      "loss": 0.6827,
+      "step": 925
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.7938749262319187e-05,
+      "loss": 0.6628,
+      "step": 926
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.7933934829917404e-05,
+      "loss": 0.6339,
+      "step": 927
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.792911542922104e-05,
+      "loss": 0.7272,
+      "step": 928
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.7924291063248048e-05,
+      "loss": 0.6178,
+      "step": 929
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.791946173501948e-05,
+      "loss": 0.6318,
+      "step": 930
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.7914627447559504e-05,
+      "loss": 0.7219,
+      "step": 931
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.7909788203895392e-05,
+      "loss": 0.6476,
+      "step": 932
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.790494400705751e-05,
+      "loss": 0.7274,
+      "step": 933
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.790009486007934e-05,
+      "loss": 0.6759,
+      "step": 934
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.7895240765997455e-05,
+      "loss": 0.7109,
+      "step": 935
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.789038172785153e-05,
+      "loss": 0.5928,
+      "step": 936
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5,
+      "learning_rate": 1.7885517748684335e-05,
+      "loss": 0.7083,
+      "step": 937
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.788064883154173e-05,
+      "loss": 0.6736,
+      "step": 938
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.787577497947267e-05,
+      "loss": 0.6915,
+      "step": 939
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.7870896195529205e-05,
+      "loss": 0.6734,
+      "step": 940
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.7866012482766466e-05,
+      "loss": 0.6978,
+      "step": 941
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.7861123844242675e-05,
+      "loss": 0.6486,
+      "step": 942
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.7856230283019138e-05,
+      "loss": 0.6665,
+      "step": 943
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.785133180216024e-05,
+      "loss": 0.6918,
+      "step": 944
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.7846428404733456e-05,
+      "loss": 0.6142,
+      "step": 945
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.7841520093809322e-05,
+      "loss": 0.659,
+      "step": 946
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.7836606872461474e-05,
+      "loss": 0.6012,
+      "step": 947
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.7831688743766605e-05,
+      "loss": 0.6533,
+      "step": 948
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.782676571080449e-05,
+      "loss": 0.6839,
+      "step": 949
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.7821837776657968e-05,
+      "loss": 0.6569,
+      "step": 950
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.7816904944412957e-05,
+      "loss": 0.705,
+      "step": 951
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.7811967217158432e-05,
+      "loss": 0.7441,
+      "step": 952
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.7807024597986443e-05,
+      "loss": 0.6413,
+      "step": 953
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.7802077089992095e-05,
+      "loss": 0.6798,
+      "step": 954
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.7797124696273553e-05,
+      "loss": 0.6427,
+      "step": 955
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.7792167419932056e-05,
+      "loss": 0.6389,
+      "step": 956
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.7787205264071888e-05,
+      "loss": 0.7152,
+      "step": 957
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.7782238231800386e-05,
+      "loss": 0.6814,
+      "step": 958
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.777726632622795e-05,
+      "loss": 0.6888,
+      "step": 959
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.7772289550468033e-05,
+      "loss": 0.714,
+      "step": 960
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.7767307907637122e-05,
+      "loss": 0.7151,
+      "step": 961
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.7762321400854766e-05,
+      "loss": 0.6565,
+      "step": 962
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.775733003324356e-05,
+      "loss": 0.6462,
+      "step": 963
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.7752333807929136e-05,
+      "loss": 0.6844,
+      "step": 964
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.774733272804017e-05,
+      "loss": 0.6613,
+      "step": 965
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.774232679670838e-05,
+      "loss": 0.5949,
+      "step": 966
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.7737316017068528e-05,
+      "loss": 0.6513,
+      "step": 967
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.773230039225839e-05,
+      "loss": 0.693,
+      "step": 968
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.77272799254188e-05,
+      "loss": 0.6771,
+      "step": 969
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.772225461969362e-05,
+      "loss": 0.6092,
+      "step": 970
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.771722447822973e-05,
+      "loss": 0.6515,
+      "step": 971
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.7712189504177046e-05,
+      "loss": 0.6459,
+      "step": 972
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.770714970068851e-05,
+      "loss": 0.6673,
+      "step": 973
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.770210507092009e-05,
+      "loss": 0.6924,
+      "step": 974
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.7697055618030777e-05,
+      "loss": 0.6821,
+      "step": 975
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.769200134518257e-05,
+      "loss": 0.7186,
+      "step": 976
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.7686942255540506e-05,
+      "loss": 0.685,
+      "step": 977
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.7681878352272625e-05,
+      "loss": 0.6842,
+      "step": 978
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.767680963854998e-05,
+      "loss": 0.6301,
+      "step": 979
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.7671736117546646e-05,
+      "loss": 0.6639,
+      "step": 980
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.7666657792439702e-05,
+      "loss": 0.7159,
+      "step": 981
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.7661574666409232e-05,
+      "loss": 0.6312,
+      "step": 982
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.765648674263834e-05,
+      "loss": 0.6469,
+      "step": 983
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.7651394024313114e-05,
+      "loss": 0.6544,
+      "step": 984
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.7646296514622666e-05,
+      "loss": 0.6428,
+      "step": 985
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.7641194216759096e-05,
+      "loss": 0.6876,
+      "step": 986
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.7636087133917494e-05,
+      "loss": 0.6739,
+      "step": 987
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.763097526929597e-05,
+      "loss": 0.6704,
+      "step": 988
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.762585862609561e-05,
+      "loss": 0.7166,
+      "step": 989
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.76207372075205e-05,
+      "loss": 0.6969,
+      "step": 990
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.7615611016777708e-05,
+      "loss": 0.6087,
+      "step": 991
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.76104800570773e-05,
+      "loss": 0.6945,
+      "step": 992
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.7605344331632328e-05,
+      "loss": 0.6944,
+      "step": 993
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.760020384365882e-05,
+      "loss": 0.6665,
+      "step": 994
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.75950585963758e-05,
+      "loss": 0.7028,
+      "step": 995
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.758990859300525e-05,
+      "loss": 0.7026,
+      "step": 996
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.7584753836772152e-05,
+      "loss": 0.6685,
+      "step": 997
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.7579594330904456e-05,
+      "loss": 0.6653,
+      "step": 998
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.25,
+      "learning_rate": 1.757443007863309e-05,
+      "loss": 0.657,
+      "step": 999
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.7569261083191942e-05,
+      "loss": 0.6941,
+      "step": 1000
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.7564087347817886e-05,
+      "loss": 0.644,
+      "step": 1001
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.7558908875750753e-05,
+      "loss": 0.6866,
+      "step": 1002
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.755372567023335e-05,
+      "loss": 0.6703,
+      "step": 1003
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.7548537734511438e-05,
+      "loss": 0.6246,
+      "step": 1004
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.7543345071833745e-05,
+      "loss": 0.6515,
+      "step": 1005
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.753814768545196e-05,
+      "loss": 0.674,
+      "step": 1006
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.7532945578620725e-05,
+      "loss": 0.7032,
+      "step": 1007
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.7527738754597647e-05,
+      "loss": 0.6848,
+      "step": 1008
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.752252721664328e-05,
+      "loss": 0.7007,
+      "step": 1009
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.751731096802113e-05,
+      "loss": 0.6983,
+      "step": 1010
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.7512090011997656e-05,
+      "loss": 0.6563,
+      "step": 1011
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.750686435184226e-05,
+      "loss": 0.7025,
+      "step": 1012
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.7501633990827296e-05,
+      "loss": 0.6883,
+      "step": 1013
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.749639893222806e-05,
+      "loss": 0.6942,
+      "step": 1014
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.7491159179322785e-05,
+      "loss": 0.717,
+      "step": 1015
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.7485914735392646e-05,
+      "loss": 0.6559,
+      "step": 1016
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.748066560372176e-05,
+      "loss": 0.6632,
+      "step": 1017
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.7475411787597173e-05,
+      "loss": 0.685,
+      "step": 1018
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 0.6667066216468811,
+      "eval_runtime": 568.8454,
+      "eval_samples_per_second": 25.411,
+      "eval_steps_per_second": 25.411,
+      "step": 1018
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.747015329030887e-05,
+      "loss": 0.6581,
+      "step": 1019
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.746489011514976e-05,
+      "loss": 0.6694,
+      "step": 1020
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.7459622265415695e-05,
+      "loss": 0.6706,
+      "step": 1021
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.745434974440544e-05,
+      "loss": 0.6321,
+      "step": 1022
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.7449072555420688e-05,
+      "loss": 0.7313,
+      "step": 1023
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.7443790701766065e-05,
+      "loss": 0.6392,
+      "step": 1024
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.7438504186749105e-05,
+      "loss": 0.655,
+      "step": 1025
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.7433213013680266e-05,
+      "loss": 0.6308,
+      "step": 1026
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.7427917185872923e-05,
+      "loss": 0.659,
+      "step": 1027
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.7422616706643374e-05,
+      "loss": 0.6855,
+      "step": 1028
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.7417311579310817e-05,
+      "loss": 0.692,
+      "step": 1029
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.7412001807197362e-05,
+      "loss": 0.6695,
+      "step": 1030
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.740668739362804e-05,
+      "loss": 0.6312,
+      "step": 1031
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.7401368341930777e-05,
+      "loss": 0.7427,
+      "step": 1032
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.7396044655436403e-05,
+      "loss": 0.6648,
+      "step": 1033
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.7390716337478655e-05,
+      "loss": 0.6557,
+      "step": 1034
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.7385383391394174e-05,
+      "loss": 0.6871,
+      "step": 1035
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.7380045820522486e-05,
+      "loss": 0.6613,
+      "step": 1036
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.7374703628206027e-05,
+      "loss": 0.6634,
+      "step": 1037
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.736935681779012e-05,
+      "loss": 0.7328,
+      "step": 1038
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.7364005392622982e-05,
+      "loss": 0.675,
+      "step": 1039
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.735864935605572e-05,
+      "loss": 0.69,
+      "step": 1040
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.7353288711442325e-05,
+      "loss": 0.641,
+      "step": 1041
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.734792346213968e-05,
+      "loss": 0.6668,
+      "step": 1042
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.734255361150754e-05,
+      "loss": 0.6936,
+      "step": 1043
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.7337179162908556e-05,
+      "loss": 0.5925,
+      "step": 1044
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.733180011970825e-05,
+      "loss": 0.6607,
+      "step": 1045
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.7326416485275023e-05,
+      "loss": 0.6999,
+      "step": 1046
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.732102826298015e-05,
+      "loss": 0.6503,
+      "step": 1047
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.7315635456197783e-05,
+      "loss": 0.6339,
+      "step": 1048
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.7310238068304933e-05,
+      "loss": 0.6669,
+      "step": 1049
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.7304836102681494e-05,
+      "loss": 0.7431,
+      "step": 1050
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.7299429562710223e-05,
+      "loss": 0.6752,
+      "step": 1051
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.729401845177674e-05,
+      "loss": 0.6387,
+      "step": 1052
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.7288602773269518e-05,
+      "loss": 0.6998,
+      "step": 1053
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.7283182530579905e-05,
+      "loss": 0.6795,
+      "step": 1054
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.72777577271021e-05,
+      "loss": 0.6867,
+      "step": 1055
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.7272328366233164e-05,
+      "loss": 0.5977,
+      "step": 1056
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.7266894451373e-05,
+      "loss": 0.6419,
+      "step": 1057
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.7261455985924373e-05,
+      "loss": 0.6178,
+      "step": 1058
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.725601297329289e-05,
+      "loss": 0.5874,
+      "step": 1059
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.7250565416887016e-05,
+      "loss": 0.7233,
+      "step": 1060
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.7245113320118055e-05,
+      "loss": 0.7044,
+      "step": 1061
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.723965668640015e-05,
+      "loss": 0.6675,
+      "step": 1062
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.7234195519150293e-05,
+      "loss": 0.6479,
+      "step": 1063
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.7228729821788305e-05,
+      "loss": 0.7211,
+      "step": 1064
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.7223259597736863e-05,
+      "loss": 0.741,
+      "step": 1065
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.721778485042145e-05,
+      "loss": 0.6295,
+      "step": 1066
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.721230558327041e-05,
+      "loss": 0.6592,
+      "step": 1067
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.72068217997149e-05,
+      "loss": 0.6482,
+      "step": 1068
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.7201333503188905e-05,
+      "loss": 0.7414,
+      "step": 1069
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.7195840697129252e-05,
+      "loss": 0.7175,
+      "step": 1070
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.719034338497557e-05,
+      "loss": 0.7125,
+      "step": 1071
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.718484157017033e-05,
+      "loss": 0.6797,
+      "step": 1072
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.7179335256158805e-05,
+      "loss": 0.6708,
+      "step": 1073
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.7173824446389098e-05,
+      "loss": 0.6822,
+      "step": 1074
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.716830914431212e-05,
+      "loss": 0.665,
+      "step": 1075
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.7162789353381604e-05,
+      "loss": 0.619,
+      "step": 1076
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.7157265077054083e-05,
+      "loss": 0.6572,
+      "step": 1077
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.7151736318788902e-05,
+      "loss": 0.6542,
+      "step": 1078
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.25,
+      "learning_rate": 1.7146203082048223e-05,
+      "loss": 0.648,
+      "step": 1079
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.7140665370296995e-05,
+      "loss": 0.6667,
+      "step": 1080
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.713512318700298e-05,
+      "loss": 0.6957,
+      "step": 1081
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.375,
+      "learning_rate": 1.712957653563674e-05,
+      "loss": 0.6996,
+      "step": 1082
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.7124025419671638e-05,
+      "loss": 0.6758,
+      "step": 1083
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.711846984258382e-05,
+      "loss": 0.6305,
+      "step": 1084
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.7112909807852237e-05,
+      "loss": 0.6511,
+      "step": 1085
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.7107345318958626e-05,
+      "loss": 0.64,
+      "step": 1086
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.7101776379387518e-05,
+      "loss": 0.6926,
+      "step": 1087
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.7096202992626228e-05,
+      "loss": 0.6726,
+      "step": 1088
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.7090625162164855e-05,
+      "loss": 0.6299,
+      "step": 1089
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.7085042891496283e-05,
+      "loss": 0.6646,
+      "step": 1090
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.707945618411617e-05,
+      "loss": 0.6452,
+      "step": 1091
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.7073865043522965e-05,
+      "loss": 0.6823,
+      "step": 1092
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.375,
+      "learning_rate": 1.7068269473217886e-05,
+      "loss": 0.7104,
+      "step": 1093
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.7062669476704922e-05,
+      "loss": 0.6379,
+      "step": 1094
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.375,
+      "learning_rate": 1.705706505749083e-05,
+      "loss": 0.6919,
+      "step": 1095
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.7051456219085154e-05,
+      "loss": 0.7108,
+      "step": 1096
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.7045842965000185e-05,
+      "loss": 0.609,
+      "step": 1097
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.7040225298750996e-05,
+      "loss": 0.7229,
+      "step": 1098
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.703460322385541e-05,
+      "loss": 0.7307,
+      "step": 1099
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.702897674383402e-05,
+      "loss": 0.7227,
+      "step": 1100
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.702334586221017e-05,
+      "loss": 0.651,
+      "step": 1101
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.701771058250997e-05,
+      "loss": 0.7018,
+      "step": 1102
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.701207090826227e-05,
+      "loss": 0.6952,
+      "step": 1103
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.7006426842998683e-05,
+      "loss": 0.632,
+      "step": 1104
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.7000778390253566e-05,
+      "loss": 0.6741,
+      "step": 1105
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.6995125553564035e-05,
+      "loss": 0.6444,
+      "step": 1106
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.6989468336469933e-05,
+      "loss": 0.704,
+      "step": 1107
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.375,
+      "learning_rate": 1.6983806742513864e-05,
+      "loss": 0.7869,
+      "step": 1108
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.697814077524115e-05,
+      "loss": 0.704,
+      "step": 1109
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.6972470438199883e-05,
+      "loss": 0.6807,
+      "step": 1110
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.25,
+      "learning_rate": 1.696679573494086e-05,
+      "loss": 0.6483,
+      "step": 1111
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.696111666901764e-05,
+      "loss": 0.6747,
+      "step": 1112
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.695543324398649e-05,
+      "loss": 0.7064,
+      "step": 1113
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.25,
+      "learning_rate": 1.694974546340642e-05,
+      "loss": 0.6138,
+      "step": 1114
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.375,
+      "learning_rate": 1.6944053330839164e-05,
+      "loss": 0.6244,
+      "step": 1115
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.693835684984918e-05,
+      "loss": 0.6951,
+      "step": 1116
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.6932656024003654e-05,
+      "loss": 0.695,
+      "step": 1117
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.692695085687249e-05,
+      "loss": 0.5995,
+      "step": 1118
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.6921241352028308e-05,
+      "loss": 0.5518,
+      "step": 1119
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.6915527513046445e-05,
+      "loss": 0.7076,
+      "step": 1120
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.690980934350496e-05,
+      "loss": 0.716,
+      "step": 1121
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.690408684698461e-05,
+      "loss": 0.6554,
+      "step": 1122
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.6898360027068874e-05,
+      "loss": 0.6547,
+      "step": 1123
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.6892628887343928e-05,
+      "loss": 0.7486,
+      "step": 1124
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.6886893431398664e-05,
+      "loss": 0.6808,
+      "step": 1125
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.6881153662824665e-05,
+      "loss": 0.672,
+      "step": 1126
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.6875409585216228e-05,
+      "loss": 0.6536,
+      "step": 1127
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.6869661202170334e-05,
+      "loss": 0.7055,
+      "step": 1128
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.6863908517286675e-05,
+      "loss": 0.6458,
+      "step": 1129
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.685815153416762e-05,
+      "loss": 0.6679,
+      "step": 1130
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.375,
+      "learning_rate": 1.6852390256418246e-05,
+      "loss": 0.6977,
+      "step": 1131
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.684662468764631e-05,
+      "loss": 0.7157,
+      "step": 1132
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.684085483146226e-05,
+      "loss": 0.7225,
+      "step": 1133
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.683508069147922e-05,
+      "loss": 0.6772,
+      "step": 1134
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.375,
+      "learning_rate": 1.6829302271313012e-05,
+      "loss": 0.6667,
+      "step": 1135
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.6823519574582124e-05,
+      "loss": 0.6868,
+      "step": 1136
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.681773260490773e-05,
+      "loss": 0.6645,
+      "step": 1137
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.6811941365913676e-05,
+      "loss": 0.6647,
+      "step": 1138
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.6806145861226487e-05,
+      "loss": 0.6405,
+      "step": 1139
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.680034609447535e-05,
+      "loss": 0.6639,
+      "step": 1140
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.6794542069292128e-05,
+      "loss": 0.7057,
+      "step": 1141
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.678873378931135e-05,
+      "loss": 0.622,
+      "step": 1142
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.6782921258170207e-05,
+      "loss": 0.6802,
+      "step": 1143
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.6777104479508553e-05,
+      "loss": 0.7465,
+      "step": 1144
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.6771283456968905e-05,
+      "loss": 0.7149,
+      "step": 1145
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.6765458194196434e-05,
+      "loss": 0.6351,
+      "step": 1146
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.6759628694838963e-05,
+      "loss": 0.6851,
+      "step": 1147
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.6753794962546976e-05,
+      "loss": 0.6842,
+      "step": 1148
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.6747957000973602e-05,
+      "loss": 0.7133,
+      "step": 1149
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.674211481377462e-05,
+      "loss": 0.6431,
+      "step": 1150
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.6736268404608455e-05,
+      "loss": 0.6966,
+      "step": 1151
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.673041777713618e-05,
+      "loss": 0.6391,
+      "step": 1152
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.67245629350215e-05,
+      "loss": 0.7157,
+      "step": 1153
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.6718703881930764e-05,
+      "loss": 0.6501,
+      "step": 1154
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.6712840621532963e-05,
+      "loss": 0.6681,
+      "step": 1155
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.6706973157499716e-05,
+      "loss": 0.6164,
+      "step": 1156
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.375,
+      "learning_rate": 1.670110149350528e-05,
+      "loss": 0.682,
+      "step": 1157
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.6695225633226538e-05,
+      "loss": 0.6277,
+      "step": 1158
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.6689345580343e-05,
+      "loss": 0.6516,
+      "step": 1159
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.66834613385368e-05,
+      "loss": 0.6172,
+      "step": 1160
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.66775729114927e-05,
+      "loss": 0.6667,
+      "step": 1161
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.375,
+      "learning_rate": 1.6671680302898088e-05,
+      "loss": 0.6663,
+      "step": 1162
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.6665783516442956e-05,
+      "loss": 0.6489,
+      "step": 1163
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.6659882555819917e-05,
+      "loss": 0.6579,
+      "step": 1164
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.25,
+      "learning_rate": 1.6653977424724208e-05,
+      "loss": 0.6469,
+      "step": 1165
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.6648068126853667e-05,
+      "loss": 0.6743,
+      "step": 1166
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.6642154665908747e-05,
+      "loss": 0.6182,
+      "step": 1167
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.6636237045592497e-05,
+      "loss": 0.5786,
+      "step": 1168
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.6630315269610587e-05,
+      "loss": 0.6914,
+      "step": 1169
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.662438934167128e-05,
+      "loss": 0.6614,
+      "step": 1170
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.661845926548544e-05,
+      "loss": 0.6857,
+      "step": 1171
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.6612525044766527e-05,
+      "loss": 0.6418,
+      "step": 1172
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.6606586683230602e-05,
+      "loss": 0.6481,
+      "step": 1173
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.660064418459631e-05,
+      "loss": 0.6846,
+      "step": 1174
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.65946975525849e-05,
+      "loss": 0.6299,
+      "step": 1175
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.658874679092019e-05,
+      "loss": 0.7765,
+      "step": 1176
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.658279190332861e-05,
+      "loss": 0.6393,
+      "step": 1177
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.6576832893539147e-05,
+      "loss": 0.671,
+      "step": 1178
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.6570869765283385e-05,
+      "loss": 0.6612,
+      "step": 1179
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.6564902522295484e-05,
+      "loss": 0.5909,
+      "step": 1180
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.655893116831218e-05,
+      "loss": 0.6282,
+      "step": 1181
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.6552955707072786e-05,
+      "loss": 0.703,
+      "step": 1182
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.654697614231918e-05,
+      "loss": 0.6658,
+      "step": 1183
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.375,
+      "learning_rate": 1.6540992477795813e-05,
+      "loss": 0.6324,
+      "step": 1184
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.6535004717249713e-05,
+      "loss": 0.7287,
+      "step": 1185
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.652901286443046e-05,
+      "loss": 0.6442,
+      "step": 1186
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.6523016923090195e-05,
+      "loss": 0.6943,
+      "step": 1187
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.25,
+      "learning_rate": 1.651701689698363e-05,
+      "loss": 0.5961,
+      "step": 1188
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.6511012789868033e-05,
+      "loss": 0.6663,
+      "step": 1189
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.25,
+      "learning_rate": 1.6505004605503227e-05,
+      "loss": 0.6647,
+      "step": 1190
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.6498992347651572e-05,
+      "loss": 0.6235,
+      "step": 1191
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.649297602007801e-05,
+      "loss": 0.6423,
+      "step": 1192
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.648695562655e-05,
+      "loss": 0.62,
+      "step": 1193
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.6480931170837572e-05,
+      "loss": 0.6129,
+      "step": 1194
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.25,
+      "learning_rate": 1.647490265671328e-05,
+      "loss": 0.6558,
+      "step": 1195
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.6468870087952237e-05,
+      "loss": 0.6497,
+      "step": 1196
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.6462833468332084e-05,
+      "loss": 0.6619,
+      "step": 1197
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.6456792801633e-05,
+      "loss": 0.6667,
+      "step": 1198
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.64507480916377e-05,
+      "loss": 0.7923,
+      "step": 1199
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.644469934213143e-05,
+      "loss": 0.6883,
+      "step": 1200
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.643864655690197e-05,
+      "loss": 0.7018,
+      "step": 1201
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.643258973973962e-05,
+      "loss": 0.6817,
+      "step": 1202
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.6426528894437203e-05,
+      "loss": 0.6402,
+      "step": 1203
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.642046402479008e-05,
+      "loss": 0.6374,
+      "step": 1204
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.641439513459612e-05,
+      "loss": 0.6564,
+      "step": 1205
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.6408322227655704e-05,
+      "loss": 0.6232,
+      "step": 1206
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.640224530777174e-05,
+      "loss": 0.6327,
+      "step": 1207
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.6396164378749646e-05,
+      "loss": 0.6937,
+      "step": 1208
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.639007944439735e-05,
+      "loss": 0.6474,
+      "step": 1209
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.6383990508525283e-05,
+      "loss": 0.6811,
+      "step": 1210
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.6377897574946388e-05,
+      "loss": 0.6243,
+      "step": 1211
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.6371800647476116e-05,
+      "loss": 0.7272,
+      "step": 1212
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.6365699729932403e-05,
+      "loss": 0.6132,
+      "step": 1213
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.6359594826135698e-05,
+      "loss": 0.6601,
+      "step": 1214
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.635348593990894e-05,
+      "loss": 0.6987,
+      "step": 1215
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.6347373075077566e-05,
+      "loss": 0.6431,
+      "step": 1216
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.63412562354695e-05,
+      "loss": 0.6711,
+      "step": 1217
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.6335135424915153e-05,
+      "loss": 0.6747,
+      "step": 1218
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.6329010647247433e-05,
+      "loss": 0.68,
+      "step": 1219
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.6322881906301724e-05,
+      "loss": 0.624,
+      "step": 1220
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.6316749205915888e-05,
+      "loss": 0.6135,
+      "step": 1221
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.631061254993028e-05,
+      "loss": 0.6388,
+      "step": 1222
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.6304471942187717e-05,
+      "loss": 0.5953,
+      "step": 1223
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.62983273865335e-05,
+      "loss": 0.6095,
+      "step": 1224
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.62921788868154e-05,
+      "loss": 0.6288,
+      "step": 1225
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.6286026446883664e-05,
+      "loss": 0.6709,
+      "step": 1226
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.6279870070590985e-05,
+      "loss": 0.6401,
+      "step": 1227
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.627370976179255e-05,
+      "loss": 0.6841,
+      "step": 1228
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.6267545524345988e-05,
+      "loss": 0.6789,
+      "step": 1229
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.6261377362111396e-05,
+      "loss": 0.645,
+      "step": 1230
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.6255205278951328e-05,
+      "loss": 0.5944,
+      "step": 1231
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.624902927873079e-05,
+      "loss": 0.6161,
+      "step": 1232
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.624284936531725e-05,
+      "loss": 0.6765,
+      "step": 1233
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.6236665542580613e-05,
+      "loss": 0.6302,
+      "step": 1234
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.623047781439324e-05,
+      "loss": 0.6193,
+      "step": 1235
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.622428618462994e-05,
+      "loss": 0.6463,
+      "step": 1236
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.6218090657167963e-05,
+      "loss": 0.6754,
+      "step": 1237
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.6211891235886992e-05,
+      "loss": 0.6223,
+      "step": 1238
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.6205687924669162e-05,
+      "loss": 0.6904,
+      "step": 1239
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.6199480727399035e-05,
+      "loss": 0.6217,
+      "step": 1240
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.6193269647963608e-05,
+      "loss": 0.6023,
+      "step": 1241
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.6187054690252306e-05,
+      "loss": 0.7008,
+      "step": 1242
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.6180835858156992e-05,
+      "loss": 0.6577,
+      "step": 1243
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.6174613155571948e-05,
+      "loss": 0.7738,
+      "step": 1244
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.375,
+      "learning_rate": 1.616838658639388e-05,
+      "loss": 0.6541,
+      "step": 1245
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.6162156154521915e-05,
+      "loss": 0.6904,
+      "step": 1246
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.61559218638576e-05,
+      "loss": 0.646,
+      "step": 1247
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.6149683718304904e-05,
+      "loss": 0.6168,
+      "step": 1248
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.6143441721770196e-05,
+      "loss": 0.6509,
+      "step": 1249
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.6137195878162267e-05,
+      "loss": 0.6899,
+      "step": 1250
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.6130946191392323e-05,
+      "loss": 0.6533,
+      "step": 1251
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.612469266537396e-05,
+      "loss": 0.6265,
+      "step": 1252
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.6118435304023195e-05,
+      "loss": 0.6538,
+      "step": 1253
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.611217411125843e-05,
+      "loss": 0.652,
+      "step": 1254
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.610590909100048e-05,
+      "loss": 0.6371,
+      "step": 1255
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.609964024717255e-05,
+      "loss": 0.7017,
+      "step": 1256
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.609336758370024e-05,
+      "loss": 0.6353,
+      "step": 1257
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.25,
+      "learning_rate": 1.6087091104511542e-05,
+      "loss": 0.6408,
+      "step": 1258
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.6080810813536844e-05,
+      "loss": 0.6534,
+      "step": 1259
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.6074526714708913e-05,
+      "loss": 0.6494,
+      "step": 1260
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.60682388119629e-05,
+      "loss": 0.6344,
+      "step": 1261
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.606194710923634e-05,
+      "loss": 0.6793,
+      "step": 1262
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.605565161046916e-05,
+      "loss": 0.5996,
+      "step": 1263
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.604935231960364e-05,
+      "loss": 0.6203,
+      "step": 1264
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.6043049240584445e-05,
+      "loss": 0.6852,
+      "step": 1265
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.6036742377358626e-05,
+      "loss": 0.7155,
+      "step": 1266
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.6030431733875583e-05,
+      "loss": 0.6935,
+      "step": 1267
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.375,
+      "learning_rate": 1.60241173140871e-05,
+      "loss": 0.6943,
+      "step": 1268
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.601779912194731e-05,
+      "loss": 0.6618,
+      "step": 1269
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.6011477161412724e-05,
+      "loss": 0.6908,
+      "step": 1270
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.60051514364422e-05,
+      "loss": 0.6522,
+      "step": 1271
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.5998821950996955e-05,
+      "loss": 0.6982,
+      "step": 1272
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.5992488709040577e-05,
+      "loss": 0.6622,
+      "step": 1273
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.598615171453898e-05,
+      "loss": 0.6811,
+      "step": 1274
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.597981097146045e-05,
+      "loss": 0.6497,
+      "step": 1275
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.597346648377561e-05,
+      "loss": 0.5887,
+      "step": 1276
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.596711825545743e-05,
+      "loss": 0.6406,
+      "step": 1277
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.5960766290481216e-05,
+      "loss": 0.6606,
+      "step": 1278
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.5954410592824633e-05,
+      "loss": 0.6474,
+      "step": 1279
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.594805116646766e-05,
+      "loss": 0.6577,
+      "step": 1280
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.5941688015392622e-05,
+      "loss": 0.5972,
+      "step": 1281
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.593532114358418e-05,
+      "loss": 0.703,
+      "step": 1282
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.5928950555029322e-05,
+      "loss": 0.649,
+      "step": 1283
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.5922576253717362e-05,
+      "loss": 0.6424,
+      "step": 1284
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.5916198243639933e-05,
+      "loss": 0.675,
+      "step": 1285
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.5909816528791002e-05,
+      "loss": 0.6383,
+      "step": 1286
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.375,
+      "learning_rate": 1.5903431113166853e-05,
+      "loss": 0.686,
+      "step": 1287
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.5897042000766082e-05,
+      "loss": 0.6626,
+      "step": 1288
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.5890649195589598e-05,
+      "loss": 0.6889,
+      "step": 1289
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.5884252701640638e-05,
+      "loss": 0.6598,
+      "step": 1290
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.5877852522924733e-05,
+      "loss": 0.6596,
+      "step": 1291
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.587144866344973e-05,
+      "loss": 0.6536,
+      "step": 1292
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.25,
+      "learning_rate": 1.5865041127225772e-05,
+      "loss": 0.6244,
+      "step": 1293
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.25,
+      "learning_rate": 1.5858629918265315e-05,
+      "loss": 0.6168,
+      "step": 1294
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.585221504058311e-05,
+      "loss": 0.6602,
+      "step": 1295
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.5845796498196206e-05,
+      "loss": 0.6654,
+      "step": 1296
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.5839374295123943e-05,
+      "loss": 0.628,
+      "step": 1297
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.5832948435387962e-05,
+      "loss": 0.7059,
+      "step": 1298
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.5826518923012186e-05,
+      "loss": 0.7018,
+      "step": 1299
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.5820085762022827e-05,
+      "loss": 0.6946,
+      "step": 1300
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.5813648956448383e-05,
+      "loss": 0.6504,
+      "step": 1301
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.5807208510319633e-05,
+      "loss": 0.6929,
+      "step": 1302
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.580076442766964e-05,
+      "loss": 0.692,
+      "step": 1303
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.5794316712533744e-05,
+      "loss": 0.6718,
+      "step": 1304
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.578786536894955e-05,
+      "loss": 0.6688,
+      "step": 1305
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.578141040095694e-05,
+      "loss": 0.6805,
+      "step": 1306
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.5774951812598076e-05,
+      "loss": 0.6945,
+      "step": 1307
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.5768489607917376e-05,
+      "loss": 0.6432,
+      "step": 1308
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.576202379096152e-05,
+      "loss": 0.6601,
+      "step": 1309
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.5755554365779458e-05,
+      "loss": 0.7129,
+      "step": 1310
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.5749081336422397e-05,
+      "loss": 0.6755,
+      "step": 1311
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.5742604706943803e-05,
+      "loss": 0.6364,
+      "step": 1312
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.5736124481399394e-05,
+      "loss": 0.6283,
+      "step": 1313
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.5729640663847136e-05,
+      "loss": 0.6377,
+      "step": 1314
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.572315325834725e-05,
+      "loss": 0.6449,
+      "step": 1315
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.5716662268962205e-05,
+      "loss": 0.6391,
+      "step": 1316
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.5710167699756707e-05,
+      "loss": 0.6377,
+      "step": 1317
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.5703669554797712e-05,
+      "loss": 0.6665,
+      "step": 1318
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.5697167838154408e-05,
+      "loss": 0.6607,
+      "step": 1319
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.5690662553898224e-05,
+      "loss": 0.6967,
+      "step": 1320
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.5684153706102826e-05,
+      "loss": 0.6493,
+      "step": 1321
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.56776412988441e-05,
+      "loss": 0.6944,
+      "step": 1322
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.5671125336200172e-05,
+      "loss": 0.6905,
+      "step": 1323
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.5664605822251392e-05,
+      "loss": 0.6395,
+      "step": 1324
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.5658082761080332e-05,
+      "loss": 0.6017,
+      "step": 1325
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.5651556156771786e-05,
+      "loss": 0.7098,
+      "step": 1326
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.5645026013412765e-05,
+      "loss": 0.5936,
+      "step": 1327
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.56384923350925e-05,
+      "loss": 0.6431,
+      "step": 1328
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.5631955125902438e-05,
+      "loss": 0.6926,
+      "step": 1329
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.562541438993622e-05,
+      "loss": 0.5962,
+      "step": 1330
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.561887013128972e-05,
+      "loss": 0.5875,
+      "step": 1331
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.5612322354061e-05,
+      "loss": 0.6327,
+      "step": 1332
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.5605771062350335e-05,
+      "loss": 0.6953,
+      "step": 1333
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.25,
+      "learning_rate": 1.5599216260260196e-05,
+      "loss": 0.6147,
+      "step": 1334
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.559265795189525e-05,
+      "loss": 0.6249,
+      "step": 1335
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.5586096141362367e-05,
+      "loss": 0.7038,
+      "step": 1336
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.5579530832770604e-05,
+      "loss": 0.6775,
+      "step": 1337
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.5572962030231217e-05,
+      "loss": 0.674,
+      "step": 1338
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.556638973785763e-05,
+      "loss": 0.7744,
+      "step": 1339
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.5559813959765482e-05,
+      "loss": 0.6077,
+      "step": 1340
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.5553234700072567e-05,
+      "loss": 0.6764,
+      "step": 1341
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.375,
+      "learning_rate": 1.5546651962898876e-05,
+      "loss": 0.6392,
+      "step": 1342
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.554006575236657e-05,
+      "loss": 0.6318,
+      "step": 1343
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.5533476072599994e-05,
+      "loss": 0.6688,
+      "step": 1344
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.5526882927725652e-05,
+      "loss": 0.606,
+      "step": 1345
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.5520286321872228e-05,
+      "loss": 0.6811,
+      "step": 1346
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.5513686259170566e-05,
+      "loss": 0.6745,
+      "step": 1347
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.5507082743753688e-05,
+      "loss": 0.6653,
+      "step": 1348
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.550047577975676e-05,
+      "loss": 0.6512,
+      "step": 1349
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.5493865371317125e-05,
+      "loss": 0.647,
+      "step": 1350
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.25,
+      "learning_rate": 1.5487251522574266e-05,
+      "loss": 0.6613,
+      "step": 1351
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.5480634237669834e-05,
+      "loss": 0.651,
+      "step": 1352
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.5474013520747626e-05,
+      "loss": 0.6741,
+      "step": 1353
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.5467389375953592e-05,
+      "loss": 0.6726,
+      "step": 1354
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.5460761807435817e-05,
+      "loss": 0.6085,
+      "step": 1355
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.5454130819344545e-05,
+      "loss": 0.6362,
+      "step": 1356
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.5447496415832153e-05,
+      "loss": 0.6467,
+      "step": 1357
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.5440858601053158e-05,
+      "loss": 0.6176,
+      "step": 1358
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.5434217379164214e-05,
+      "loss": 0.6753,
+      "step": 1359
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.542757275432411e-05,
+      "loss": 0.6611,
+      "step": 1360
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.5420924730693764e-05,
+      "loss": 0.6577,
+      "step": 1361
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.5414273312436216e-05,
+      "loss": 0.597,
+      "step": 1362
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.5407618503716647e-05,
+      "loss": 0.6642,
+      "step": 1363
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.5400960308702348e-05,
+      "loss": 0.6271,
+      "step": 1364
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.5394298731562736e-05,
+      "loss": 0.6065,
+      "step": 1365
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.538763377646934e-05,
+      "loss": 0.7082,
+      "step": 1366
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.5380965447595823e-05,
+      "loss": 0.6873,
+      "step": 1367
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.537429374911793e-05,
+      "loss": 0.614,
+      "step": 1368
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.5367618685213545e-05,
+      "loss": 0.7099,
+      "step": 1369
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.536094026006264e-05,
+      "loss": 0.5959,
+      "step": 1370
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.5354258477847303e-05,
+      "loss": 0.6768,
+      "step": 1371
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.5347573342751717e-05,
+      "loss": 0.6782,
+      "step": 1372
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.534088485896217e-05,
+      "loss": 0.6544,
+      "step": 1373
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.5334193030667045e-05,
+      "loss": 0.694,
+      "step": 1374
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.5327497862056825e-05,
+      "loss": 0.678,
+      "step": 1375
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.532079935732407e-05,
+      "loss": 0.6419,
+      "step": 1376
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.5314097520663446e-05,
+      "loss": 0.6902,
+      "step": 1377
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.5307392356271687e-05,
+      "loss": 0.6916,
+      "step": 1378
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.530068386834763e-05,
+      "loss": 0.6451,
+      "step": 1379
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.5293972061092187e-05,
+      "loss": 0.6323,
+      "step": 1380
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.5287256938708338e-05,
+      "loss": 0.6312,
+      "step": 1381
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.5280538505401152e-05,
+      "loss": 0.6518,
+      "step": 1382
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.5273816765377768e-05,
+      "loss": 0.691,
+      "step": 1383
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.526709172284739e-05,
+      "loss": 0.6524,
+      "step": 1384
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.52603633820213e-05,
+      "loss": 0.6528,
+      "step": 1385
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.5253631747112837e-05,
+      "loss": 0.6718,
+      "step": 1386
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.5246896822337402e-05,
+      "loss": 0.5985,
+      "step": 1387
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.5240158611912468e-05,
+      "loss": 0.6488,
+      "step": 1388
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.5233417120057547e-05,
+      "loss": 0.6436,
+      "step": 1389
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.5226672350994222e-05,
+      "loss": 0.643,
+      "step": 1390
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.5219924308946122e-05,
+      "loss": 0.6499,
+      "step": 1391
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.5213172998138928e-05,
+      "loss": 0.6629,
+      "step": 1392
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.520641842280036e-05,
+      "loss": 0.6264,
+      "step": 1393
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.5199660587160195e-05,
+      "loss": 0.6422,
+      "step": 1394
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.5192899495450237e-05,
+      "loss": 0.6837,
+      "step": 1395
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.5186135151904347e-05,
+      "loss": 0.7042,
+      "step": 1396
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.5179367560758399e-05,
+      "loss": 0.6558,
+      "step": 1397
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.5172596726250323e-05,
+      "loss": 0.661,
+      "step": 1398
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.5165822652620068e-05,
+      "loss": 0.621,
+      "step": 1399
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.25,
+      "learning_rate": 1.5159045344109613e-05,
+      "loss": 0.6374,
+      "step": 1400
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.5152264804962964e-05,
+      "loss": 0.6763,
+      "step": 1401
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.5145481039426154e-05,
+      "loss": 0.6241,
+      "step": 1402
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.5138694051747226e-05,
+      "loss": 0.6665,
+      "step": 1403
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.5131903846176253e-05,
+      "loss": 0.6408,
+      "step": 1404
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.512511042696531e-05,
+      "loss": 0.6691,
+      "step": 1405
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.5118313798368498e-05,
+      "loss": 0.6585,
+      "step": 1406
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.5111513964641917e-05,
+      "loss": 0.6489,
+      "step": 1407
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.5104710930043682e-05,
+      "loss": 0.6606,
+      "step": 1408
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.5097904698833902e-05,
+      "loss": 0.6139,
+      "step": 1409
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.5091095275274701e-05,
+      "loss": 0.6252,
+      "step": 1410
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.5084282663630192e-05,
+      "loss": 0.6789,
+      "step": 1411
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.5077466868166493e-05,
+      "loss": 0.6426,
+      "step": 1412
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.5070647893151702e-05,
+      "loss": 0.6366,
+      "step": 1413
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.5063825742855924e-05,
+      "loss": 0.6472,
+      "step": 1414
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.505700042155124e-05,
+      "loss": 0.6364,
+      "step": 1415
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.5050171933511727e-05,
+      "loss": 0.6586,
+      "step": 1416
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.5043340283013435e-05,
+      "loss": 0.6731,
+      "step": 1417
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.5036505474334401e-05,
+      "loss": 0.7051,
+      "step": 1418
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.5029667511754638e-05,
+      "loss": 0.654,
+      "step": 1419
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.5022826399556135e-05,
+      "loss": 0.6722,
+      "step": 1420
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.5015982142022848e-05,
+      "loss": 0.6652,
+      "step": 1421
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.5009134743440712e-05,
+      "loss": 0.6912,
+      "step": 1422
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.5002284208097621e-05,
+      "loss": 0.6361,
+      "step": 1423
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.4995430540283436e-05,
+      "loss": 0.6159,
+      "step": 1424
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.25,
+      "learning_rate": 1.498857374428998e-05,
+      "loss": 0.6182,
+      "step": 1425
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.4981713824411035e-05,
+      "loss": 0.7147,
+      "step": 1426
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.4974850784942338e-05,
+      "loss": 0.6405,
+      "step": 1427
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.496798463018158e-05,
+      "loss": 0.6357,
+      "step": 1428
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.4961115364428403e-05,
+      "loss": 0.695,
+      "step": 1429
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.4954242991984398e-05,
+      "loss": 0.6741,
+      "step": 1430
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.49473675171531e-05,
+      "loss": 0.6502,
+      "step": 1431
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.4940488944239986e-05,
+      "loss": 0.6411,
+      "step": 1432
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.4933607277552478e-05,
+      "loss": 0.6567,
+      "step": 1433
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.4926722521399927e-05,
+      "loss": 0.6514,
+      "step": 1434
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.4919834680093628e-05,
+      "loss": 0.696,
+      "step": 1435
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.4912943757946798e-05,
+      "loss": 0.665,
+      "step": 1436
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.4906049759274594e-05,
+      "loss": 0.6862,
+      "step": 1437
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.489915268839409e-05,
+      "loss": 0.6178,
+      "step": 1438
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.4892252549624292e-05,
+      "loss": 0.6776,
+      "step": 1439
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.4885349347286118e-05,
+      "loss": 0.6868,
+      "step": 1440
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.4878443085702415e-05,
+      "loss": 0.6732,
+      "step": 1441
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.4871533769197934e-05,
+      "loss": 0.6402,
+      "step": 1442
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.4864621402099353e-05,
+      "loss": 0.6938,
+      "step": 1443
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.4857705988735244e-05,
+      "loss": 0.7139,
+      "step": 1444
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.4850787533436101e-05,
+      "loss": 0.6426,
+      "step": 1445
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.4843866040534317e-05,
+      "loss": 0.6832,
+      "step": 1446
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.4836941514364183e-05,
+      "loss": 0.6474,
+      "step": 1447
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.4830013959261895e-05,
+      "loss": 0.7397,
+      "step": 1448
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.482308337956555e-05,
+      "loss": 0.6753,
+      "step": 1449
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.4816149779615128e-05,
+      "loss": 0.6358,
+      "step": 1450
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.4809213163752505e-05,
+      "loss": 0.6726,
+      "step": 1451
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.4802273536321448e-05,
+      "loss": 0.6792,
+      "step": 1452
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.4795330901667608e-05,
+      "loss": 0.6434,
+      "step": 1453
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.4788385264138523e-05,
+      "loss": 0.5785,
+      "step": 1454
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.4781436628083601e-05,
+      "loss": 0.6283,
+      "step": 1455
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.477448499785414e-05,
+      "loss": 0.664,
+      "step": 1456
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.47675303778033e-05,
+      "loss": 0.7725,
+      "step": 1457
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.4760572772286132e-05,
+      "loss": 0.5917,
+      "step": 1458
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.4753612185659536e-05,
+      "loss": 0.6474,
+      "step": 1459
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.4746648622282294e-05,
+      "loss": 0.6223,
+      "step": 1460
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.25,
+      "learning_rate": 1.4739682086515044e-05,
+      "loss": 0.6345,
+      "step": 1461
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.4732712582720283e-05,
+      "loss": 0.6447,
+      "step": 1462
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.4725740115262376e-05,
+      "loss": 0.6474,
+      "step": 1463
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.4718764688507538e-05,
+      "loss": 0.6689,
+      "step": 1464
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.4711786306823838e-05,
+      "loss": 0.6551,
+      "step": 1465
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.4704804974581196e-05,
+      "loss": 0.5983,
+      "step": 1466
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.375,
+      "learning_rate": 1.4697820696151373e-05,
+      "loss": 0.6831,
+      "step": 1467
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.4690833475907991e-05,
+      "loss": 0.6294,
+      "step": 1468
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.4683843318226498e-05,
+      "loss": 0.6342,
+      "step": 1469
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.4676850227484191e-05,
+      "loss": 0.6664,
+      "step": 1470
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.4669854208060195e-05,
+      "loss": 0.6502,
+      "step": 1471
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.4662855264335481e-05,
+      "loss": 0.6665,
+      "step": 1472
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.4655853400692838e-05,
+      "loss": 0.6633,
+      "step": 1473
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.4648848621516894e-05,
+      "loss": 0.6765,
+      "step": 1474
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.46418409311941e-05,
+      "loss": 0.6696,
+      "step": 1475
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.4634830334112722e-05,
+      "loss": 0.652,
+      "step": 1476
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.4627816834662862e-05,
+      "loss": 0.6664,
+      "step": 1477
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.4620800437236426e-05,
+      "loss": 0.5617,
+      "step": 1478
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.4613781146227137e-05,
+      "loss": 0.6502,
+      "step": 1479
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.4606758966030536e-05,
+      "loss": 0.6756,
+      "step": 1480
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.4599733901043968e-05,
+      "loss": 0.6077,
+      "step": 1481
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.4592705955666587e-05,
+      "loss": 0.651,
+      "step": 1482
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.4585675134299344e-05,
+      "loss": 0.632,
+      "step": 1483
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.4578641441345003e-05,
+      "loss": 0.7071,
+      "step": 1484
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.4571604881208118e-05,
+      "loss": 0.6405,
+      "step": 1485
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.4564565458295042e-05,
+      "loss": 0.6497,
+      "step": 1486
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.4557523177013913e-05,
+      "loss": 0.6834,
+      "step": 1487
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.4550478041774671e-05,
+      "loss": 0.6834,
+      "step": 1488
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.4543430056989035e-05,
+      "loss": 0.6424,
+      "step": 1489
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.453637922707051e-05,
+      "loss": 0.614,
+      "step": 1490
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.4529325556434381e-05,
+      "loss": 0.7005,
+      "step": 1491
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.4522269049497722e-05,
+      "loss": 0.6341,
+      "step": 1492
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.4515209710679363e-05,
+      "loss": 0.5657,
+      "step": 1493
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.4508147544399928e-05,
+      "loss": 0.6566,
+      "step": 1494
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.4501082555081802e-05,
+      "loss": 0.5961,
+      "step": 1495
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.4494014747149137e-05,
+      "loss": 0.613,
+      "step": 1496
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.4486944125027854e-05,
+      "loss": 0.6605,
+      "step": 1497
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.447987069314563e-05,
+      "loss": 0.5936,
+      "step": 1498
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.447279445593191e-05,
+      "loss": 0.6614,
+      "step": 1499
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.4465715417817889e-05,
+      "loss": 0.7445,
+      "step": 1500
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.445863358323652e-05,
+      "loss": 0.6502,
+      "step": 1501
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.4451548956622504e-05,
+      "loss": 0.7306,
+      "step": 1502
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.444446154241229e-05,
+      "loss": 0.6453,
+      "step": 1503
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.443737134504408e-05,
+      "loss": 0.6264,
+      "step": 1504
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.4430278368957809e-05,
+      "loss": 0.6725,
+      "step": 1505
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.4423182618595155e-05,
+      "loss": 0.5855,
+      "step": 1506
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.4416084098399542e-05,
+      "loss": 0.6719,
+      "step": 1507
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.4408982812816112e-05,
+      "loss": 0.6398,
+      "step": 1508
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.4401878766291754e-05,
+      "loss": 0.6418,
+      "step": 1509
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.4394771963275079e-05,
+      "loss": 0.6578,
+      "step": 1510
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.4387662408216421e-05,
+      "loss": 0.7189,
+      "step": 1511
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4380550105567843e-05,
+      "loss": 0.7166,
+      "step": 1512
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.4373435059783132e-05,
+      "loss": 0.7168,
+      "step": 1513
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.4366317275317778e-05,
+      "loss": 0.6764,
+      "step": 1514
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.4359196756629006e-05,
+      "loss": 0.6577,
+      "step": 1515
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.4352073508175737e-05,
+      "loss": 0.6687,
+      "step": 1516
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.4344947534418607e-05,
+      "loss": 0.6249,
+      "step": 1517
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.433781883981996e-05,
+      "loss": 0.6129,
+      "step": 1518
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.433068742884384e-05,
+      "loss": 0.7305,
+      "step": 1519
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.4323553305956e-05,
+      "loss": 0.6541,
+      "step": 1520
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.431641647562388e-05,
+      "loss": 0.6208,
+      "step": 1521
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.4309276942316623e-05,
+      "loss": 0.6516,
+      "step": 1522
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.4302134710505066e-05,
+      "loss": 0.686,
+      "step": 1523
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.4294989784661733e-05,
+      "loss": 0.6416,
+      "step": 1524
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.4287842169260827e-05,
+      "loss": 0.6201,
+      "step": 1525
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.4280691868778252e-05,
+      "loss": 0.6048,
+      "step": 1526
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.4273538887691581e-05,
+      "loss": 0.6842,
+      "step": 1527
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.4266383230480067e-05,
+      "loss": 0.6688,
+      "step": 1528
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.4259224901624642e-05,
+      "loss": 0.6183,
+      "step": 1529
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.425206390560791e-05,
+      "loss": 0.6243,
+      "step": 1530
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.4244900246914146e-05,
+      "loss": 0.6752,
+      "step": 1531
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.4237733930029292e-05,
+      "loss": 0.6652,
+      "step": 1532
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.4230564959440949e-05,
+      "loss": 0.6696,
+      "step": 1533
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.4223393339638392e-05,
+      "loss": 0.6196,
+      "step": 1534
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.4216219075112542e-05,
+      "loss": 0.6971,
+      "step": 1535
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.4209042170355987e-05,
+      "loss": 0.6155,
+      "step": 1536
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.4201862629862954e-05,
+      "loss": 0.6271,
+      "step": 1537
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.4194680458129342e-05,
+      "loss": 0.6761,
+      "step": 1538
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.4187495659652675e-05,
+      "loss": 0.6278,
+      "step": 1539
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.4180308238932137e-05,
+      "loss": 0.6242,
+      "step": 1540
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.4173118200468547e-05,
+      "loss": 0.6738,
+      "step": 1541
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.375,
+      "learning_rate": 1.416592554876437e-05,
+      "loss": 0.6333,
+      "step": 1542
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.4158730288323704e-05,
+      "loss": 0.684,
+      "step": 1543
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.4151532423652272e-05,
+      "loss": 0.7104,
+      "step": 1544
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.375,
+      "learning_rate": 1.4144331959257438e-05,
+      "loss": 0.7063,
+      "step": 1545
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.4137128899648197e-05,
+      "loss": 0.6206,
+      "step": 1546
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.4129923249335157e-05,
+      "loss": 0.7004,
+      "step": 1547
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.4122715012830559e-05,
+      "loss": 0.7412,
+      "step": 1548
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.25,
+      "learning_rate": 1.4115504194648256e-05,
+      "loss": 0.6375,
+      "step": 1549
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.4108290799303721e-05,
+      "loss": 0.532,
+      "step": 1550
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.4101074831314047e-05,
+      "loss": 0.7432,
+      "step": 1551
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.4093856295197925e-05,
+      "loss": 0.6823,
+      "step": 1552
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.4086635195475666e-05,
+      "loss": 0.6246,
+      "step": 1553
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.4079411536669179e-05,
+      "loss": 0.6393,
+      "step": 1554
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.4072185323301977e-05,
+      "loss": 0.6144,
+      "step": 1555
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.4064956559899174e-05,
+      "loss": 0.6633,
+      "step": 1556
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.4057725250987485e-05,
+      "loss": 0.6618,
+      "step": 1557
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.4050491401095207e-05,
+      "loss": 0.6929,
+      "step": 1558
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.404325501475224e-05,
+      "loss": 0.6744,
+      "step": 1559
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.4036016096490066e-05,
+      "loss": 0.6997,
+      "step": 1560
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.4028774650841759e-05,
+      "loss": 0.6719,
+      "step": 1561
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.4021530682341968e-05,
+      "loss": 0.6647,
+      "step": 1562
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.4014284195526926e-05,
+      "loss": 0.6459,
+      "step": 1563
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.4007035194934437e-05,
+      "loss": 0.6306,
+      "step": 1564
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.3999783685103893e-05,
+      "loss": 0.5982,
+      "step": 1565
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.3992529670576245e-05,
+      "loss": 0.6613,
+      "step": 1566
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.3985273155894013e-05,
+      "loss": 0.6948,
+      "step": 1567
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.3978014145601285e-05,
+      "loss": 0.6228,
+      "step": 1568
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.3970752644243714e-05,
+      "loss": 0.6023,
+      "step": 1569
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.3963488656368518e-05,
+      "loss": 0.7004,
+      "step": 1570
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.3956222186524455e-05,
+      "loss": 0.7028,
+      "step": 1571
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.3948953239261854e-05,
+      "loss": 0.6193,
+      "step": 1572
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.3941681819132587e-05,
+      "loss": 0.6097,
+      "step": 1573
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.3934407930690077e-05,
+      "loss": 0.6597,
+      "step": 1574
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.3927131578489292e-05,
+      "loss": 0.61,
+      "step": 1575
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.391985276708674e-05,
+      "loss": 0.6559,
+      "step": 1576
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.3912571501040477e-05,
+      "loss": 0.6614,
+      "step": 1577
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.3905287784910087e-05,
+      "loss": 0.6367,
+      "step": 1578
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.3898001623256694e-05,
+      "loss": 0.6827,
+      "step": 1579
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.3890713020642951e-05,
+      "loss": 0.6728,
+      "step": 1580
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.3883421981633043e-05,
+      "loss": 0.638,
+      "step": 1581
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.3876128510792677e-05,
+      "loss": 0.6622,
+      "step": 1582
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.375,
+      "learning_rate": 1.3868832612689081e-05,
+      "loss": 0.7001,
+      "step": 1583
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.3861534291891009e-05,
+      "loss": 0.6036,
+      "step": 1584
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.3854233552968726e-05,
+      "loss": 0.6376,
+      "step": 1585
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.3846930400494015e-05,
+      "loss": 0.6661,
+      "step": 1586
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.3839624839040171e-05,
+      "loss": 0.6432,
+      "step": 1587
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.3832316873181995e-05,
+      "loss": 0.6493,
+      "step": 1588
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.3825006507495794e-05,
+      "loss": 0.6362,
+      "step": 1589
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.3817693746559382e-05,
+      "loss": 0.7273,
+      "step": 1590
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.3810378594952065e-05,
+      "loss": 0.5579,
+      "step": 1591
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.3803061057254653e-05,
+      "loss": 0.6302,
+      "step": 1592
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.379574113804945e-05,
+      "loss": 0.6742,
+      "step": 1593
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.3788418841920244e-05,
+      "loss": 0.6557,
+      "step": 1594
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.3781094173452316e-05,
+      "loss": 0.5966,
+      "step": 1595
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.3773767137232436e-05,
+      "loss": 0.6826,
+      "step": 1596
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.3766437737848854e-05,
+      "loss": 0.6557,
+      "step": 1597
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.37591059798913e-05,
+      "loss": 0.6462,
+      "step": 1598
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.3751771867950978e-05,
+      "loss": 0.6114,
+      "step": 1599
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.3744435406620571e-05,
+      "loss": 0.6968,
+      "step": 1600
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.3737096600494232e-05,
+      "loss": 0.6758,
+      "step": 1601
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.3729755454167574e-05,
+      "loss": 0.6058,
+      "step": 1602
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.375,
+      "learning_rate": 1.3722411972237691e-05,
+      "loss": 0.7095,
+      "step": 1603
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.3715066159303125e-05,
+      "loss": 0.6022,
+      "step": 1604
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.3707718019963887e-05,
+      "loss": 0.6503,
+      "step": 1605
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.3700367558821444e-05,
+      "loss": 0.6663,
+      "step": 1606
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.3693014780478708e-05,
+      "loss": 0.6973,
+      "step": 1607
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.3685659689540054e-05,
+      "loss": 0.6339,
+      "step": 1608
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.3678302290611298e-05,
+      "loss": 0.6128,
+      "step": 1609
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.3670942588299708e-05,
+      "loss": 0.6928,
+      "step": 1610
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.3663580587213977e-05,
+      "loss": 0.6354,
+      "step": 1611
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.3656216291964261e-05,
+      "loss": 0.6388,
+      "step": 1612
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.3648849707162142e-05,
+      "loss": 0.6778,
+      "step": 1613
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.3641480837420626e-05,
+      "loss": 0.6683,
+      "step": 1614
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.3634109687354169e-05,
+      "loss": 0.6148,
+      "step": 1615
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.362673626157864e-05,
+      "loss": 0.603,
+      "step": 1616
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.3619360564711341e-05,
+      "loss": 0.6835,
+      "step": 1617
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.361198260137099e-05,
+      "loss": 0.6772,
+      "step": 1618
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.3604602376177729e-05,
+      "loss": 0.7534,
+      "step": 1619
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.3597219893753119e-05,
+      "loss": 0.627,
+      "step": 1620
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.3589835158720126e-05,
+      "loss": 0.6532,
+      "step": 1621
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.3582448175703131e-05,
+      "loss": 0.6538,
+      "step": 1622
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.3575058949327925e-05,
+      "loss": 0.5842,
+      "step": 1623
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.3567667484221702e-05,
+      "loss": 0.6436,
+      "step": 1624
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.3560273785013057e-05,
+      "loss": 0.7005,
+      "step": 1625
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.3552877856331982e-05,
+      "loss": 0.6459,
+      "step": 1626
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.3545479702809874e-05,
+      "loss": 0.6677,
+      "step": 1627
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.3538079329079513e-05,
+      "loss": 0.5925,
+      "step": 1628
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.3530676739775071e-05,
+      "loss": 0.6329,
+      "step": 1629
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.3523271939532112e-05,
+      "loss": 0.7101,
+      "step": 1630
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.3515864932987583e-05,
+      "loss": 0.6728,
+      "step": 1631
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.3508455724779813e-05,
+      "loss": 0.6117,
+      "step": 1632
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.3501044319548502e-05,
+      "loss": 0.6652,
+      "step": 1633
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.3493630721934739e-05,
+      "loss": 0.6263,
+      "step": 1634
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.3486214936580977e-05,
+      "loss": 0.6289,
+      "step": 1635
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.3478796968131043e-05,
+      "loss": 0.6532,
+      "step": 1636
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.3471376821230127e-05,
+      "loss": 0.5864,
+      "step": 1637
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.3463954500524783e-05,
+      "loss": 0.6319,
+      "step": 1638
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.3456530010662933e-05,
+      "loss": 0.7017,
+      "step": 1639
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.3449103356293853e-05,
+      "loss": 0.6195,
+      "step": 1640
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.3441674542068173e-05,
+      "loss": 0.8728,
+      "step": 1641
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.3434243572637874e-05,
+      "loss": 0.6706,
+      "step": 1642
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.3426810452656293e-05,
+      "loss": 0.5672,
+      "step": 1643
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.3419375186778113e-05,
+      "loss": 0.6401,
+      "step": 1644
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.375,
+      "learning_rate": 1.341193777965935e-05,
+      "loss": 0.6525,
+      "step": 1645
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.3404498235957377e-05,
+      "loss": 0.6305,
+      "step": 1646
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.3397056560330893e-05,
+      "loss": 0.636,
+      "step": 1647
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.3389612757439935e-05,
+      "loss": 0.6625,
+      "step": 1648
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.3382166831945873e-05,
+      "loss": 0.6934,
+      "step": 1649
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.3374718788511412e-05,
+      "loss": 0.625,
+      "step": 1650
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.336726863180057e-05,
+      "loss": 0.7077,
+      "step": 1651
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.3359816366478695e-05,
+      "loss": 0.6063,
+      "step": 1652
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.3352361997212464e-05,
+      "loss": 0.6861,
+      "step": 1653
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.3344905528669858e-05,
+      "loss": 0.5958,
+      "step": 1654
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.25,
+      "learning_rate": 1.3337446965520183e-05,
+      "loss": 0.645,
+      "step": 1655
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.25,
+      "learning_rate": 1.3329986312434045e-05,
+      "loss": 0.6169,
+      "step": 1656
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.3322523574083368e-05,
+      "loss": 0.6359,
+      "step": 1657
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.3315058755141386e-05,
+      "loss": 0.6144,
+      "step": 1658
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.3307591860282623e-05,
+      "loss": 0.626,
+      "step": 1659
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.375,
+      "learning_rate": 1.330012289418291e-05,
+      "loss": 0.625,
+      "step": 1660
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.375,
+      "learning_rate": 1.3292651861519372e-05,
+      "loss": 0.6355,
+      "step": 1661
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.3285178766970437e-05,
+      "loss": 0.5897,
+      "step": 1662
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.3277703615215813e-05,
+      "loss": 0.6268,
+      "step": 1663
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.25,
+      "learning_rate": 1.3270226410936503e-05,
+      "loss": 0.6415,
+      "step": 1664
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.3262747158814791e-05,
+      "loss": 0.592,
+      "step": 1665
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.3255265863534248e-05,
+      "loss": 0.5926,
+      "step": 1666
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.3247782529779723e-05,
+      "loss": 0.6086,
+      "step": 1667
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.3240297162237334e-05,
+      "loss": 0.6033,
+      "step": 1668
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.3232809765594486e-05,
+      "loss": 0.6609,
+      "step": 1669
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.3225320344539845e-05,
+      "loss": 0.6541,
+      "step": 1670
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.3217828903763345e-05,
+      "loss": 0.6986,
+      "step": 1671
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.3210335447956193e-05,
+      "loss": 0.6416,
+      "step": 1672
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.3202839981810846e-05,
+      "loss": 0.7179,
+      "step": 1673
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.3195342510021032e-05,
+      "loss": 0.6529,
+      "step": 1674
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.3187843037281725e-05,
+      "loss": 0.676,
+      "step": 1675
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.25,
+      "learning_rate": 1.3180341568289155e-05,
+      "loss": 0.6514,
+      "step": 1676
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.3172838107740805e-05,
+      "loss": 0.6649,
+      "step": 1677
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.3165332660335403e-05,
+      "loss": 0.6389,
+      "step": 1678
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.3157825230772923e-05,
+      "loss": 0.6428,
+      "step": 1679
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.3150315823754572e-05,
+      "loss": 0.6505,
+      "step": 1680
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.3142804443982805e-05,
+      "loss": 0.7413,
+      "step": 1681
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.3135291096161314e-05,
+      "loss": 0.61,
+      "step": 1682
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.3127775784995016e-05,
+      "loss": 0.6955,
+      "step": 1683
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.3120258515190054e-05,
+      "loss": 0.6515,
+      "step": 1684
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.3112739291453812e-05,
+      "loss": 0.6589,
+      "step": 1685
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.25,
+      "learning_rate": 1.3105218118494887e-05,
+      "loss": 0.6078,
+      "step": 1686
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.3097695001023097e-05,
+      "loss": 0.5646,
+      "step": 1687
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.3090169943749475e-05,
+      "loss": 0.5844,
+      "step": 1688
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.3082642951386282e-05,
+      "loss": 0.655,
+      "step": 1689
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.25,
+      "learning_rate": 1.3075114028646976e-05,
+      "loss": 0.6521,
+      "step": 1690
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.3067583180246233e-05,
+      "loss": 0.6824,
+      "step": 1691
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.3060050410899927e-05,
+      "loss": 0.5668,
+      "step": 1692
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.3052515725325145e-05,
+      "loss": 0.6551,
+      "step": 1693
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.3044979128240167e-05,
+      "loss": 0.6961,
+      "step": 1694
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.3037440624364468e-05,
+      "loss": 0.585,
+      "step": 1695
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.3029900218418723e-05,
+      "loss": 0.6569,
+      "step": 1696
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.3022357915124798e-05,
+      "loss": 0.6099,
+      "step": 1697
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.3014813719205738e-05,
+      "loss": 0.6368,
+      "step": 1698
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.3007267635385787e-05,
+      "loss": 0.612,
+      "step": 1699
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.299971966839036e-05,
+      "loss": 0.6979,
+      "step": 1700
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.2992169822946057e-05,
+      "loss": 0.573,
+      "step": 1701
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.25,
+      "learning_rate": 1.2984618103780652e-05,
+      "loss": 0.6525,
+      "step": 1702
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.2977064515623097e-05,
+      "loss": 0.6219,
+      "step": 1703
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.375,
+      "learning_rate": 1.2969509063203503e-05,
+      "loss": 0.6804,
+      "step": 1704
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.2961951751253158e-05,
+      "loss": 0.648,
+      "step": 1705
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.2954392584504518e-05,
+      "loss": 0.614,
+      "step": 1706
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.2946831567691187e-05,
+      "loss": 0.6012,
+      "step": 1707
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.293926870554794e-05,
+      "loss": 0.5967,
+      "step": 1708
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.29317040028107e-05,
+      "loss": 0.6082,
+      "step": 1709
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.292413746421655e-05,
+      "loss": 0.6528,
+      "step": 1710
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.2916569094503713e-05,
+      "loss": 0.6237,
+      "step": 1711
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.290899889841157e-05,
+      "loss": 0.6215,
+      "step": 1712
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.2901426880680629e-05,
+      "loss": 0.616,
+      "step": 1713
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.289385304605256e-05,
+      "loss": 0.6498,
+      "step": 1714
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.2886277399270153e-05,
+      "loss": 0.6606,
+      "step": 1715
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.2878699945077344e-05,
+      "loss": 0.599,
+      "step": 1716
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.2871120688219192e-05,
+      "loss": 0.6483,
+      "step": 1717
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.2863539633441895e-05,
+      "loss": 0.6241,
+      "step": 1718
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.2855956785492763e-05,
+      "loss": 0.6401,
+      "step": 1719
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.2848372149120248e-05,
+      "loss": 0.632,
+      "step": 1720
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.2840785729073902e-05,
+      "loss": 0.6061,
+      "step": 1721
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.2833197530104409e-05,
+      "loss": 0.6326,
+      "step": 1722
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.2825607556963555e-05,
+      "loss": 0.6665,
+      "step": 1723
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.2818015814404248e-05,
+      "loss": 0.6193,
+      "step": 1724
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.2810422307180497e-05,
+      "loss": 0.5856,
+      "step": 1725
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.2802827040047417e-05,
+      "loss": 0.6652,
+      "step": 1726
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.25,
+      "learning_rate": 1.2795230017761227e-05,
+      "loss": 0.6899,
+      "step": 1727
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.2787631245079247e-05,
+      "loss": 0.6539,
+      "step": 1728
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.2780030726759885e-05,
+      "loss": 0.616,
+      "step": 1729
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.2772428467562653e-05,
+      "loss": 0.6164,
+      "step": 1730
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.25,
+      "learning_rate": 1.2764824472248143e-05,
+      "loss": 0.6673,
+      "step": 1731
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.275721874557804e-05,
+      "loss": 0.6489,
+      "step": 1732
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.2749611292315116e-05,
+      "loss": 0.657,
+      "step": 1733
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.2742002117223215e-05,
+      "loss": 0.6415,
+      "step": 1734
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.25,
+      "learning_rate": 1.2734391225067264e-05,
+      "loss": 0.589,
+      "step": 1735
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.2726778620613268e-05,
+      "loss": 0.6112,
+      "step": 1736
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.25,
+      "learning_rate": 1.2719164308628304e-05,
+      "loss": 0.6294,
+      "step": 1737
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.2711548293880514e-05,
+      "loss": 0.6971,
+      "step": 1738
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.2703930581139106e-05,
+      "loss": 0.623,
+      "step": 1739
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.2696311175174358e-05,
+      "loss": 0.631,
+      "step": 1740
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.2688690080757602e-05,
+      "loss": 0.617,
+      "step": 1741
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.2681067302661228e-05,
+      "loss": 0.6603,
+      "step": 1742
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.2673442845658683e-05,
+      "loss": 0.6663,
+      "step": 1743
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.2665816714524465e-05,
+      "loss": 0.6099,
+      "step": 1744
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.265818891403412e-05,
+      "loss": 0.6745,
+      "step": 1745
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.2650559448964232e-05,
+      "loss": 0.6343,
+      "step": 1746
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.2642928324092445e-05,
+      "loss": 0.608,
+      "step": 1747
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.2635295544197424e-05,
+      "loss": 0.6428,
+      "step": 1748
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.2627661114058883e-05,
+      "loss": 0.6398,
+      "step": 1749
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.2620025038457555e-05,
+      "loss": 0.7136,
+      "step": 1750
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.2612387322175223e-05,
+      "loss": 0.6399,
+      "step": 1751
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.2604747969994682e-05,
+      "loss": 0.6553,
+      "step": 1752
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.25,
+      "learning_rate": 1.2597106986699759e-05,
+      "loss": 0.604,
+      "step": 1753
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.2589464377075295e-05,
+      "loss": 0.602,
+      "step": 1754
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.258182014590716e-05,
+      "loss": 0.6207,
+      "step": 1755
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.2574174297982231e-05,
+      "loss": 0.668,
+      "step": 1756
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.2566526838088401e-05,
+      "loss": 0.6707,
+      "step": 1757
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.255887777101457e-05,
+      "loss": 0.6381,
+      "step": 1758
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.2551227101550651e-05,
+      "loss": 0.648,
+      "step": 1759
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.2543574834487551e-05,
+      "loss": 0.6686,
+      "step": 1760
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.2535920974617181e-05,
+      "loss": 0.6903,
+      "step": 1761
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.2528265526732452e-05,
+      "loss": 0.59,
+      "step": 1762
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.25,
+      "learning_rate": 1.252060849562727e-05,
+      "loss": 0.595,
+      "step": 1763
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.2512949886096529e-05,
+      "loss": 0.6151,
+      "step": 1764
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.2505289702936109e-05,
+      "loss": 0.62,
+      "step": 1765
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.2497627950942883e-05,
+      "loss": 0.62,
+      "step": 1766
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.25,
+      "learning_rate": 1.2489964634914703e-05,
+      "loss": 0.5963,
+      "step": 1767
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.2482299759650399e-05,
+      "loss": 0.6101,
+      "step": 1768
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.2474633329949776e-05,
+      "loss": 0.6633,
+      "step": 1769
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.2466965350613615e-05,
+      "loss": 0.6672,
+      "step": 1770
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.2459295826443671e-05,
+      "loss": 0.6521,
+      "step": 1771
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.245162476224266e-05,
+      "loss": 0.6065,
+      "step": 1772
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.2443952162814263e-05,
+      "loss": 0.6161,
+      "step": 1773
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.2436278032963125e-05,
+      "loss": 0.6591,
+      "step": 1774
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.2428602377494851e-05,
+      "loss": 0.6588,
+      "step": 1775
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.2420925201215999e-05,
+      "loss": 0.7286,
+      "step": 1776
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.2413246508934075e-05,
+      "loss": 0.6285,
+      "step": 1777
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.2405566305457544e-05,
+      "loss": 0.6976,
+      "step": 1778
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.2397884595595808e-05,
+      "loss": 0.6852,
+      "step": 1779
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.239020138415922e-05,
+      "loss": 0.6022,
+      "step": 1780
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.2382516675959062e-05,
+      "loss": 0.5849,
+      "step": 1781
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.237483047580757e-05,
+      "loss": 0.6432,
+      "step": 1782
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.23671427885179e-05,
+      "loss": 0.7115,
+      "step": 1783
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.2359453618904144e-05,
+      "loss": 0.6792,
+      "step": 1784
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.2351762971781324e-05,
+      "loss": 0.5808,
+      "step": 1785
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.2344070851965388e-05,
+      "loss": 0.5925,
+      "step": 1786
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.2336377264273202e-05,
+      "loss": 0.6726,
+      "step": 1787
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.2328682213522554e-05,
+      "loss": 0.6291,
+      "step": 1788
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.2320985704532143e-05,
+      "loss": 0.6343,
+      "step": 1789
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.2313287742121592e-05,
+      "loss": 0.6304,
+      "step": 1790
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.2305588331111425e-05,
+      "loss": 0.603,
+      "step": 1791
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.2297887476323073e-05,
+      "loss": 0.6487,
+      "step": 1792
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.2290185182578877e-05,
+      "loss": 0.6667,
+      "step": 1793
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.2282481454702078e-05,
+      "loss": 0.687,
+      "step": 1794
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.227477629751681e-05,
+      "loss": 0.625,
+      "step": 1795
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.2267069715848108e-05,
+      "loss": 0.6409,
+      "step": 1796
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.2259361714521891e-05,
+      "loss": 0.5752,
+      "step": 1797
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.2251652298364978e-05,
+      "loss": 0.6048,
+      "step": 1798
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.2243941472205066e-05,
+      "loss": 0.6711,
+      "step": 1799
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.2236229240870731e-05,
+      "loss": 0.6277,
+      "step": 1800
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.2228515609191446e-05,
+      "loss": 0.68,
+      "step": 1801
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.2220800581997542e-05,
+      "loss": 0.6789,
+      "step": 1802
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.2213084164120236e-05,
+      "loss": 0.6011,
+      "step": 1803
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.2205366360391611e-05,
+      "loss": 0.5835,
+      "step": 1804
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.2197647175644618e-05,
+      "loss": 0.6455,
+      "step": 1805
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.2189926614713073e-05,
+      "loss": 0.647,
+      "step": 1806
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.2182204682431653e-05,
+      "loss": 0.607,
+      "step": 1807
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.2174481383635894e-05,
+      "loss": 0.6129,
+      "step": 1808
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.2166756723162195e-05,
+      "loss": 0.5868,
+      "step": 1809
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.375,
+      "learning_rate": 1.2159030705847792e-05,
+      "loss": 0.6834,
+      "step": 1810
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.215130333653079e-05,
+      "loss": 0.6162,
+      "step": 1811
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.2143574620050119e-05,
+      "loss": 0.6349,
+      "step": 1812
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.2135844561245576e-05,
+      "loss": 0.6562,
+      "step": 1813
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.212811316495778e-05,
+      "loss": 0.5985,
+      "step": 1814
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.2120380436028194e-05,
+      "loss": 0.6173,
+      "step": 1815
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.2112646379299118e-05,
+      "loss": 0.6815,
+      "step": 1816
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.25,
+      "learning_rate": 1.2104910999613682e-05,
+      "loss": 0.6216,
+      "step": 1817
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.209717430181584e-05,
+      "loss": 0.685,
+      "step": 1818
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.208943629075038e-05,
+      "loss": 0.5858,
+      "step": 1819
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.2081696971262905e-05,
+      "loss": 0.6381,
+      "step": 1820
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.2073956348199838e-05,
+      "loss": 0.5931,
+      "step": 1821
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.2066214426408425e-05,
+      "loss": 0.6363,
+      "step": 1822
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.2058471210736716e-05,
+      "loss": 0.6675,
+      "step": 1823
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.2050726706033575e-05,
+      "loss": 0.7045,
+      "step": 1824
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.204298091714868e-05,
+      "loss": 0.6668,
+      "step": 1825
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.2035233848932501e-05,
+      "loss": 0.6045,
+      "step": 1826
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.2027485506236319e-05,
+      "loss": 0.6532,
+      "step": 1827
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.2019735893912206e-05,
+      "loss": 0.6552,
+      "step": 1828
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.2011985016813033e-05,
+      "loss": 0.6957,
+      "step": 1829
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.2004232879792465e-05,
+      "loss": 0.679,
+      "step": 1830
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.1996479487704953e-05,
+      "loss": 0.6314,
+      "step": 1831
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.198872484540573e-05,
+      "loss": 0.6832,
+      "step": 1832
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.1980968957750823e-05,
+      "loss": 0.5996,
+      "step": 1833
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.1973211829597024e-05,
+      "loss": 0.6546,
+      "step": 1834
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.1965453465801916e-05,
+      "loss": 0.6525,
+      "step": 1835
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.1957693871223846e-05,
+      "loss": 0.6665,
+      "step": 1836
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.1949933050721938e-05,
+      "loss": 0.5906,
+      "step": 1837
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.1942171009156078e-05,
+      "loss": 0.6196,
+      "step": 1838
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.193440775138692e-05,
+      "loss": 0.6606,
+      "step": 1839
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.1926643282275884e-05,
+      "loss": 0.6618,
+      "step": 1840
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.1918877606685142e-05,
+      "loss": 0.654,
+      "step": 1841
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.191111072947762e-05,
+      "loss": 0.6176,
+      "step": 1842
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.1903342655516997e-05,
+      "loss": 0.6305,
+      "step": 1843
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.1895573389667712e-05,
+      "loss": 0.626,
+      "step": 1844
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.1887802936794938e-05,
+      "loss": 0.6472,
+      "step": 1845
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.1880031301764598e-05,
+      "loss": 0.6729,
+      "step": 1846
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.1872258489443352e-05,
+      "loss": 0.614,
+      "step": 1847
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.1864484504698594e-05,
+      "loss": 0.586,
+      "step": 1848
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.1856709352398465e-05,
+      "loss": 0.6377,
+      "step": 1849
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.1848933037411825e-05,
+      "loss": 0.6236,
+      "step": 1850
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.1841155564608262e-05,
+      "loss": 0.5826,
+      "step": 1851
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.1833376938858099e-05,
+      "loss": 0.7116,
+      "step": 1852
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.182559716503237e-05,
+      "loss": 0.6794,
+      "step": 1853
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.1817816248002835e-05,
+      "loss": 0.6261,
+      "step": 1854
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.25,
+      "learning_rate": 1.1810034192641969e-05,
+      "loss": 0.6152,
+      "step": 1855
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.1802251003822953e-05,
+      "loss": 0.6669,
+      "step": 1856
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.1794466686419687e-05,
+      "loss": 0.6161,
+      "step": 1857
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.1786681245306774e-05,
+      "loss": 0.688,
+      "step": 1858
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.1778894685359522e-05,
+      "loss": 0.6535,
+      "step": 1859
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.1771107011453934e-05,
+      "loss": 0.6495,
+      "step": 1860
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.176331822846672e-05,
+      "loss": 0.6308,
+      "step": 1861
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.1755528341275278e-05,
+      "loss": 0.6358,
+      "step": 1862
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.174773735475769e-05,
+      "loss": 0.6467,
+      "step": 1863
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.173994527379275e-05,
+      "loss": 0.6152,
+      "step": 1864
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.25,
+      "learning_rate": 1.1732152103259913e-05,
+      "loss": 0.6136,
+      "step": 1865
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.1724357848039326e-05,
+      "loss": 0.6335,
+      "step": 1866
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.1716562513011815e-05,
+      "loss": 0.6101,
+      "step": 1867
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.1708766103058889e-05,
+      "loss": 0.7098,
+      "step": 1868
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.1700968623062716e-05,
+      "loss": 0.6464,
+      "step": 1869
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.1693170077906145e-05,
+      "loss": 0.5972,
+      "step": 1870
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.1685370472472685e-05,
+      "loss": 0.6692,
+      "step": 1871
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.1677569811646518e-05,
+      "loss": 0.6482,
+      "step": 1872
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.166976810031248e-05,
+      "loss": 0.5957,
+      "step": 1873
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.166196534335606e-05,
+      "loss": 0.6169,
+      "step": 1874
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.1654161545663413e-05,
+      "loss": 0.5836,
+      "step": 1875
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.1646356712121341e-05,
+      "loss": 0.6053,
+      "step": 1876
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.1638550847617294e-05,
+      "loss": 0.6357,
+      "step": 1877
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.1630743957039364e-05,
+      "loss": 0.6322,
+      "step": 1878
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.1622936045276297e-05,
+      "loss": 0.6756,
+      "step": 1879
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.25,
+      "learning_rate": 1.1615127117217465e-05,
+      "loss": 0.577,
+      "step": 1880
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.160731717775288e-05,
+      "loss": 0.6182,
+      "step": 1881
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.1599506231773196e-05,
+      "loss": 0.7013,
+      "step": 1882
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.159169428416969e-05,
+      "loss": 0.6478,
+      "step": 1883
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.1583881339834266e-05,
+      "loss": 0.5915,
+      "step": 1884
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.157606740365945e-05,
+      "loss": 0.6629,
+      "step": 1885
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.1568252480538396e-05,
+      "loss": 0.6169,
+      "step": 1886
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.1560436575364874e-05,
+      "loss": 0.6278,
+      "step": 1887
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.1552619693033264e-05,
+      "loss": 0.6116,
+      "step": 1888
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.1544801838438564e-05,
+      "loss": 0.6203,
+      "step": 1889
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.25,
+      "learning_rate": 1.1536983016476375e-05,
+      "loss": 0.6329,
+      "step": 1890
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.152916323204291e-05,
+      "loss": 0.6059,
+      "step": 1891
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.1521342490034982e-05,
+      "loss": 0.5858,
+      "step": 1892
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.151352079535e-05,
+      "loss": 0.678,
+      "step": 1893
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.1505698152885974e-05,
+      "loss": 0.6392,
+      "step": 1894
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.1497874567541505e-05,
+      "loss": 0.6441,
+      "step": 1895
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.1490050044215791e-05,
+      "loss": 0.6068,
+      "step": 1896
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.1482224587808608e-05,
+      "loss": 0.6152,
+      "step": 1897
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.1474398203220322e-05,
+      "loss": 0.6324,
+      "step": 1898
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.146657089535188e-05,
+      "loss": 0.6098,
+      "step": 1899
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.1458742669104806e-05,
+      "loss": 0.6597,
+      "step": 1900
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.1450913529381192e-05,
+      "loss": 0.6917,
+      "step": 1901
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.1443083481083717e-05,
+      "loss": 0.5949,
+      "step": 1902
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.143525252911562e-05,
+      "loss": 0.6245,
+      "step": 1903
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.1427420678380701e-05,
+      "loss": 0.7013,
+      "step": 1904
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.1419587933783334e-05,
+      "loss": 0.6754,
+      "step": 1905
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.1411754300228449e-05,
+      "loss": 0.593,
+      "step": 1906
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.1403919782621526e-05,
+      "loss": 0.6638,
+      "step": 1907
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.375,
+      "learning_rate": 1.1396084385868607e-05,
+      "loss": 0.6682,
+      "step": 1908
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.1388248114876279e-05,
+      "loss": 0.7277,
+      "step": 1909
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.1380410974551683e-05,
+      "loss": 0.6245,
+      "step": 1910
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.1372572969802496e-05,
+      "loss": 0.6221,
+      "step": 1911
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.1364734105536939e-05,
+      "loss": 0.6502,
+      "step": 1912
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.1356894386663778e-05,
+      "loss": 0.5901,
+      "step": 1913
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.1349053818092304e-05,
+      "loss": 0.6421,
+      "step": 1914
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.1341212404732348e-05,
+      "loss": 0.6115,
+      "step": 1915
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.1333370151494267e-05,
+      "loss": 0.7278,
+      "step": 1916
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.1325527063288939e-05,
+      "loss": 0.6604,
+      "step": 1917
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.1317683145027775e-05,
+      "loss": 0.6265,
+      "step": 1918
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.1309838401622698e-05,
+      "loss": 0.5815,
+      "step": 1919
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.1301992837986154e-05,
+      "loss": 0.6127,
+      "step": 1920
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.129414645903109e-05,
+      "loss": 0.6091,
+      "step": 1921
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.1286299269670982e-05,
+      "loss": 0.619,
+      "step": 1922
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.25,
+      "learning_rate": 1.1278451274819799e-05,
+      "loss": 0.5875,
+      "step": 1923
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.127060247939202e-05,
+      "loss": 0.6102,
+      "step": 1924
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.1262752888302624e-05,
+      "loss": 0.6306,
+      "step": 1925
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.125490250646709e-05,
+      "loss": 0.62,
+      "step": 1926
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.1247051338801392e-05,
+      "loss": 0.6428,
+      "step": 1927
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.1239199390221992e-05,
+      "loss": 0.6454,
+      "step": 1928
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.1231346665645845e-05,
+      "loss": 0.6538,
+      "step": 1929
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.1223493169990394e-05,
+      "loss": 0.6062,
+      "step": 1930
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.1215638908173558e-05,
+      "loss": 0.6293,
+      "step": 1931
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.1207783885113743e-05,
+      "loss": 0.6086,
+      "step": 1932
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.1199928105729825e-05,
+      "loss": 0.6302,
+      "step": 1933
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.1192071574941162e-05,
+      "loss": 0.6074,
+      "step": 1934
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.1184214297667571e-05,
+      "loss": 0.6374,
+      "step": 1935
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.1176356278829346e-05,
+      "loss": 0.631,
+      "step": 1936
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.1168497523347247e-05,
+      "loss": 0.599,
+      "step": 1937
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.1160638036142484e-05,
+      "loss": 0.6923,
+      "step": 1938
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.1152777822136733e-05,
+      "loss": 0.6629,
+      "step": 1939
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.1144916886252126e-05,
+      "loss": 0.5698,
+      "step": 1940
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.1137055233411242e-05,
+      "loss": 0.5902,
+      "step": 1941
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.112919286853712e-05,
+      "loss": 0.6455,
+      "step": 1942
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.1121329796553228e-05,
+      "loss": 0.6215,
+      "step": 1943
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.1113466022383488e-05,
+      "loss": 0.6352,
+      "step": 1944
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.110560155095226e-05,
+      "loss": 0.5916,
+      "step": 1945
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.1097736387184343e-05,
+      "loss": 0.679,
+      "step": 1946
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.1089870536004965e-05,
+      "loss": 0.584,
+      "step": 1947
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.1082004002339782e-05,
+      "loss": 0.6433,
+      "step": 1948
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.1074136791114884e-05,
+      "loss": 0.6961,
+      "step": 1949
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.1066268907256783e-05,
+      "loss": 0.6523,
+      "step": 1950
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.1058400355692413e-05,
+      "loss": 0.6292,
+      "step": 1951
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.1050531141349123e-05,
+      "loss": 0.707,
+      "step": 1952
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.1042661269154682e-05,
+      "loss": 0.6185,
+      "step": 1953
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.25,
+      "learning_rate": 1.1034790744037267e-05,
+      "loss": 0.6696,
+      "step": 1954
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.1026919570925463e-05,
+      "loss": 0.6943,
+      "step": 1955
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.375,
+      "learning_rate": 1.1019047754748261e-05,
+      "loss": 0.6686,
+      "step": 1956
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.1011175300435064e-05,
+      "loss": 0.6189,
+      "step": 1957
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.1003302212915659e-05,
+      "loss": 0.6716,
+      "step": 1958
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.099542849712024e-05,
+      "loss": 0.6392,
+      "step": 1959
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.0987554157979392e-05,
+      "loss": 0.6501,
+      "step": 1960
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.0979679200424094e-05,
+      "loss": 0.6697,
+      "step": 1961
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.0971803629385704e-05,
+      "loss": 0.7781,
+      "step": 1962
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.0963927449795968e-05,
+      "loss": 0.6521,
+      "step": 1963
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.095605066658702e-05,
+      "loss": 0.6826,
+      "step": 1964
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.0948173284691359e-05,
+      "loss": 0.7059,
+      "step": 1965
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.0940295309041869e-05,
+      "loss": 0.6001,
+      "step": 1966
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.0932416744571801e-05,
+      "loss": 0.6266,
+      "step": 1967
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.0924537596214777e-05,
+      "loss": 0.654,
+      "step": 1968
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.0916657868904784e-05,
+      "loss": 0.6429,
+      "step": 1969
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.090877756757617e-05,
+      "loss": 0.6492,
+      "step": 1970
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.0900896697163645e-05,
+      "loss": 0.6198,
+      "step": 1971
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.0893015262602275e-05,
+      "loss": 0.6779,
+      "step": 1972
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.0885133268827479e-05,
+      "loss": 0.626,
+      "step": 1973
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.0877250720775026e-05,
+      "loss": 0.6013,
+      "step": 1974
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.0869367623381026e-05,
+      "loss": 0.6785,
+      "step": 1975
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.0861483981581947e-05,
+      "loss": 0.6668,
+      "step": 1976
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.0853599800314582e-05,
+      "loss": 0.6581,
+      "step": 1977
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.0845715084516073e-05,
+      "loss": 0.6823,
+      "step": 1978
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.0837829839123893e-05,
+      "loss": 0.6493,
+      "step": 1979
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.375,
+      "learning_rate": 1.0829944069075848e-05,
+      "loss": 0.6922,
+      "step": 1980
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.0822057779310073e-05,
+      "loss": 0.5462,
+      "step": 1981
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.0814170974765023e-05,
+      "loss": 0.6174,
+      "step": 1982
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.080628366037948e-05,
+      "loss": 0.6267,
+      "step": 1983
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.0798395841092545e-05,
+      "loss": 0.6628,
+      "step": 1984
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.0790507521843636e-05,
+      "loss": 0.6185,
+      "step": 1985
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.0782618707572481e-05,
+      "loss": 0.6406,
+      "step": 1986
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.077472940321912e-05,
+      "loss": 0.6478,
+      "step": 1987
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.0766839613723897e-05,
+      "loss": 0.6623,
+      "step": 1988
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.0758949344027467e-05,
+      "loss": 0.6658,
+      "step": 1989
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.0751058599070782e-05,
+      "loss": 0.6582,
+      "step": 1990
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.0743167383795086e-05,
+      "loss": 0.6741,
+      "step": 1991
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.0735275703141923e-05,
+      "loss": 0.5944,
+      "step": 1992
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.0727383562053134e-05,
+      "loss": 0.6208,
+      "step": 1993
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.0719490965470829e-05,
+      "loss": 0.6504,
+      "step": 1994
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.071159791833743e-05,
+      "loss": 0.6401,
+      "step": 1995
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.0703704425595623e-05,
+      "loss": 0.6147,
+      "step": 1996
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.069581049218837e-05,
+      "loss": 0.6346,
+      "step": 1997
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.0687916123058924e-05,
+      "loss": 0.5931,
+      "step": 1998
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.0680021323150805e-05,
+      "loss": 0.638,
+      "step": 1999
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.0672126097407796e-05,
+      "loss": 0.6331,
+      "step": 2000
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.0664230450773954e-05,
+      "loss": 0.6319,
+      "step": 2001
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.0656334388193592e-05,
+      "loss": 0.6575,
+      "step": 2002
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.25,
+      "learning_rate": 1.0648437914611298e-05,
+      "loss": 0.624,
+      "step": 2003
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.0640541034971901e-05,
+      "loss": 0.6438,
+      "step": 2004
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.0632643754220496e-05,
+      "loss": 0.6148,
+      "step": 2005
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.0624746077302416e-05,
+      "loss": 0.6776,
+      "step": 2006
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.061684800916326e-05,
+      "loss": 0.6401,
+      "step": 2007
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.060894955474886e-05,
+      "loss": 0.6302,
+      "step": 2008
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.060105071900529e-05,
+      "loss": 0.6181,
+      "step": 2009
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.0593151506878868e-05,
+      "loss": 0.5919,
+      "step": 2010
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.0585251923316143e-05,
+      "loss": 0.6422,
+      "step": 2011
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.0577351973263898e-05,
+      "loss": 0.6686,
+      "step": 2012
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.056945166166915e-05,
+      "loss": 0.6864,
+      "step": 2013
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.0561550993479132e-05,
+      "loss": 0.6104,
+      "step": 2014
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.0553649973641314e-05,
+      "loss": 0.6154,
+      "step": 2015
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.0545748607103371e-05,
+      "loss": 0.6206,
+      "step": 2016
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.0537846898813213e-05,
+      "loss": 0.6216,
+      "step": 2017
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.25,
+      "learning_rate": 1.0529944853718944e-05,
+      "loss": 0.6827,
+      "step": 2018
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.0522042476768896e-05,
+      "loss": 0.6177,
+      "step": 2019
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.0514139772911598e-05,
+      "loss": 0.5835,
+      "step": 2020
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.050623674709579e-05,
+      "loss": 0.6197,
+      "step": 2021
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.0498333404270409e-05,
+      "loss": 0.6228,
+      "step": 2022
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.375,
+      "learning_rate": 1.0490429749384593e-05,
+      "loss": 0.7181,
+      "step": 2023
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.0482525787387674e-05,
+      "loss": 0.6228,
+      "step": 2024
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.0474621523229182e-05,
+      "loss": 0.6514,
+      "step": 2025
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.0466716961858825e-05,
+      "loss": 0.6123,
+      "step": 2026
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.0458812108226511e-05,
+      "loss": 0.6115,
+      "step": 2027
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.0450906967282317e-05,
+      "loss": 0.7127,
+      "step": 2028
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.044300154397651e-05,
+      "loss": 0.63,
+      "step": 2029
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.0435095843259531e-05,
+      "loss": 0.6466,
+      "step": 2030
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.042718987008199e-05,
+      "loss": 0.6538,
+      "step": 2031
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.041928362939468e-05,
+      "loss": 0.6505,
+      "step": 2032
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.0411377126148541e-05,
+      "loss": 0.6173,
+      "step": 2033
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.0403470365294701e-05,
+      "loss": 0.6681,
+      "step": 2034
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.0395563351784433e-05,
+      "loss": 0.6615,
+      "step": 2035
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 4070,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 2035,
+  "total_flos": 6.005399895744184e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}