diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,140034 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.32,
+  "eval_steps": 500,
+  "global_step": 20000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.6e-05,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0,
+      "loss": 0.174,
+      "step": 1
+    },
+    {
+      "epoch": 3.2e-05,
+      "grad_norm": 1.1328125,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 0.1894,
+      "step": 2
+    },
+    {
+      "epoch": 4.8e-05,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 0.1369,
+      "step": 3
+    },
+    {
+      "epoch": 6.4e-05,
+      "grad_norm": 0.7890625,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 0.1449,
+      "step": 4
+    },
+    {
+      "epoch": 8e-05,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 0.1653,
+      "step": 5
+    },
+    {
+      "epoch": 9.6e-05,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.152,
+      "step": 6
+    },
+    {
+      "epoch": 0.000112,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 0.167,
+      "step": 7
+    },
+    {
+      "epoch": 0.000128,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 0.1683,
+      "step": 8
+    },
+    {
+      "epoch": 0.000144,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 0.1367,
+      "step": 9
+    },
+    {
+      "epoch": 0.00016,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.8e-06,
+      "loss": 0.1824,
+      "step": 10
+    },
+    {
+      "epoch": 0.000176,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.133,
+      "step": 11
+    },
+    {
+      "epoch": 0.000192,
+      "grad_norm": 1.171875,
+      "learning_rate": 2.2e-06,
+      "loss": 0.175,
+      "step": 12
+    },
+    {
+      "epoch": 0.000208,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.1658,
+      "step": 13
+    },
+    {
+      "epoch": 0.000224,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.6e-06,
+      "loss": 0.1192,
+      "step": 14
+    },
+    {
+      "epoch": 0.00024,
+      "grad_norm": 1.1171875,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 0.1685,
+      "step": 15
+    },
+    {
+      "epoch": 0.000256,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3e-06,
+      "loss": 0.1557,
+      "step": 16
+    },
+    {
+      "epoch": 0.000272,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 0.1482,
+      "step": 17
+    },
+    {
+      "epoch": 0.000288,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 0.2086,
+      "step": 18
+    },
+    {
+      "epoch": 0.000304,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.6e-06,
+      "loss": 0.1731,
+      "step": 19
+    },
+    {
+      "epoch": 0.00032,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.8e-06,
+      "loss": 0.1784,
+      "step": 20
+    },
+    {
+      "epoch": 0.000336,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.1804,
+      "step": 21
+    },
+    {
+      "epoch": 0.000352,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 0.1808,
+      "step": 22
+    },
+    {
+      "epoch": 0.000368,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.4e-06,
+      "loss": 0.1685,
+      "step": 23
+    },
+    {
+      "epoch": 0.000384,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.6e-06,
+      "loss": 0.2063,
+      "step": 24
+    },
+    {
+      "epoch": 0.0004,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 0.1327,
+      "step": 25
+    },
+    {
+      "epoch": 0.000416,
+      "grad_norm": 1.28125,
+      "learning_rate": 5e-06,
+      "loss": 0.1434,
+      "step": 26
+    },
+    {
+      "epoch": 0.000432,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.2e-06,
+      "loss": 0.1345,
+      "step": 27
+    },
+    {
+      "epoch": 0.000448,
+      "grad_norm": 0.77734375,
+      "learning_rate": 5.4e-06,
+      "loss": 0.1699,
+      "step": 28
+    },
+    {
+      "epoch": 0.000464,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 0.2125,
+      "step": 29
+    },
+    {
+      "epoch": 0.00048,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.8e-06,
+      "loss": 0.1713,
+      "step": 30
+    },
+    {
+      "epoch": 0.000496,
+      "grad_norm": 1.1484375,
+      "learning_rate": 6e-06,
+      "loss": 0.1814,
+      "step": 31
+    },
+    {
+      "epoch": 0.000512,
+      "grad_norm": 0.8359375,
+      "learning_rate": 6.2e-06,
+      "loss": 0.158,
+      "step": 32
+    },
+    {
+      "epoch": 0.000528,
+      "grad_norm": 1.3984375,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 0.1556,
+      "step": 33
+    },
+    {
+      "epoch": 0.000544,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.6e-06,
+      "loss": 0.144,
+      "step": 34
+    },
+    {
+      "epoch": 0.00056,
+      "grad_norm": 0.8984375,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 0.1726,
+      "step": 35
+    },
+    {
+      "epoch": 0.000576,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 0.123,
+      "step": 36
+    },
+    {
+      "epoch": 0.000592,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.2e-06,
+      "loss": 0.1859,
+      "step": 37
+    },
+    {
+      "epoch": 0.000608,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.4e-06,
+      "loss": 0.1612,
+      "step": 38
+    },
+    {
+      "epoch": 0.000624,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.6e-06,
+      "loss": 0.1631,
+      "step": 39
+    },
+    {
+      "epoch": 0.00064,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.8e-06,
+      "loss": 0.1854,
+      "step": 40
+    },
+    {
+      "epoch": 0.000656,
+      "grad_norm": 1.328125,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.2083,
+      "step": 41
+    },
+    {
+      "epoch": 0.000672,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.200000000000001e-06,
+      "loss": 0.1447,
+      "step": 42
+    },
+    {
+      "epoch": 0.000688,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 0.1509,
+      "step": 43
+    },
+    {
+      "epoch": 0.000704,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.599999999999999e-06,
+      "loss": 0.1864,
+      "step": 44
+    },
+    {
+      "epoch": 0.00072,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.8e-06,
+      "loss": 0.132,
+      "step": 45
+    },
+    {
+      "epoch": 0.000736,
+      "grad_norm": 1.0625,
+      "learning_rate": 9e-06,
+      "loss": 0.1832,
+      "step": 46
+    },
+    {
+      "epoch": 0.000752,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.2e-06,
+      "loss": 0.1795,
+      "step": 47
+    },
+    {
+      "epoch": 0.000768,
+      "grad_norm": 0.875,
+      "learning_rate": 9.4e-06,
+      "loss": 0.1725,
+      "step": 48
+    },
+    {
+      "epoch": 0.000784,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 0.1437,
+      "step": 49
+    },
+    {
+      "epoch": 0.0008,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 0.166,
+      "step": 50
+    },
+    {
+      "epoch": 0.000816,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1e-05,
+      "loss": 0.1476,
+      "step": 51
+    },
+    {
+      "epoch": 0.000832,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.02e-05,
+      "loss": 0.1428,
+      "step": 52
+    },
+    {
+      "epoch": 0.000848,
+      "grad_norm": 1.125,
+      "learning_rate": 1.04e-05,
+      "loss": 0.171,
+      "step": 53
+    },
+    {
+      "epoch": 0.000864,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.06e-05,
+      "loss": 0.1706,
+      "step": 54
+    },
+    {
+      "epoch": 0.00088,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.08e-05,
+      "loss": 0.1609,
+      "step": 55
+    },
+    {
+      "epoch": 0.000896,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 0.1793,
+      "step": 56
+    },
+    {
+      "epoch": 0.000912,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 0.1266,
+      "step": 57
+    },
+    {
+      "epoch": 0.000928,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.1400000000000001e-05,
+      "loss": 0.1899,
+      "step": 58
+    },
+    {
+      "epoch": 0.000944,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.16e-05,
+      "loss": 0.1401,
+      "step": 59
+    },
+    {
+      "epoch": 0.00096,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.18e-05,
+      "loss": 0.1821,
+      "step": 60
+    },
+    {
+      "epoch": 0.000976,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.2e-05,
+      "loss": 0.1617,
+      "step": 61
+    },
+    {
+      "epoch": 0.000992,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.22e-05,
+      "loss": 0.1581,
+      "step": 62
+    },
+    {
+      "epoch": 0.001008,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.24e-05,
+      "loss": 0.1682,
+      "step": 63
+    },
+    {
+      "epoch": 0.001024,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.2600000000000001e-05,
+      "loss": 0.1505,
+      "step": 64
+    },
+    {
+      "epoch": 0.00104,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 0.1694,
+      "step": 65
+    },
+    {
+      "epoch": 0.001056,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 0.1492,
+      "step": 66
+    },
+    {
+      "epoch": 0.001072,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.32e-05,
+      "loss": 0.1371,
+      "step": 67
+    },
+    {
+      "epoch": 0.001088,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.3400000000000002e-05,
+      "loss": 0.1462,
+      "step": 68
+    },
+    {
+      "epoch": 0.001104,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.3600000000000002e-05,
+      "loss": 0.1775,
+      "step": 69
+    },
+    {
+      "epoch": 0.00112,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.3800000000000002e-05,
+      "loss": 0.1499,
+      "step": 70
+    },
+    {
+      "epoch": 0.001136,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 0.2359,
+      "step": 71
+    },
+    {
+      "epoch": 0.001152,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.42e-05,
+      "loss": 0.1679,
+      "step": 72
+    },
+    {
+      "epoch": 0.001168,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.44e-05,
+      "loss": 0.1869,
+      "step": 73
+    },
+    {
+      "epoch": 0.001184,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.4599999999999999e-05,
+      "loss": 0.1773,
+      "step": 74
+    },
+    {
+      "epoch": 0.0012,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.48e-05,
+      "loss": 0.2028,
+      "step": 75
+    },
+    {
+      "epoch": 0.001216,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.5e-05,
+      "loss": 0.1804,
+      "step": 76
+    },
+    {
+      "epoch": 0.001232,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.52e-05,
+      "loss": 0.1782,
+      "step": 77
+    },
+    {
+      "epoch": 0.001248,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.54e-05,
+      "loss": 0.1598,
+      "step": 78
+    },
+    {
+      "epoch": 0.001264,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.56e-05,
+      "loss": 0.1331,
+      "step": 79
+    },
+    {
+      "epoch": 0.00128,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.58e-05,
+      "loss": 0.1671,
+      "step": 80
+    },
+    {
+      "epoch": 0.001296,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.2039,
+      "step": 81
+    },
+    {
+      "epoch": 0.001312,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.62e-05,
+      "loss": 0.1371,
+      "step": 82
+    },
+    {
+      "epoch": 0.001328,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.6400000000000002e-05,
+      "loss": 0.1956,
+      "step": 83
+    },
+    {
+      "epoch": 0.001344,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.66e-05,
+      "loss": 0.1622,
+      "step": 84
+    },
+    {
+      "epoch": 0.00136,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 0.1371,
+      "step": 85
+    },
+    {
+      "epoch": 0.001376,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 0.1759,
+      "step": 86
+    },
+    {
+      "epoch": 0.001392,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.7199999999999998e-05,
+      "loss": 0.1703,
+      "step": 87
+    },
+    {
+      "epoch": 0.001408,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.74e-05,
+      "loss": 0.1557,
+      "step": 88
+    },
+    {
+      "epoch": 0.001424,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.76e-05,
+      "loss": 0.1653,
+      "step": 89
+    },
+    {
+      "epoch": 0.00144,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.78e-05,
+      "loss": 0.198,
+      "step": 90
+    },
+    {
+      "epoch": 0.001456,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.8e-05,
+      "loss": 0.1726,
+      "step": 91
+    },
+    {
+      "epoch": 0.001472,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.8200000000000002e-05,
+      "loss": 0.1674,
+      "step": 92
+    },
+    {
+      "epoch": 0.001488,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.84e-05,
+      "loss": 0.1423,
+      "step": 93
+    },
+    {
+      "epoch": 0.001504,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.86e-05,
+      "loss": 0.1453,
+      "step": 94
+    },
+    {
+      "epoch": 0.00152,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.88e-05,
+      "loss": 0.1467,
+      "step": 95
+    },
+    {
+      "epoch": 0.001536,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.9e-05,
+      "loss": 0.1738,
+      "step": 96
+    },
+    {
+      "epoch": 0.001552,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 0.1828,
+      "step": 97
+    },
+    {
+      "epoch": 0.001568,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.94e-05,
+      "loss": 0.1639,
+      "step": 98
+    },
+    {
+      "epoch": 0.001584,
+      "grad_norm": 1.5,
+      "learning_rate": 1.9600000000000002e-05,
+      "loss": 0.1891,
+      "step": 99
+    },
+    {
+      "epoch": 0.0016,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.9800000000000004e-05,
+      "loss": 0.1952,
+      "step": 100
+    },
+    {
+      "epoch": 0.001616,
+      "grad_norm": 0.84375,
+      "learning_rate": 2e-05,
+      "loss": 0.1669,
+      "step": 101
+    },
+    {
+      "epoch": 0.001632,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.0200000000000003e-05,
+      "loss": 0.1208,
+      "step": 102
+    },
+    {
+      "epoch": 0.001648,
+      "grad_norm": 0.98046875,
+      "learning_rate": 2.04e-05,
+      "loss": 0.1962,
+      "step": 103
+    },
+    {
+      "epoch": 0.001664,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.06e-05,
+      "loss": 0.1711,
+      "step": 104
+    },
+    {
+      "epoch": 0.00168,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.08e-05,
+      "loss": 0.147,
+      "step": 105
+    },
+    {
+      "epoch": 0.001696,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.1e-05,
+      "loss": 0.1624,
+      "step": 106
+    },
+    {
+      "epoch": 0.001712,
+      "grad_norm": 0.96875,
+      "learning_rate": 2.12e-05,
+      "loss": 0.1629,
+      "step": 107
+    },
+    {
+      "epoch": 0.001728,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.1400000000000002e-05,
+      "loss": 0.1447,
+      "step": 108
+    },
+    {
+      "epoch": 0.001744,
+      "grad_norm": 1.0625,
+      "learning_rate": 2.16e-05,
+      "loss": 0.1665,
+      "step": 109
+    },
+    {
+      "epoch": 0.00176,
+      "grad_norm": 1.0859375,
+      "learning_rate": 2.18e-05,
+      "loss": 0.1938,
+      "step": 110
+    },
+    {
+      "epoch": 0.001776,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 0.1281,
+      "step": 111
+    },
+    {
+      "epoch": 0.001792,
+      "grad_norm": 1.125,
+      "learning_rate": 2.22e-05,
+      "loss": 0.1806,
+      "step": 112
+    },
+    {
+      "epoch": 0.001808,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.2400000000000002e-05,
+      "loss": 0.1742,
+      "step": 113
+    },
+    {
+      "epoch": 0.001824,
+      "grad_norm": 1.5,
+      "learning_rate": 2.26e-05,
+      "loss": 0.1644,
+      "step": 114
+    },
+    {
+      "epoch": 0.00184,
+      "grad_norm": 1.28125,
+      "learning_rate": 2.2800000000000002e-05,
+      "loss": 0.1629,
+      "step": 115
+    },
+    {
+      "epoch": 0.001856,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 0.1904,
+      "step": 116
+    },
+    {
+      "epoch": 0.001872,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.32e-05,
+      "loss": 0.1586,
+      "step": 117
+    },
+    {
+      "epoch": 0.001888,
+      "grad_norm": 1.0859375,
+      "learning_rate": 2.3400000000000003e-05,
+      "loss": 0.1533,
+      "step": 118
+    },
+    {
+      "epoch": 0.001904,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.36e-05,
+      "loss": 0.1656,
+      "step": 119
+    },
+    {
+      "epoch": 0.00192,
+      "grad_norm": 1.2734375,
+      "learning_rate": 2.38e-05,
+      "loss": 0.2069,
+      "step": 120
+    },
+    {
+      "epoch": 0.001936,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.4e-05,
+      "loss": 0.1847,
+      "step": 121
+    },
+    {
+      "epoch": 0.001952,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.4200000000000002e-05,
+      "loss": 0.1357,
+      "step": 122
+    },
+    {
+      "epoch": 0.001968,
+      "grad_norm": 1.3671875,
+      "learning_rate": 2.44e-05,
+      "loss": 0.1611,
+      "step": 123
+    },
+    {
+      "epoch": 0.001984,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.46e-05,
+      "loss": 0.1302,
+      "step": 124
+    },
+    {
+      "epoch": 0.002,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.48e-05,
+      "loss": 0.1717,
+      "step": 125
+    },
+    {
+      "epoch": 0.002016,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.5e-05,
+      "loss": 0.1391,
+      "step": 126
+    },
+    {
+      "epoch": 0.002032,
+      "grad_norm": 1.34375,
+      "learning_rate": 2.5200000000000003e-05,
+      "loss": 0.1786,
+      "step": 127
+    },
+    {
+      "epoch": 0.002048,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.54e-05,
+      "loss": 0.2045,
+      "step": 128
+    },
+    {
+      "epoch": 0.002064,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.5600000000000002e-05,
+      "loss": 0.1627,
+      "step": 129
+    },
+    {
+      "epoch": 0.00208,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.58e-05,
+      "loss": 0.1702,
+      "step": 130
+    },
+    {
+      "epoch": 0.002096,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 0.1836,
+      "step": 131
+    },
+    {
+      "epoch": 0.002112,
+      "grad_norm": 1.1484375,
+      "learning_rate": 2.6200000000000003e-05,
+      "loss": 0.141,
+      "step": 132
+    },
+    {
+      "epoch": 0.002128,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.64e-05,
+      "loss": 0.1592,
+      "step": 133
+    },
+    {
+      "epoch": 0.002144,
+      "grad_norm": 1.0,
+      "learning_rate": 2.6600000000000003e-05,
+      "loss": 0.1565,
+      "step": 134
+    },
+    {
+      "epoch": 0.00216,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.6800000000000004e-05,
+      "loss": 0.1939,
+      "step": 135
+    },
+    {
+      "epoch": 0.002176,
+      "grad_norm": 1.1796875,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 0.1388,
+      "step": 136
+    },
+    {
+      "epoch": 0.002192,
+      "grad_norm": 1.2421875,
+      "learning_rate": 2.7200000000000004e-05,
+      "loss": 0.1161,
+      "step": 137
+    },
+    {
+      "epoch": 0.002208,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.7400000000000002e-05,
+      "loss": 0.1408,
+      "step": 138
+    },
+    {
+      "epoch": 0.002224,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.7600000000000003e-05,
+      "loss": 0.1761,
+      "step": 139
+    },
+    {
+      "epoch": 0.00224,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.7800000000000005e-05,
+      "loss": 0.1733,
+      "step": 140
+    },
+    {
+      "epoch": 0.002256,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 0.1632,
+      "step": 141
+    },
+    {
+      "epoch": 0.002272,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.8199999999999998e-05,
+      "loss": 0.1258,
+      "step": 142
+    },
+    {
+      "epoch": 0.002288,
+      "grad_norm": 0.91796875,
+      "learning_rate": 2.84e-05,
+      "loss": 0.1679,
+      "step": 143
+    },
+    {
+      "epoch": 0.002304,
+      "grad_norm": 1.1796875,
+      "learning_rate": 2.86e-05,
+      "loss": 0.1344,
+      "step": 144
+    },
+    {
+      "epoch": 0.00232,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.88e-05,
+      "loss": 0.1395,
+      "step": 145
+    },
+    {
+      "epoch": 0.002336,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.9e-05,
+      "loss": 0.1829,
+      "step": 146
+    },
+    {
+      "epoch": 0.002352,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.9199999999999998e-05,
+      "loss": 0.1236,
+      "step": 147
+    },
+    {
+      "epoch": 0.002368,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.94e-05,
+      "loss": 0.1365,
+      "step": 148
+    },
+    {
+      "epoch": 0.002384,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.96e-05,
+      "loss": 0.1918,
+      "step": 149
+    },
+    {
+      "epoch": 0.0024,
+      "grad_norm": 1.359375,
+      "learning_rate": 2.98e-05,
+      "loss": 0.1845,
+      "step": 150
+    },
+    {
+      "epoch": 0.002416,
+      "grad_norm": 1.5625,
+      "learning_rate": 3e-05,
+      "loss": 0.1969,
+      "step": 151
+    },
+    {
+      "epoch": 0.002432,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.02e-05,
+      "loss": 0.2335,
+      "step": 152
+    },
+    {
+      "epoch": 0.002448,
+      "grad_norm": 1.5859375,
+      "learning_rate": 3.04e-05,
+      "loss": 0.1608,
+      "step": 153
+    },
+    {
+      "epoch": 0.002464,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.06e-05,
+      "loss": 0.1574,
+      "step": 154
+    },
+    {
+      "epoch": 0.00248,
+      "grad_norm": 1.15625,
+      "learning_rate": 3.08e-05,
+      "loss": 0.1992,
+      "step": 155
+    },
+    {
+      "epoch": 0.002496,
+      "grad_norm": 1.109375,
+      "learning_rate": 3.1e-05,
+      "loss": 0.1659,
+      "step": 156
+    },
+    {
+      "epoch": 0.002512,
+      "grad_norm": 1.203125,
+      "learning_rate": 3.12e-05,
+      "loss": 0.1573,
+      "step": 157
+    },
+    {
+      "epoch": 0.002528,
+      "grad_norm": 1.15625,
+      "learning_rate": 3.1400000000000004e-05,
+      "loss": 0.1253,
+      "step": 158
+    },
+    {
+      "epoch": 0.002544,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.16e-05,
+      "loss": 0.1616,
+      "step": 159
+    },
+    {
+      "epoch": 0.00256,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.18e-05,
+      "loss": 0.166,
+      "step": 160
+    },
+    {
+      "epoch": 0.002576,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 0.1635,
+      "step": 161
+    },
+    {
+      "epoch": 0.002592,
+      "grad_norm": 1.40625,
+      "learning_rate": 3.2200000000000003e-05,
+      "loss": 0.1571,
+      "step": 162
+    },
+    {
+      "epoch": 0.002608,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.24e-05,
+      "loss": 0.1529,
+      "step": 163
+    },
+    {
+      "epoch": 0.002624,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.26e-05,
+      "loss": 0.1824,
+      "step": 164
+    },
+    {
+      "epoch": 0.00264,
+      "grad_norm": 1.3984375,
+      "learning_rate": 3.2800000000000004e-05,
+      "loss": 0.1937,
+      "step": 165
+    },
+    {
+      "epoch": 0.002656,
+      "grad_norm": 1.2734375,
+      "learning_rate": 3.3e-05,
+      "loss": 0.1855,
+      "step": 166
+    },
+    {
+      "epoch": 0.002672,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.32e-05,
+      "loss": 0.1935,
+      "step": 167
+    },
+    {
+      "epoch": 0.002688,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.3400000000000005e-05,
+      "loss": 0.178,
+      "step": 168
+    },
+    {
+      "epoch": 0.002704,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.3600000000000004e-05,
+      "loss": 0.1569,
+      "step": 169
+    },
+    {
+      "epoch": 0.00272,
+      "grad_norm": 1.453125,
+      "learning_rate": 3.38e-05,
+      "loss": 0.1673,
+      "step": 170
+    },
+    {
+      "epoch": 0.002736,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 0.1317,
+      "step": 171
+    },
+    {
+      "epoch": 0.002752,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.4200000000000005e-05,
+      "loss": 0.1505,
+      "step": 172
+    },
+    {
+      "epoch": 0.002768,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.4399999999999996e-05,
+      "loss": 0.1696,
+      "step": 173
+    },
+    {
+      "epoch": 0.002784,
+      "grad_norm": 1.0,
+      "learning_rate": 3.46e-05,
+      "loss": 0.176,
+      "step": 174
+    },
+    {
+      "epoch": 0.0028,
+      "grad_norm": 1.125,
+      "learning_rate": 3.48e-05,
+      "loss": 0.1443,
+      "step": 175
+    },
+    {
+      "epoch": 0.002816,
+      "grad_norm": 1.4140625,
+      "learning_rate": 3.5e-05,
+      "loss": 0.1985,
+      "step": 176
+    },
+    {
+      "epoch": 0.002832,
+      "grad_norm": 1.21875,
+      "learning_rate": 3.52e-05,
+      "loss": 0.1553,
+      "step": 177
+    },
+    {
+      "epoch": 0.002848,
+      "grad_norm": 1.2578125,
+      "learning_rate": 3.54e-05,
+      "loss": 0.1247,
+      "step": 178
+    },
+    {
+      "epoch": 0.002864,
+      "grad_norm": 1.578125,
+      "learning_rate": 3.56e-05,
+      "loss": 0.1442,
+      "step": 179
+    },
+    {
+      "epoch": 0.00288,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.58e-05,
+      "loss": 0.1703,
+      "step": 180
+    },
+    {
+      "epoch": 0.002896,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.6e-05,
+      "loss": 0.1625,
+      "step": 181
+    },
+    {
+      "epoch": 0.002912,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.62e-05,
+      "loss": 0.1524,
+      "step": 182
+    },
+    {
+      "epoch": 0.002928,
+      "grad_norm": 1.34375,
+      "learning_rate": 3.6400000000000004e-05,
+      "loss": 0.1852,
+      "step": 183
+    },
+    {
+      "epoch": 0.002944,
+      "grad_norm": 1.2890625,
+      "learning_rate": 3.66e-05,
+      "loss": 0.171,
+      "step": 184
+    },
+    {
+      "epoch": 0.00296,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.68e-05,
+      "loss": 0.1577,
+      "step": 185
+    },
+    {
+      "epoch": 0.002976,
+      "grad_norm": 0.97265625,
+      "learning_rate": 3.7e-05,
+      "loss": 0.1775,
+      "step": 186
+    },
+    {
+      "epoch": 0.002992,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.72e-05,
+      "loss": 0.172,
+      "step": 187
+    },
+    {
+      "epoch": 0.003008,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.74e-05,
+      "loss": 0.1324,
+      "step": 188
+    },
+    {
+      "epoch": 0.003024,
+      "grad_norm": 1.0703125,
+      "learning_rate": 3.76e-05,
+      "loss": 0.1733,
+      "step": 189
+    },
+    {
+      "epoch": 0.00304,
+      "grad_norm": 1.2578125,
+      "learning_rate": 3.7800000000000004e-05,
+      "loss": 0.173,
+      "step": 190
+    },
+    {
+      "epoch": 0.003056,
+      "grad_norm": 0.6875,
+      "learning_rate": 3.8e-05,
+      "loss": 0.1535,
+      "step": 191
+    },
+    {
+      "epoch": 0.003072,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.82e-05,
+      "loss": 0.1517,
+      "step": 192
+    },
+    {
+      "epoch": 0.003088,
+      "grad_norm": 1.5703125,
+      "learning_rate": 3.8400000000000005e-05,
+      "loss": 0.1485,
+      "step": 193
+    },
+    {
+      "epoch": 0.003104,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.86e-05,
+      "loss": 0.1405,
+      "step": 194
+    },
+    {
+      "epoch": 0.00312,
+      "grad_norm": 1.1015625,
+      "learning_rate": 3.88e-05,
+      "loss": 0.1749,
+      "step": 195
+    },
+    {
+      "epoch": 0.003136,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 0.1806,
+      "step": 196
+    },
+    {
+      "epoch": 0.003152,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.9200000000000004e-05,
+      "loss": 0.1921,
+      "step": 197
+    },
+    {
+      "epoch": 0.003168,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.94e-05,
+      "loss": 0.1632,
+      "step": 198
+    },
+    {
+      "epoch": 0.003184,
+      "grad_norm": 0.97265625,
+      "learning_rate": 3.960000000000001e-05,
+      "loss": 0.1807,
+      "step": 199
+    },
+    {
+      "epoch": 0.0032,
+      "grad_norm": 1.703125,
+      "learning_rate": 3.9800000000000005e-05,
+      "loss": 0.162,
+      "step": 200
+    },
+    {
+      "epoch": 0.003216,
+      "grad_norm": 1.6328125,
+      "learning_rate": 4e-05,
+      "loss": 0.1903,
+      "step": 201
+    },
+    {
+      "epoch": 0.003232,
+      "grad_norm": 1.5390625,
+      "learning_rate": 4.02e-05,
+      "loss": 0.15,
+      "step": 202
+    },
+    {
+      "epoch": 0.003248,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.0400000000000006e-05,
+      "loss": 0.1587,
+      "step": 203
+    },
+    {
+      "epoch": 0.003264,
+      "grad_norm": 1.34375,
+      "learning_rate": 4.0600000000000004e-05,
+      "loss": 0.1641,
+      "step": 204
+    },
+    {
+      "epoch": 0.00328,
+      "grad_norm": 0.875,
+      "learning_rate": 4.08e-05,
+      "loss": 0.1442,
+      "step": 205
+    },
+    {
+      "epoch": 0.003296,
+      "grad_norm": 0.875,
+      "learning_rate": 4.1e-05,
+      "loss": 0.1871,
+      "step": 206
+    },
+    {
+      "epoch": 0.003312,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.12e-05,
+      "loss": 0.1665,
+      "step": 207
+    },
+    {
+      "epoch": 0.003328,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.14e-05,
+      "loss": 0.188,
+      "step": 208
+    },
+    {
+      "epoch": 0.003344,
+      "grad_norm": 1.2265625,
+      "learning_rate": 4.16e-05,
+      "loss": 0.1708,
+      "step": 209
+    },
+    {
+      "epoch": 0.00336,
+      "grad_norm": 1.7421875,
+      "learning_rate": 4.18e-05,
+      "loss": 0.1738,
+      "step": 210
+    },
+    {
+      "epoch": 0.003376,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.2e-05,
+      "loss": 0.1465,
+      "step": 211
+    },
+    {
+      "epoch": 0.003392,
+      "grad_norm": 1.1875,
+      "learning_rate": 4.22e-05,
+      "loss": 0.1777,
+      "step": 212
+    },
+    {
+      "epoch": 0.003408,
+      "grad_norm": 1.53125,
+      "learning_rate": 4.24e-05,
+      "loss": 0.1629,
+      "step": 213
+    },
+    {
+      "epoch": 0.003424,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.26e-05,
+      "loss": 0.1491,
+      "step": 214
+    },
+    {
+      "epoch": 0.00344,
+      "grad_norm": 1.328125,
+      "learning_rate": 4.2800000000000004e-05,
+      "loss": 0.1601,
+      "step": 215
+    },
+    {
+      "epoch": 0.003456,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.3e-05,
+      "loss": 0.157,
+      "step": 216
+    },
+    {
+      "epoch": 0.003472,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.32e-05,
+      "loss": 0.1577,
+      "step": 217
+    },
+    {
+      "epoch": 0.003488,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.3400000000000005e-05,
+      "loss": 0.2011,
+      "step": 218
+    },
+    {
+      "epoch": 0.003504,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.36e-05,
+      "loss": 0.165,
+      "step": 219
+    },
+    {
+      "epoch": 0.00352,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.38e-05,
+      "loss": 0.1751,
+      "step": 220
+    },
+    {
+      "epoch": 0.003536,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.1521,
+      "step": 221
+    },
+    {
+      "epoch": 0.003552,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.4200000000000004e-05,
+      "loss": 0.1781,
+      "step": 222
+    },
+    {
+      "epoch": 0.003568,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.44e-05,
+      "loss": 0.1545,
+      "step": 223
+    },
+    {
+      "epoch": 0.003584,
+      "grad_norm": 1.3125,
+      "learning_rate": 4.46e-05,
+      "loss": 0.1892,
+      "step": 224
+    },
+    {
+      "epoch": 0.0036,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.4800000000000005e-05,
+      "loss": 0.157,
+      "step": 225
+    },
+    {
+      "epoch": 0.003616,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.5e-05,
+      "loss": 0.1149,
+      "step": 226
+    },
+    {
+      "epoch": 0.003632,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.52e-05,
+      "loss": 0.1543,
+      "step": 227
+    },
+    {
+      "epoch": 0.003648,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.5400000000000006e-05,
+      "loss": 0.1527,
+      "step": 228
+    },
+    {
+      "epoch": 0.003664,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.5600000000000004e-05,
+      "loss": 0.1731,
+      "step": 229
+    },
+    {
+      "epoch": 0.00368,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.58e-05,
+      "loss": 0.181,
+      "step": 230
+    },
+    {
+      "epoch": 0.003696,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.1346,
+      "step": 231
+    },
+    {
+      "epoch": 0.003712,
+      "grad_norm": 1.34375,
+      "learning_rate": 4.6200000000000005e-05,
+      "loss": 0.1624,
+      "step": 232
+    },
+    {
+      "epoch": 0.003728,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.64e-05,
+      "loss": 0.1739,
+      "step": 233
+    },
+    {
+      "epoch": 0.003744,
+      "grad_norm": 1.3828125,
+      "learning_rate": 4.660000000000001e-05,
+      "loss": 0.1737,
+      "step": 234
+    },
+    {
+      "epoch": 0.00376,
+      "grad_norm": 1.2578125,
+      "learning_rate": 4.6800000000000006e-05,
+      "loss": 0.1791,
+      "step": 235
+    },
+    {
+      "epoch": 0.003776,
+      "grad_norm": 1.515625,
+      "learning_rate": 4.7e-05,
+      "loss": 0.1629,
+      "step": 236
+    },
+    {
+      "epoch": 0.003792,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.72e-05,
+      "loss": 0.1376,
+      "step": 237
+    },
+    {
+      "epoch": 0.003808,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.74e-05,
+      "loss": 0.1914,
+      "step": 238
+    },
+    {
+      "epoch": 0.003824,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.76e-05,
+      "loss": 0.2158,
+      "step": 239
+    },
+    {
+      "epoch": 0.00384,
+      "grad_norm": 1.3984375,
+      "learning_rate": 4.78e-05,
+      "loss": 0.1639,
+      "step": 240
+    },
+    {
+      "epoch": 0.003856,
+      "grad_norm": 1.140625,
+      "learning_rate": 4.8e-05,
+      "loss": 0.1623,
+      "step": 241
+    },
+    {
+      "epoch": 0.003872,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.82e-05,
+      "loss": 0.182,
+      "step": 242
+    },
+    {
+      "epoch": 0.003888,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.8400000000000004e-05,
+      "loss": 0.1513,
+      "step": 243
+    },
+    {
+      "epoch": 0.003904,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.86e-05,
+      "loss": 0.1746,
+      "step": 244
+    },
+    {
+      "epoch": 0.00392,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.88e-05,
+      "loss": 0.1398,
+      "step": 245
+    },
+    {
+      "epoch": 0.003936,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.9e-05,
+      "loss": 0.161,
+      "step": 246
+    },
+    {
+      "epoch": 0.003952,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.92e-05,
+      "loss": 0.1681,
+      "step": 247
+    },
+    {
+      "epoch": 0.003968,
+      "grad_norm": 1.3984375,
+      "learning_rate": 4.94e-05,
+      "loss": 0.1586,
+      "step": 248
+    },
+    {
+      "epoch": 0.003984,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.96e-05,
+      "loss": 0.1859,
+      "step": 249
+    },
+    {
+      "epoch": 0.004,
+      "grad_norm": 1.21875,
+      "learning_rate": 4.9800000000000004e-05,
+      "loss": 0.1903,
+      "step": 250
+    },
+    {
+      "epoch": 0.004016,
+      "grad_norm": 0.734375,
+      "learning_rate": 5e-05,
+      "loss": 0.1602,
+      "step": 251
+    },
+    {
+      "epoch": 0.004032,
+      "grad_norm": 1.21875,
+      "learning_rate": 5.02e-05,
+      "loss": 0.1552,
+      "step": 252
+    },
+    {
+      "epoch": 0.004048,
+      "grad_norm": 1.796875,
+      "learning_rate": 5.0400000000000005e-05,
+      "loss": 0.1569,
+      "step": 253
+    },
+    {
+      "epoch": 0.004064,
+      "grad_norm": 0.9921875,
+      "learning_rate": 5.0600000000000003e-05,
+      "loss": 0.1393,
+      "step": 254
+    },
+    {
+      "epoch": 0.00408,
+      "grad_norm": 1.1796875,
+      "learning_rate": 5.08e-05,
+      "loss": 0.1806,
+      "step": 255
+    },
+    {
+      "epoch": 0.004096,
+      "grad_norm": 1.4140625,
+      "learning_rate": 5.1000000000000006e-05,
+      "loss": 0.1881,
+      "step": 256
+    },
+    {
+      "epoch": 0.004112,
+      "grad_norm": 0.921875,
+      "learning_rate": 5.1200000000000004e-05,
+      "loss": 0.1707,
+      "step": 257
+    },
+    {
+      "epoch": 0.004128,
+      "grad_norm": 0.81640625,
+      "learning_rate": 5.14e-05,
+      "loss": 0.1334,
+      "step": 258
+    },
+    {
+      "epoch": 0.004144,
+      "grad_norm": 1.15625,
+      "learning_rate": 5.16e-05,
+      "loss": 0.1567,
+      "step": 259
+    },
+    {
+      "epoch": 0.00416,
+      "grad_norm": 0.92578125,
+      "learning_rate": 5.1800000000000005e-05,
+      "loss": 0.1968,
+      "step": 260
+    },
+    {
+      "epoch": 0.004176,
+      "grad_norm": 0.82421875,
+      "learning_rate": 5.2000000000000004e-05,
+      "loss": 0.1845,
+      "step": 261
+    },
+    {
+      "epoch": 0.004192,
+      "grad_norm": 1.4375,
+      "learning_rate": 5.22e-05,
+      "loss": 0.178,
+      "step": 262
+    },
+    {
+      "epoch": 0.004208,
+      "grad_norm": 1.359375,
+      "learning_rate": 5.2400000000000007e-05,
+      "loss": 0.1573,
+      "step": 263
+    },
+    {
+      "epoch": 0.004224,
+      "grad_norm": 1.203125,
+      "learning_rate": 5.2600000000000005e-05,
+      "loss": 0.199,
+      "step": 264
+    },
+    {
+      "epoch": 0.00424,
+      "grad_norm": 1.1875,
+      "learning_rate": 5.28e-05,
+      "loss": 0.1837,
+      "step": 265
+    },
+    {
+      "epoch": 0.004256,
+      "grad_norm": 1.1953125,
+      "learning_rate": 5.300000000000001e-05,
+      "loss": 0.1973,
+      "step": 266
+    },
+    {
+      "epoch": 0.004272,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.3200000000000006e-05,
+      "loss": 0.1716,
+      "step": 267
+    },
+    {
+      "epoch": 0.004288,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.3400000000000004e-05,
+      "loss": 0.1815,
+      "step": 268
+    },
+    {
+      "epoch": 0.004304,
+      "grad_norm": 1.2578125,
+      "learning_rate": 5.360000000000001e-05,
+      "loss": 0.182,
+      "step": 269
+    },
+    {
+      "epoch": 0.00432,
+      "grad_norm": 1.390625,
+      "learning_rate": 5.380000000000001e-05,
+      "loss": 0.1986,
+      "step": 270
+    },
+    {
+      "epoch": 0.004336,
+      "grad_norm": 1.5,
+      "learning_rate": 5.4000000000000005e-05,
+      "loss": 0.1955,
+      "step": 271
+    },
+    {
+      "epoch": 0.004352,
+      "grad_norm": 1.4296875,
+      "learning_rate": 5.420000000000001e-05,
+      "loss": 0.1624,
+      "step": 272
+    },
+    {
+      "epoch": 0.004368,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.440000000000001e-05,
+      "loss": 0.1641,
+      "step": 273
+    },
+    {
+      "epoch": 0.004384,
+      "grad_norm": 13.25,
+      "learning_rate": 5.4600000000000006e-05,
+      "loss": 0.2348,
+      "step": 274
+    },
+    {
+      "epoch": 0.0044,
+      "grad_norm": 0.98046875,
+      "learning_rate": 5.4800000000000004e-05,
+      "loss": 0.195,
+      "step": 275
+    },
+    {
+      "epoch": 0.004416,
+      "grad_norm": 0.90625,
+      "learning_rate": 5.500000000000001e-05,
+      "loss": 0.1762,
+      "step": 276
+    },
+    {
+      "epoch": 0.004432,
+      "grad_norm": 0.79296875,
+      "learning_rate": 5.520000000000001e-05,
+      "loss": 0.191,
+      "step": 277
+    },
+    {
+      "epoch": 0.004448,
+      "grad_norm": 1.4765625,
+      "learning_rate": 5.5400000000000005e-05,
+      "loss": 0.1919,
+      "step": 278
+    },
+    {
+      "epoch": 0.004464,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.560000000000001e-05,
+      "loss": 0.1617,
+      "step": 279
+    },
+    {
+      "epoch": 0.00448,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.580000000000001e-05,
+      "loss": 0.1588,
+      "step": 280
+    },
+    {
+      "epoch": 0.004496,
+      "grad_norm": 0.8984375,
+      "learning_rate": 5.6000000000000006e-05,
+      "loss": 0.1669,
+      "step": 281
+    },
+    {
+      "epoch": 0.004512,
+      "grad_norm": 0.9140625,
+      "learning_rate": 5.620000000000001e-05,
+      "loss": 0.1729,
+      "step": 282
+    },
+    {
+      "epoch": 0.004528,
+      "grad_norm": 1.2890625,
+      "learning_rate": 5.6399999999999995e-05,
+      "loss": 0.1824,
+      "step": 283
+    },
+    {
+      "epoch": 0.004544,
+      "grad_norm": 1.0,
+      "learning_rate": 5.66e-05,
+      "loss": 0.15,
+      "step": 284
+    },
+    {
+      "epoch": 0.00456,
+      "grad_norm": 1.1328125,
+      "learning_rate": 5.68e-05,
+      "loss": 0.1749,
+      "step": 285
+    },
+    {
+      "epoch": 0.004576,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.6999999999999996e-05,
+      "loss": 0.1679,
+      "step": 286
+    },
+    {
+      "epoch": 0.004592,
+      "grad_norm": 1.265625,
+      "learning_rate": 5.72e-05,
+      "loss": 0.1455,
+      "step": 287
+    },
+    {
+      "epoch": 0.004608,
+      "grad_norm": 1.453125,
+      "learning_rate": 5.74e-05,
+      "loss": 0.2471,
+      "step": 288
+    },
+    {
+      "epoch": 0.004624,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.76e-05,
+      "loss": 0.218,
+      "step": 289
+    },
+    {
+      "epoch": 0.00464,
+      "grad_norm": 0.921875,
+      "learning_rate": 5.7799999999999995e-05,
+      "loss": 0.1709,
+      "step": 290
+    },
+    {
+      "epoch": 0.004656,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.8e-05,
+      "loss": 0.1917,
+      "step": 291
+    },
+    {
+      "epoch": 0.004672,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.82e-05,
+      "loss": 0.1716,
+      "step": 292
+    },
+    {
+      "epoch": 0.004688,
+      "grad_norm": 1.2421875,
+      "learning_rate": 5.8399999999999997e-05,
+      "loss": 0.1675,
+      "step": 293
+    },
+    {
+      "epoch": 0.004704,
+      "grad_norm": 1.3125,
+      "learning_rate": 5.86e-05,
+      "loss": 0.1524,
+      "step": 294
+    },
+    {
+      "epoch": 0.00472,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.88e-05,
+      "loss": 0.1235,
+      "step": 295
+    },
+    {
+      "epoch": 0.004736,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.9e-05,
+      "loss": 0.22,
+      "step": 296
+    },
+    {
+      "epoch": 0.004752,
+      "grad_norm": 1.359375,
+      "learning_rate": 5.92e-05,
+      "loss": 0.1908,
+      "step": 297
+    },
+    {
+      "epoch": 0.004768,
+      "grad_norm": 1.3359375,
+      "learning_rate": 5.94e-05,
+      "loss": 0.1698,
+      "step": 298
+    },
+    {
+      "epoch": 0.004784,
+      "grad_norm": 1.8125,
+      "learning_rate": 5.96e-05,
+      "loss": 0.2789,
+      "step": 299
+    },
+    {
+      "epoch": 0.0048,
+      "grad_norm": 1.296875,
+      "learning_rate": 5.9800000000000003e-05,
+      "loss": 0.188,
+      "step": 300
+    },
+    {
+      "epoch": 0.004816,
+      "grad_norm": 1.1328125,
+      "learning_rate": 6e-05,
+      "loss": 0.1685,
+      "step": 301
+    },
+    {
+      "epoch": 0.004832,
+      "grad_norm": 1.3125,
+      "learning_rate": 6.02e-05,
+      "loss": 0.217,
+      "step": 302
+    },
+    {
+      "epoch": 0.004848,
+      "grad_norm": 1.34375,
+      "learning_rate": 6.04e-05,
+      "loss": 0.188,
+      "step": 303
+    },
+    {
+      "epoch": 0.004864,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.06e-05,
+      "loss": 0.1696,
+      "step": 304
+    },
+    {
+      "epoch": 0.00488,
+      "grad_norm": 1.1953125,
+      "learning_rate": 6.08e-05,
+      "loss": 0.1866,
+      "step": 305
+    },
+    {
+      "epoch": 0.004896,
+      "grad_norm": 1.921875,
+      "learning_rate": 6.1e-05,
+      "loss": 0.2091,
+      "step": 306
+    },
+    {
+      "epoch": 0.004912,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.12e-05,
+      "loss": 0.2341,
+      "step": 307
+    },
+    {
+      "epoch": 0.004928,
+      "grad_norm": 1.171875,
+      "learning_rate": 6.14e-05,
+      "loss": 0.2002,
+      "step": 308
+    },
+    {
+      "epoch": 0.004944,
+      "grad_norm": 1.2890625,
+      "learning_rate": 6.16e-05,
+      "loss": 0.1621,
+      "step": 309
+    },
+    {
+      "epoch": 0.00496,
+      "grad_norm": 1.2265625,
+      "learning_rate": 6.18e-05,
+      "loss": 0.2535,
+      "step": 310
+    },
+    {
+      "epoch": 0.004976,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.2e-05,
+      "loss": 0.1658,
+      "step": 311
+    },
+    {
+      "epoch": 0.004992,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.220000000000001e-05,
+      "loss": 0.1721,
+      "step": 312
+    },
+    {
+      "epoch": 0.005008,
+      "grad_norm": 1.3046875,
+      "learning_rate": 6.24e-05,
+      "loss": 0.1947,
+      "step": 313
+    },
+    {
+      "epoch": 0.005024,
+      "grad_norm": 1.765625,
+      "learning_rate": 6.26e-05,
+      "loss": 0.2242,
+      "step": 314
+    },
+    {
+      "epoch": 0.00504,
+      "grad_norm": 1.5,
+      "learning_rate": 6.280000000000001e-05,
+      "loss": 0.2143,
+      "step": 315
+    },
+    {
+      "epoch": 0.005056,
+      "grad_norm": 1.8984375,
+      "learning_rate": 6.3e-05,
+      "loss": 0.219,
+      "step": 316
+    },
+    {
+      "epoch": 0.005072,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.32e-05,
+      "loss": 0.175,
+      "step": 317
+    },
+    {
+      "epoch": 0.005088,
+      "grad_norm": 1.4765625,
+      "learning_rate": 6.340000000000001e-05,
+      "loss": 0.2109,
+      "step": 318
+    },
+    {
+      "epoch": 0.005104,
+      "grad_norm": 1.5078125,
+      "learning_rate": 6.36e-05,
+      "loss": 0.2011,
+      "step": 319
+    },
+    {
+      "epoch": 0.00512,
+      "grad_norm": 1.234375,
+      "learning_rate": 6.38e-05,
+      "loss": 0.1508,
+      "step": 320
+    },
+    {
+      "epoch": 0.005136,
+      "grad_norm": 1.1953125,
+      "learning_rate": 6.400000000000001e-05,
+      "loss": 0.2092,
+      "step": 321
+    },
+    {
+      "epoch": 0.005152,
+      "grad_norm": 1.3828125,
+      "learning_rate": 6.42e-05,
+      "loss": 0.1644,
+      "step": 322
+    },
+    {
+      "epoch": 0.005168,
+      "grad_norm": 1.5546875,
+      "learning_rate": 6.440000000000001e-05,
+      "loss": 0.1869,
+      "step": 323
+    },
+    {
+      "epoch": 0.005184,
+      "grad_norm": 1.3828125,
+      "learning_rate": 6.460000000000001e-05,
+      "loss": 0.209,
+      "step": 324
+    },
+    {
+      "epoch": 0.0052,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.48e-05,
+      "loss": 0.2072,
+      "step": 325
+    },
+    {
+      "epoch": 0.005216,
+      "grad_norm": 0.8359375,
+      "learning_rate": 6.500000000000001e-05,
+      "loss": 0.1649,
+      "step": 326
+    },
+    {
+      "epoch": 0.005232,
+      "grad_norm": 1.2421875,
+      "learning_rate": 6.52e-05,
+      "loss": 0.1741,
+      "step": 327
+    },
+    {
+      "epoch": 0.005248,
+      "grad_norm": 1.75,
+      "learning_rate": 6.54e-05,
+      "loss": 0.2492,
+      "step": 328
+    },
+    {
+      "epoch": 0.005264,
+      "grad_norm": 1.1953125,
+      "learning_rate": 6.560000000000001e-05,
+      "loss": 0.1856,
+      "step": 329
+    },
+    {
+      "epoch": 0.00528,
+      "grad_norm": 1.265625,
+      "learning_rate": 6.58e-05,
+      "loss": 0.1559,
+      "step": 330
+    },
+    {
+      "epoch": 0.005296,
+      "grad_norm": 1.7265625,
+      "learning_rate": 6.6e-05,
+      "loss": 0.1944,
+      "step": 331
+    },
+    {
+      "epoch": 0.005312,
+      "grad_norm": 2.34375,
+      "learning_rate": 6.620000000000001e-05,
+      "loss": 0.2133,
+      "step": 332
+    },
+    {
+      "epoch": 0.005328,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.64e-05,
+      "loss": 0.1602,
+      "step": 333
+    },
+    {
+      "epoch": 0.005344,
+      "grad_norm": 3.875,
+      "learning_rate": 6.66e-05,
+      "loss": 0.2797,
+      "step": 334
+    },
+    {
+      "epoch": 0.00536,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.680000000000001e-05,
+      "loss": 0.1907,
+      "step": 335
+    },
+    {
+      "epoch": 0.005376,
+      "grad_norm": 1.390625,
+      "learning_rate": 6.7e-05,
+      "loss": 0.17,
+      "step": 336
+    },
+    {
+      "epoch": 0.005392,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.720000000000001e-05,
+      "loss": 0.1652,
+      "step": 337
+    },
+    {
+      "epoch": 0.005408,
+      "grad_norm": 1.890625,
+      "learning_rate": 6.740000000000001e-05,
+      "loss": 0.1889,
+      "step": 338
+    },
+    {
+      "epoch": 0.005424,
+      "grad_norm": 1.2578125,
+      "learning_rate": 6.76e-05,
+      "loss": 0.1817,
+      "step": 339
+    },
+    {
+      "epoch": 0.00544,
+      "grad_norm": 1.140625,
+      "learning_rate": 6.780000000000001e-05,
+      "loss": 0.1477,
+      "step": 340
+    },
+    {
+      "epoch": 0.005456,
+      "grad_norm": 1.296875,
+      "learning_rate": 6.800000000000001e-05,
+      "loss": 0.2106,
+      "step": 341
+    },
+    {
+      "epoch": 0.005472,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.82e-05,
+      "loss": 0.206,
+      "step": 342
+    },
+    {
+      "epoch": 0.005488,
+      "grad_norm": 1.4140625,
+      "learning_rate": 6.840000000000001e-05,
+      "loss": 0.2199,
+      "step": 343
+    },
+    {
+      "epoch": 0.005504,
+      "grad_norm": 1.4921875,
+      "learning_rate": 6.860000000000001e-05,
+      "loss": 0.1613,
+      "step": 344
+    },
+    {
+      "epoch": 0.00552,
+      "grad_norm": 1.390625,
+      "learning_rate": 6.879999999999999e-05,
+      "loss": 0.2386,
+      "step": 345
+    },
+    {
+      "epoch": 0.005536,
+      "grad_norm": 1.375,
+      "learning_rate": 6.9e-05,
+      "loss": 0.2058,
+      "step": 346
+    },
+    {
+      "epoch": 0.005552,
+      "grad_norm": 3.84375,
+      "learning_rate": 6.92e-05,
+      "loss": 0.2339,
+      "step": 347
+    },
+    {
+      "epoch": 0.005568,
+      "grad_norm": 1.4765625,
+      "learning_rate": 6.939999999999999e-05,
+      "loss": 0.2376,
+      "step": 348
+    },
+    {
+      "epoch": 0.005584,
+      "grad_norm": 1.3671875,
+      "learning_rate": 6.96e-05,
+      "loss": 0.2087,
+      "step": 349
+    },
+    {
+      "epoch": 0.0056,
+      "grad_norm": 1.296875,
+      "learning_rate": 6.98e-05,
+      "loss": 0.1991,
+      "step": 350
+    },
+    {
+      "epoch": 0.005616,
+      "grad_norm": 1.1875,
+      "learning_rate": 7e-05,
+      "loss": 0.2139,
+      "step": 351
+    },
+    {
+      "epoch": 0.005632,
+      "grad_norm": 1.125,
+      "learning_rate": 7.02e-05,
+      "loss": 0.2344,
+      "step": 352
+    },
+    {
+      "epoch": 0.005648,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.04e-05,
+      "loss": 0.1728,
+      "step": 353
+    },
+    {
+      "epoch": 0.005664,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.06e-05,
+      "loss": 0.2156,
+      "step": 354
+    },
+    {
+      "epoch": 0.00568,
+      "grad_norm": 1.96875,
+      "learning_rate": 7.08e-05,
+      "loss": 0.2952,
+      "step": 355
+    },
+    {
+      "epoch": 0.005696,
+      "grad_norm": 1.5234375,
+      "learning_rate": 7.1e-05,
+      "loss": 0.2214,
+      "step": 356
+    },
+    {
+      "epoch": 0.005712,
+      "grad_norm": 1.296875,
+      "learning_rate": 7.12e-05,
+      "loss": 0.193,
+      "step": 357
+    },
+    {
+      "epoch": 0.005728,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.14e-05,
+      "loss": 0.2172,
+      "step": 358
+    },
+    {
+      "epoch": 0.005744,
+      "grad_norm": 1.375,
+      "learning_rate": 7.16e-05,
+      "loss": 0.1934,
+      "step": 359
+    },
+    {
+      "epoch": 0.00576,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.18e-05,
+      "loss": 0.1801,
+      "step": 360
+    },
+    {
+      "epoch": 0.005776,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.2e-05,
+      "loss": 0.2162,
+      "step": 361
+    },
+    {
+      "epoch": 0.005792,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.22e-05,
+      "loss": 0.2739,
+      "step": 362
+    },
+    {
+      "epoch": 0.005808,
+      "grad_norm": 1.25,
+      "learning_rate": 7.24e-05,
+      "loss": 0.208,
+      "step": 363
+    },
+    {
+      "epoch": 0.005824,
+      "grad_norm": 1.4375,
+      "learning_rate": 7.26e-05,
+      "loss": 0.24,
+      "step": 364
+    },
+    {
+      "epoch": 0.00584,
+      "grad_norm": 1.25,
+      "learning_rate": 7.280000000000001e-05,
+      "loss": 0.2156,
+      "step": 365
+    },
+    {
+      "epoch": 0.005856,
+      "grad_norm": 1.125,
+      "learning_rate": 7.3e-05,
+      "loss": 0.1786,
+      "step": 366
+    },
+    {
+      "epoch": 0.005872,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.32e-05,
+      "loss": 0.1968,
+      "step": 367
+    },
+    {
+      "epoch": 0.005888,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.340000000000001e-05,
+      "loss": 0.1776,
+      "step": 368
+    },
+    {
+      "epoch": 0.005904,
+      "grad_norm": 2.0,
+      "learning_rate": 7.36e-05,
+      "loss": 0.248,
+      "step": 369
+    },
+    {
+      "epoch": 0.00592,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.38e-05,
+      "loss": 0.206,
+      "step": 370
+    },
+    {
+      "epoch": 0.005936,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.4e-05,
+      "loss": 0.1962,
+      "step": 371
+    },
+    {
+      "epoch": 0.005952,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.42e-05,
+      "loss": 0.1737,
+      "step": 372
+    },
+    {
+      "epoch": 0.005968,
+      "grad_norm": 5.25,
+      "learning_rate": 7.44e-05,
+      "loss": 0.2248,
+      "step": 373
+    },
+    {
+      "epoch": 0.005984,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.46e-05,
+      "loss": 0.1612,
+      "step": 374
+    },
+    {
+      "epoch": 0.006,
+      "grad_norm": 1.53125,
+      "learning_rate": 7.48e-05,
+      "loss": 0.1882,
+      "step": 375
+    },
+    {
+      "epoch": 0.006016,
+      "grad_norm": 1.375,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 0.2168,
+      "step": 376
+    },
+    {
+      "epoch": 0.006032,
+      "grad_norm": 3.390625,
+      "learning_rate": 7.52e-05,
+      "loss": 0.2457,
+      "step": 377
+    },
+    {
+      "epoch": 0.006048,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.54e-05,
+      "loss": 0.2265,
+      "step": 378
+    },
+    {
+      "epoch": 0.006064,
+      "grad_norm": 1.4921875,
+      "learning_rate": 7.560000000000001e-05,
+      "loss": 0.1963,
+      "step": 379
+    },
+    {
+      "epoch": 0.00608,
+      "grad_norm": 1.2421875,
+      "learning_rate": 7.58e-05,
+      "loss": 0.1976,
+      "step": 380
+    },
+    {
+      "epoch": 0.006096,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.6e-05,
+      "loss": 0.2828,
+      "step": 381
+    },
+    {
+      "epoch": 0.006112,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.620000000000001e-05,
+      "loss": 0.1964,
+      "step": 382
+    },
+    {
+      "epoch": 0.006128,
+      "grad_norm": 1.4609375,
+      "learning_rate": 7.64e-05,
+      "loss": 0.2206,
+      "step": 383
+    },
+    {
+      "epoch": 0.006144,
+      "grad_norm": 4.59375,
+      "learning_rate": 7.66e-05,
+      "loss": 0.2297,
+      "step": 384
+    },
+    {
+      "epoch": 0.00616,
+      "grad_norm": 1.484375,
+      "learning_rate": 7.680000000000001e-05,
+      "loss": 0.2095,
+      "step": 385
+    },
+    {
+      "epoch": 0.006176,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.7e-05,
+      "loss": 0.2175,
+      "step": 386
+    },
+    {
+      "epoch": 0.006192,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.72e-05,
+      "loss": 0.1973,
+      "step": 387
+    },
+    {
+      "epoch": 0.006208,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.740000000000001e-05,
+      "loss": 0.2281,
+      "step": 388
+    },
+    {
+      "epoch": 0.006224,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.76e-05,
+      "loss": 0.266,
+      "step": 389
+    },
+    {
+      "epoch": 0.00624,
+      "grad_norm": 1.7421875,
+      "learning_rate": 7.780000000000001e-05,
+      "loss": 0.2339,
+      "step": 390
+    },
+    {
+      "epoch": 0.006256,
+      "grad_norm": 1.578125,
+      "learning_rate": 7.800000000000001e-05,
+      "loss": 0.2287,
+      "step": 391
+    },
+    {
+      "epoch": 0.006272,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.82e-05,
+      "loss": 0.2325,
+      "step": 392
+    },
+    {
+      "epoch": 0.006288,
+      "grad_norm": 1.984375,
+      "learning_rate": 7.840000000000001e-05,
+      "loss": 0.2713,
+      "step": 393
+    },
+    {
+      "epoch": 0.006304,
+      "grad_norm": 1.3046875,
+      "learning_rate": 7.860000000000001e-05,
+      "loss": 0.1938,
+      "step": 394
+    },
+    {
+      "epoch": 0.00632,
+      "grad_norm": 1.96875,
+      "learning_rate": 7.88e-05,
+      "loss": 0.2365,
+      "step": 395
+    },
+    {
+      "epoch": 0.006336,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.900000000000001e-05,
+      "loss": 0.2007,
+      "step": 396
+    },
+    {
+      "epoch": 0.006352,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.920000000000001e-05,
+      "loss": 0.1852,
+      "step": 397
+    },
+    {
+      "epoch": 0.006368,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.94e-05,
+      "loss": 0.245,
+      "step": 398
+    },
+    {
+      "epoch": 0.006384,
+      "grad_norm": 1.4609375,
+      "learning_rate": 7.960000000000001e-05,
+      "loss": 0.2187,
+      "step": 399
+    },
+    {
+      "epoch": 0.0064,
+      "grad_norm": 1.359375,
+      "learning_rate": 7.98e-05,
+      "loss": 0.1844,
+      "step": 400
+    },
+    {
+      "epoch": 0.006416,
+      "grad_norm": 1.296875,
+      "learning_rate": 8e-05,
+      "loss": 0.238,
+      "step": 401
+    },
+    {
+      "epoch": 0.006432,
+      "grad_norm": 1.8984375,
+      "learning_rate": 8.020000000000001e-05,
+      "loss": 0.1917,
+      "step": 402
+    },
+    {
+      "epoch": 0.006448,
+      "grad_norm": 1.5390625,
+      "learning_rate": 8.04e-05,
+      "loss": 0.1864,
+      "step": 403
+    },
+    {
+      "epoch": 0.006464,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.060000000000001e-05,
+      "loss": 0.2462,
+      "step": 404
+    },
+    {
+      "epoch": 0.00648,
+      "grad_norm": 1.5625,
+      "learning_rate": 8.080000000000001e-05,
+      "loss": 0.2428,
+      "step": 405
+    },
+    {
+      "epoch": 0.006496,
+      "grad_norm": 3.0,
+      "learning_rate": 8.1e-05,
+      "loss": 0.2489,
+      "step": 406
+    },
+    {
+      "epoch": 0.006512,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.120000000000001e-05,
+      "loss": 0.2184,
+      "step": 407
+    },
+    {
+      "epoch": 0.006528,
+      "grad_norm": 1.25,
+      "learning_rate": 8.14e-05,
+      "loss": 0.2057,
+      "step": 408
+    },
+    {
+      "epoch": 0.006544,
+      "grad_norm": 1.125,
+      "learning_rate": 8.16e-05,
+      "loss": 0.167,
+      "step": 409
+    },
+    {
+      "epoch": 0.00656,
+      "grad_norm": 1.7421875,
+      "learning_rate": 8.18e-05,
+      "loss": 0.1764,
+      "step": 410
+    },
+    {
+      "epoch": 0.006576,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.2e-05,
+      "loss": 0.1765,
+      "step": 411
+    },
+    {
+      "epoch": 0.006592,
+      "grad_norm": 1.28125,
+      "learning_rate": 8.22e-05,
+      "loss": 0.2193,
+      "step": 412
+    },
+    {
+      "epoch": 0.006608,
+      "grad_norm": 1.3203125,
+      "learning_rate": 8.24e-05,
+      "loss": 0.2179,
+      "step": 413
+    },
+    {
+      "epoch": 0.006624,
+      "grad_norm": 1.53125,
+      "learning_rate": 8.26e-05,
+      "loss": 0.2155,
+      "step": 414
+    },
+    {
+      "epoch": 0.00664,
+      "grad_norm": 1.5,
+      "learning_rate": 8.28e-05,
+      "loss": 0.2436,
+      "step": 415
+    },
+    {
+      "epoch": 0.006656,
+      "grad_norm": 1.328125,
+      "learning_rate": 8.3e-05,
+      "loss": 0.2089,
+      "step": 416
+    },
+    {
+      "epoch": 0.006672,
+      "grad_norm": 1.828125,
+      "learning_rate": 8.32e-05,
+      "loss": 0.2113,
+      "step": 417
+    },
+    {
+      "epoch": 0.006688,
+      "grad_norm": 1.765625,
+      "learning_rate": 8.34e-05,
+      "loss": 0.2776,
+      "step": 418
+    },
+    {
+      "epoch": 0.006704,
+      "grad_norm": 1.4453125,
+      "learning_rate": 8.36e-05,
+      "loss": 0.2434,
+      "step": 419
+    },
+    {
+      "epoch": 0.00672,
+      "grad_norm": 1.34375,
+      "learning_rate": 8.38e-05,
+      "loss": 0.2021,
+      "step": 420
+    },
+    {
+      "epoch": 0.006736,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.4e-05,
+      "loss": 0.2034,
+      "step": 421
+    },
+    {
+      "epoch": 0.006752,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.42e-05,
+      "loss": 0.2494,
+      "step": 422
+    },
+    {
+      "epoch": 0.006768,
+      "grad_norm": 2.0,
+      "learning_rate": 8.44e-05,
+      "loss": 0.3055,
+      "step": 423
+    },
+    {
+      "epoch": 0.006784,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.46e-05,
+      "loss": 0.2416,
+      "step": 424
+    },
+    {
+      "epoch": 0.0068,
+      "grad_norm": 1.453125,
+      "learning_rate": 8.48e-05,
+      "loss": 0.2095,
+      "step": 425
+    },
+    {
+      "epoch": 0.006816,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.5e-05,
+      "loss": 0.2085,
+      "step": 426
+    },
+    {
+      "epoch": 0.006832,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.52e-05,
+      "loss": 0.1958,
+      "step": 427
+    },
+    {
+      "epoch": 0.006848,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.54e-05,
+      "loss": 0.2087,
+      "step": 428
+    },
+    {
+      "epoch": 0.006864,
+      "grad_norm": 1.3984375,
+      "learning_rate": 8.560000000000001e-05,
+      "loss": 0.2584,
+      "step": 429
+    },
+    {
+      "epoch": 0.00688,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.58e-05,
+      "loss": 0.239,
+      "step": 430
+    },
+    {
+      "epoch": 0.006896,
+      "grad_norm": 1.421875,
+      "learning_rate": 8.6e-05,
+      "loss": 0.1848,
+      "step": 431
+    },
+    {
+      "epoch": 0.006912,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.620000000000001e-05,
+      "loss": 0.2342,
+      "step": 432
+    },
+    {
+      "epoch": 0.006928,
+      "grad_norm": 1.375,
+      "learning_rate": 8.64e-05,
+      "loss": 0.1857,
+      "step": 433
+    },
+    {
+      "epoch": 0.006944,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.66e-05,
+      "loss": 0.2663,
+      "step": 434
+    },
+    {
+      "epoch": 0.00696,
+      "grad_norm": 1.484375,
+      "learning_rate": 8.680000000000001e-05,
+      "loss": 0.272,
+      "step": 435
+    },
+    {
+      "epoch": 0.006976,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.7e-05,
+      "loss": 0.2859,
+      "step": 436
+    },
+    {
+      "epoch": 0.006992,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.72e-05,
+      "loss": 0.2416,
+      "step": 437
+    },
+    {
+      "epoch": 0.007008,
+      "grad_norm": 1.328125,
+      "learning_rate": 8.740000000000001e-05,
+      "loss": 0.2192,
+      "step": 438
+    },
+    {
+      "epoch": 0.007024,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.76e-05,
+      "loss": 0.2133,
+      "step": 439
+    },
+    {
+      "epoch": 0.00704,
+      "grad_norm": 1.34375,
+      "learning_rate": 8.78e-05,
+      "loss": 0.2125,
+      "step": 440
+    },
+    {
+      "epoch": 0.007056,
+      "grad_norm": 1.328125,
+      "learning_rate": 8.800000000000001e-05,
+      "loss": 0.2122,
+      "step": 441
+    },
+    {
+      "epoch": 0.007072,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.82e-05,
+      "loss": 0.168,
+      "step": 442
+    },
+    {
+      "epoch": 0.007088,
+      "grad_norm": 1.484375,
+      "learning_rate": 8.840000000000001e-05,
+      "loss": 0.2246,
+      "step": 443
+    },
+    {
+      "epoch": 0.007104,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.86e-05,
+      "loss": 0.1966,
+      "step": 444
+    },
+    {
+      "epoch": 0.00712,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.88e-05,
+      "loss": 0.2148,
+      "step": 445
+    },
+    {
+      "epoch": 0.007136,
+      "grad_norm": 1.296875,
+      "learning_rate": 8.900000000000001e-05,
+      "loss": 0.2048,
+      "step": 446
+    },
+    {
+      "epoch": 0.007152,
+      "grad_norm": 1.859375,
+      "learning_rate": 8.92e-05,
+      "loss": 0.1985,
+      "step": 447
+    },
+    {
+      "epoch": 0.007168,
+      "grad_norm": 1.234375,
+      "learning_rate": 8.94e-05,
+      "loss": 0.1884,
+      "step": 448
+    },
+    {
+      "epoch": 0.007184,
+      "grad_norm": 1.5390625,
+      "learning_rate": 8.960000000000001e-05,
+      "loss": 0.2292,
+      "step": 449
+    },
+    {
+      "epoch": 0.0072,
+      "grad_norm": 1.796875,
+      "learning_rate": 8.98e-05,
+      "loss": 0.25,
+      "step": 450
+    },
+    {
+      "epoch": 0.007216,
+      "grad_norm": 1.4453125,
+      "learning_rate": 9e-05,
+      "loss": 0.2437,
+      "step": 451
+    },
+    {
+      "epoch": 0.007232,
+      "grad_norm": 2.125,
+      "learning_rate": 9.020000000000001e-05,
+      "loss": 0.2758,
+      "step": 452
+    },
+    {
+      "epoch": 0.007248,
+      "grad_norm": 1.625,
+      "learning_rate": 9.04e-05,
+      "loss": 0.2387,
+      "step": 453
+    },
+    {
+      "epoch": 0.007264,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.06e-05,
+      "loss": 0.198,
+      "step": 454
+    },
+    {
+      "epoch": 0.00728,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.080000000000001e-05,
+      "loss": 0.2135,
+      "step": 455
+    },
+    {
+      "epoch": 0.007296,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.1e-05,
+      "loss": 0.2291,
+      "step": 456
+    },
+    {
+      "epoch": 0.007312,
+      "grad_norm": 1.4296875,
+      "learning_rate": 9.120000000000001e-05,
+      "loss": 0.2706,
+      "step": 457
+    },
+    {
+      "epoch": 0.007328,
+      "grad_norm": 1.46875,
+      "learning_rate": 9.140000000000001e-05,
+      "loss": 0.211,
+      "step": 458
+    },
+    {
+      "epoch": 0.007344,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.16e-05,
+      "loss": 0.2634,
+      "step": 459
+    },
+    {
+      "epoch": 0.00736,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.180000000000001e-05,
+      "loss": 0.25,
+      "step": 460
+    },
+    {
+      "epoch": 0.007376,
+      "grad_norm": 1.5390625,
+      "learning_rate": 9.200000000000001e-05,
+      "loss": 0.2523,
+      "step": 461
+    },
+    {
+      "epoch": 0.007392,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.22e-05,
+      "loss": 0.1944,
+      "step": 462
+    },
+    {
+      "epoch": 0.007408,
+      "grad_norm": 1.421875,
+      "learning_rate": 9.240000000000001e-05,
+      "loss": 0.2117,
+      "step": 463
+    },
+    {
+      "epoch": 0.007424,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.260000000000001e-05,
+      "loss": 0.2227,
+      "step": 464
+    },
+    {
+      "epoch": 0.00744,
+      "grad_norm": 1.390625,
+      "learning_rate": 9.28e-05,
+      "loss": 0.2362,
+      "step": 465
+    },
+    {
+      "epoch": 0.007456,
+      "grad_norm": 1.7578125,
+      "learning_rate": 9.300000000000001e-05,
+      "loss": 0.2246,
+      "step": 466
+    },
+    {
+      "epoch": 0.007472,
+      "grad_norm": 1.25,
+      "learning_rate": 9.320000000000002e-05,
+      "loss": 0.1741,
+      "step": 467
+    },
+    {
+      "epoch": 0.007488,
+      "grad_norm": 2.125,
+      "learning_rate": 9.340000000000001e-05,
+      "loss": 0.2058,
+      "step": 468
+    },
+    {
+      "epoch": 0.007504,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.360000000000001e-05,
+      "loss": 0.2051,
+      "step": 469
+    },
+    {
+      "epoch": 0.00752,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.38e-05,
+      "loss": 0.2042,
+      "step": 470
+    },
+    {
+      "epoch": 0.007536,
+      "grad_norm": 3.5,
+      "learning_rate": 9.4e-05,
+      "loss": 0.2497,
+      "step": 471
+    },
+    {
+      "epoch": 0.007552,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.42e-05,
+      "loss": 0.2009,
+      "step": 472
+    },
+    {
+      "epoch": 0.007568,
+      "grad_norm": 1.4296875,
+      "learning_rate": 9.44e-05,
+      "loss": 0.1929,
+      "step": 473
+    },
+    {
+      "epoch": 0.007584,
+      "grad_norm": 1.4609375,
+      "learning_rate": 9.46e-05,
+      "loss": 0.1984,
+      "step": 474
+    },
+    {
+      "epoch": 0.0076,
+      "grad_norm": 1.890625,
+      "learning_rate": 9.48e-05,
+      "loss": 0.2615,
+      "step": 475
+    },
+    {
+      "epoch": 0.007616,
+      "grad_norm": 2.625,
+      "learning_rate": 9.5e-05,
+      "loss": 0.2284,
+      "step": 476
+    },
+    {
+      "epoch": 0.007632,
+      "grad_norm": 1.5234375,
+      "learning_rate": 9.52e-05,
+      "loss": 0.2151,
+      "step": 477
+    },
+    {
+      "epoch": 0.007648,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.54e-05,
+      "loss": 0.2091,
+      "step": 478
+    },
+    {
+      "epoch": 0.007664,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.56e-05,
+      "loss": 0.2084,
+      "step": 479
+    },
+    {
+      "epoch": 0.00768,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.58e-05,
+      "loss": 0.2107,
+      "step": 480
+    },
+    {
+      "epoch": 0.007696,
+      "grad_norm": 1.390625,
+      "learning_rate": 9.6e-05,
+      "loss": 0.2514,
+      "step": 481
+    },
+    {
+      "epoch": 0.007712,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.620000000000001e-05,
+      "loss": 0.1967,
+      "step": 482
+    },
+    {
+      "epoch": 0.007728,
+      "grad_norm": 1.7890625,
+      "learning_rate": 9.64e-05,
+      "loss": 0.2129,
+      "step": 483
+    },
+    {
+      "epoch": 0.007744,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.66e-05,
+      "loss": 0.2275,
+      "step": 484
+    },
+    {
+      "epoch": 0.00776,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.680000000000001e-05,
+      "loss": 0.1816,
+      "step": 485
+    },
+    {
+      "epoch": 0.007776,
+      "grad_norm": 1.296875,
+      "learning_rate": 9.7e-05,
+      "loss": 0.2074,
+      "step": 486
+    },
+    {
+      "epoch": 0.007792,
+      "grad_norm": 1.4765625,
+      "learning_rate": 9.72e-05,
+      "loss": 0.2636,
+      "step": 487
+    },
+    {
+      "epoch": 0.007808,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.74e-05,
+      "loss": 0.193,
+      "step": 488
+    },
+    {
+      "epoch": 0.007824,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.76e-05,
+      "loss": 0.2516,
+      "step": 489
+    },
+    {
+      "epoch": 0.00784,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.78e-05,
+      "loss": 0.2639,
+      "step": 490
+    },
+    {
+      "epoch": 0.007856,
+      "grad_norm": 1.796875,
+      "learning_rate": 9.8e-05,
+      "loss": 0.2206,
+      "step": 491
+    },
+    {
+      "epoch": 0.007872,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.82e-05,
+      "loss": 0.2616,
+      "step": 492
+    },
+    {
+      "epoch": 0.007888,
+      "grad_norm": 1.3984375,
+      "learning_rate": 9.84e-05,
+      "loss": 0.239,
+      "step": 493
+    },
+    {
+      "epoch": 0.007904,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.86e-05,
+      "loss": 0.2132,
+      "step": 494
+    },
+    {
+      "epoch": 0.00792,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.88e-05,
+      "loss": 0.2131,
+      "step": 495
+    },
+    {
+      "epoch": 0.007936,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.900000000000001e-05,
+      "loss": 0.2324,
+      "step": 496
+    },
+    {
+      "epoch": 0.007952,
+      "grad_norm": 3.640625,
+      "learning_rate": 9.92e-05,
+      "loss": 0.2457,
+      "step": 497
+    },
+    {
+      "epoch": 0.007968,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.94e-05,
+      "loss": 0.2534,
+      "step": 498
+    },
+    {
+      "epoch": 0.007984,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.960000000000001e-05,
+      "loss": 0.2008,
+      "step": 499
+    },
+    {
+      "epoch": 0.008,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.98e-05,
+      "loss": 0.2296,
+      "step": 500
+    },
+    {
+      "epoch": 0.008016,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001,
+      "loss": 0.241,
+      "step": 501
+    },
+    {
+      "epoch": 0.008032,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.999838709677419e-05,
+      "loss": 0.1918,
+      "step": 502
+    },
+    {
+      "epoch": 0.008048,
+      "grad_norm": 1.8046875,
+      "learning_rate": 9.999677419354839e-05,
+      "loss": 0.2266,
+      "step": 503
+    },
+    {
+      "epoch": 0.008064,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.999516129032258e-05,
+      "loss": 0.2164,
+      "step": 504
+    },
+    {
+      "epoch": 0.00808,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.999354838709678e-05,
+      "loss": 0.2183,
+      "step": 505
+    },
+    {
+      "epoch": 0.008096,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.999193548387096e-05,
+      "loss": 0.1946,
+      "step": 506
+    },
+    {
+      "epoch": 0.008112,
+      "grad_norm": 1.625,
+      "learning_rate": 9.999032258064516e-05,
+      "loss": 0.2632,
+      "step": 507
+    },
+    {
+      "epoch": 0.008128,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.998870967741936e-05,
+      "loss": 0.2165,
+      "step": 508
+    },
+    {
+      "epoch": 0.008144,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.998709677419356e-05,
+      "loss": 0.2158,
+      "step": 509
+    },
+    {
+      "epoch": 0.00816,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.998548387096775e-05,
+      "loss": 0.199,
+      "step": 510
+    },
+    {
+      "epoch": 0.008176,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.998387096774195e-05,
+      "loss": 0.1898,
+      "step": 511
+    },
+    {
+      "epoch": 0.008192,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.998225806451613e-05,
+      "loss": 0.2538,
+      "step": 512
+    },
+    {
+      "epoch": 0.008208,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.998064516129033e-05,
+      "loss": 0.1981,
+      "step": 513
+    },
+    {
+      "epoch": 0.008224,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.997903225806452e-05,
+      "loss": 0.2577,
+      "step": 514
+    },
+    {
+      "epoch": 0.00824,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.997741935483872e-05,
+      "loss": 0.215,
+      "step": 515
+    },
+    {
+      "epoch": 0.008256,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.99758064516129e-05,
+      "loss": 0.2181,
+      "step": 516
+    },
+    {
+      "epoch": 0.008272,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.997419354838709e-05,
+      "loss": 0.273,
+      "step": 517
+    },
+    {
+      "epoch": 0.008288,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.997258064516129e-05,
+      "loss": 0.2044,
+      "step": 518
+    },
+    {
+      "epoch": 0.008304,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.997096774193549e-05,
+      "loss": 0.2378,
+      "step": 519
+    },
+    {
+      "epoch": 0.00832,
+      "grad_norm": 1.3359375,
+      "learning_rate": 9.996935483870969e-05,
+      "loss": 0.2251,
+      "step": 520
+    },
+    {
+      "epoch": 0.008336,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.996774193548388e-05,
+      "loss": 0.2481,
+      "step": 521
+    },
+    {
+      "epoch": 0.008352,
+      "grad_norm": 1.4453125,
+      "learning_rate": 9.996612903225808e-05,
+      "loss": 0.2398,
+      "step": 522
+    },
+    {
+      "epoch": 0.008368,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.996451612903226e-05,
+      "loss": 0.2458,
+      "step": 523
+    },
+    {
+      "epoch": 0.008384,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.996290322580646e-05,
+      "loss": 0.2507,
+      "step": 524
+    },
+    {
+      "epoch": 0.0084,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.996129032258065e-05,
+      "loss": 0.257,
+      "step": 525
+    },
+    {
+      "epoch": 0.008416,
+      "grad_norm": 1.484375,
+      "learning_rate": 9.995967741935485e-05,
+      "loss": 0.2593,
+      "step": 526
+    },
+    {
+      "epoch": 0.008432,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.995806451612903e-05,
+      "loss": 0.2966,
+      "step": 527
+    },
+    {
+      "epoch": 0.008448,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.995645161290323e-05,
+      "loss": 0.2117,
+      "step": 528
+    },
+    {
+      "epoch": 0.008464,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.995483870967742e-05,
+      "loss": 0.2072,
+      "step": 529
+    },
+    {
+      "epoch": 0.00848,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.995322580645162e-05,
+      "loss": 0.1685,
+      "step": 530
+    },
+    {
+      "epoch": 0.008496,
+      "grad_norm": 1.40625,
+      "learning_rate": 9.995161290322582e-05,
+      "loss": 0.2181,
+      "step": 531
+    },
+    {
+      "epoch": 0.008512,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.995e-05,
+      "loss": 0.271,
+      "step": 532
+    },
+    {
+      "epoch": 0.008528,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.99483870967742e-05,
+      "loss": 0.2361,
+      "step": 533
+    },
+    {
+      "epoch": 0.008544,
+      "grad_norm": 1.4453125,
+      "learning_rate": 9.994677419354839e-05,
+      "loss": 0.2176,
+      "step": 534
+    },
+    {
+      "epoch": 0.00856,
+      "grad_norm": 1.4765625,
+      "learning_rate": 9.994516129032259e-05,
+      "loss": 0.2223,
+      "step": 535
+    },
+    {
+      "epoch": 0.008576,
+      "grad_norm": 1.375,
+      "learning_rate": 9.994354838709677e-05,
+      "loss": 0.2274,
+      "step": 536
+    },
+    {
+      "epoch": 0.008592,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.994193548387097e-05,
+      "loss": 0.2121,
+      "step": 537
+    },
+    {
+      "epoch": 0.008608,
+      "grad_norm": 1.8671875,
+      "learning_rate": 9.994032258064516e-05,
+      "loss": 0.2049,
+      "step": 538
+    },
+    {
+      "epoch": 0.008624,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.993870967741936e-05,
+      "loss": 0.205,
+      "step": 539
+    },
+    {
+      "epoch": 0.00864,
+      "grad_norm": 1.4296875,
+      "learning_rate": 9.993709677419355e-05,
+      "loss": 0.1844,
+      "step": 540
+    },
+    {
+      "epoch": 0.008656,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.993548387096775e-05,
+      "loss": 0.2195,
+      "step": 541
+    },
+    {
+      "epoch": 0.008672,
+      "grad_norm": 1.3125,
+      "learning_rate": 9.993387096774193e-05,
+      "loss": 0.2225,
+      "step": 542
+    },
+    {
+      "epoch": 0.008688,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.993225806451613e-05,
+      "loss": 0.2459,
+      "step": 543
+    },
+    {
+      "epoch": 0.008704,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.993064516129033e-05,
+      "loss": 0.22,
+      "step": 544
+    },
+    {
+      "epoch": 0.00872,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.992903225806453e-05,
+      "loss": 0.2087,
+      "step": 545
+    },
+    {
+      "epoch": 0.008736,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.992741935483872e-05,
+      "loss": 0.1843,
+      "step": 546
+    },
+    {
+      "epoch": 0.008752,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.992580645161292e-05,
+      "loss": 0.226,
+      "step": 547
+    },
+    {
+      "epoch": 0.008768,
+      "grad_norm": 1.40625,
+      "learning_rate": 9.99241935483871e-05,
+      "loss": 0.2528,
+      "step": 548
+    },
+    {
+      "epoch": 0.008784,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.992258064516129e-05,
+      "loss": 0.2189,
+      "step": 549
+    },
+    {
+      "epoch": 0.0088,
+      "grad_norm": 1.328125,
+      "learning_rate": 9.992096774193549e-05,
+      "loss": 0.2097,
+      "step": 550
+    },
+    {
+      "epoch": 0.008816,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.991935483870967e-05,
+      "loss": 0.2091,
+      "step": 551
+    },
+    {
+      "epoch": 0.008832,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.991774193548387e-05,
+      "loss": 0.2194,
+      "step": 552
+    },
+    {
+      "epoch": 0.008848,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.991612903225806e-05,
+      "loss": 0.2161,
+      "step": 553
+    },
+    {
+      "epoch": 0.008864,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.991451612903226e-05,
+      "loss": 0.2358,
+      "step": 554
+    },
+    {
+      "epoch": 0.00888,
+      "grad_norm": 1.8984375,
+      "learning_rate": 9.991290322580646e-05,
+      "loss": 0.2606,
+      "step": 555
+    },
+    {
+      "epoch": 0.008896,
+      "grad_norm": 1.90625,
+      "learning_rate": 9.991129032258066e-05,
+      "loss": 0.3322,
+      "step": 556
+    },
+    {
+      "epoch": 0.008912,
+      "grad_norm": 1.421875,
+      "learning_rate": 9.990967741935485e-05,
+      "loss": 0.2521,
+      "step": 557
+    },
+    {
+      "epoch": 0.008928,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.990806451612904e-05,
+      "loss": 0.1948,
+      "step": 558
+    },
+    {
+      "epoch": 0.008944,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.990645161290323e-05,
+      "loss": 0.2403,
+      "step": 559
+    },
+    {
+      "epoch": 0.00896,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.990483870967743e-05,
+      "loss": 0.2228,
+      "step": 560
+    },
+    {
+      "epoch": 0.008976,
+      "grad_norm": 1.625,
+      "learning_rate": 9.990322580645162e-05,
+      "loss": 0.2595,
+      "step": 561
+    },
+    {
+      "epoch": 0.008992,
+      "grad_norm": 6.9375,
+      "learning_rate": 9.990161290322582e-05,
+      "loss": 0.2991,
+      "step": 562
+    },
+    {
+      "epoch": 0.009008,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.99e-05,
+      "loss": 0.2297,
+      "step": 563
+    },
+    {
+      "epoch": 0.009024,
+      "grad_norm": 3.578125,
+      "learning_rate": 9.989838709677419e-05,
+      "loss": 0.2649,
+      "step": 564
+    },
+    {
+      "epoch": 0.00904,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.989677419354839e-05,
+      "loss": 0.3551,
+      "step": 565
+    },
+    {
+      "epoch": 0.009056,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.989516129032257e-05,
+      "loss": 0.2454,
+      "step": 566
+    },
+    {
+      "epoch": 0.009072,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.989354838709677e-05,
+      "loss": 0.1993,
+      "step": 567
+    },
+    {
+      "epoch": 0.009088,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.989193548387097e-05,
+      "loss": 0.1984,
+      "step": 568
+    },
+    {
+      "epoch": 0.009104,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.989032258064517e-05,
+      "loss": 0.2268,
+      "step": 569
+    },
+    {
+      "epoch": 0.00912,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.988870967741936e-05,
+      "loss": 0.2191,
+      "step": 570
+    },
+    {
+      "epoch": 0.009136,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.988709677419356e-05,
+      "loss": 0.2279,
+      "step": 571
+    },
+    {
+      "epoch": 0.009152,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.988548387096774e-05,
+      "loss": 0.2642,
+      "step": 572
+    },
+    {
+      "epoch": 0.009168,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.988387096774194e-05,
+      "loss": 0.2063,
+      "step": 573
+    },
+    {
+      "epoch": 0.009184,
+      "grad_norm": 1.34375,
+      "learning_rate": 9.988225806451613e-05,
+      "loss": 0.2621,
+      "step": 574
+    },
+    {
+      "epoch": 0.0092,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.988064516129033e-05,
+      "loss": 0.2461,
+      "step": 575
+    },
+    {
+      "epoch": 0.009216,
+      "grad_norm": 1.8203125,
+      "learning_rate": 9.987903225806452e-05,
+      "loss": 0.2417,
+      "step": 576
+    },
+    {
+      "epoch": 0.009232,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.987741935483872e-05,
+      "loss": 0.3172,
+      "step": 577
+    },
+    {
+      "epoch": 0.009248,
+      "grad_norm": 1.625,
+      "learning_rate": 9.98758064516129e-05,
+      "loss": 0.2045,
+      "step": 578
+    },
+    {
+      "epoch": 0.009264,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.98741935483871e-05,
+      "loss": 0.2102,
+      "step": 579
+    },
+    {
+      "epoch": 0.00928,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.98725806451613e-05,
+      "loss": 0.2546,
+      "step": 580
+    },
+    {
+      "epoch": 0.009296,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.987096774193549e-05,
+      "loss": 0.2855,
+      "step": 581
+    },
+    {
+      "epoch": 0.009312,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.986935483870969e-05,
+      "loss": 0.1888,
+      "step": 582
+    },
+    {
+      "epoch": 0.009328,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.986774193548387e-05,
+      "loss": 0.245,
+      "step": 583
+    },
+    {
+      "epoch": 0.009344,
+      "grad_norm": 1.0,
+      "learning_rate": 9.986612903225807e-05,
+      "loss": 0.1977,
+      "step": 584
+    },
+    {
+      "epoch": 0.00936,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.986451612903226e-05,
+      "loss": 0.239,
+      "step": 585
+    },
+    {
+      "epoch": 0.009376,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.986290322580646e-05,
+      "loss": 0.2344,
+      "step": 586
+    },
+    {
+      "epoch": 0.009392,
+      "grad_norm": 1.4765625,
+      "learning_rate": 9.986129032258064e-05,
+      "loss": 0.2706,
+      "step": 587
+    },
+    {
+      "epoch": 0.009408,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.985967741935484e-05,
+      "loss": 0.1928,
+      "step": 588
+    },
+    {
+      "epoch": 0.009424,
+      "grad_norm": 1.375,
+      "learning_rate": 9.985806451612903e-05,
+      "loss": 0.2447,
+      "step": 589
+    },
+    {
+      "epoch": 0.00944,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.985645161290323e-05,
+      "loss": 0.2096,
+      "step": 590
+    },
+    {
+      "epoch": 0.009456,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.985483870967743e-05,
+      "loss": 0.1623,
+      "step": 591
+    },
+    {
+      "epoch": 0.009472,
+      "grad_norm": 1.296875,
+      "learning_rate": 9.985322580645163e-05,
+      "loss": 0.2591,
+      "step": 592
+    },
+    {
+      "epoch": 0.009488,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.985161290322582e-05,
+      "loss": 0.2334,
+      "step": 593
+    },
+    {
+      "epoch": 0.009504,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.985000000000001e-05,
+      "loss": 0.2411,
+      "step": 594
+    },
+    {
+      "epoch": 0.00952,
+      "grad_norm": 1.125,
+      "learning_rate": 9.98483870967742e-05,
+      "loss": 0.1777,
+      "step": 595
+    },
+    {
+      "epoch": 0.009536,
+      "grad_norm": 1.4140625,
+      "learning_rate": 9.984677419354839e-05,
+      "loss": 0.1899,
+      "step": 596
+    },
+    {
+      "epoch": 0.009552,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.984516129032259e-05,
+      "loss": 0.2171,
+      "step": 597
+    },
+    {
+      "epoch": 0.009568,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.984354838709677e-05,
+      "loss": 0.2364,
+      "step": 598
+    },
+    {
+      "epoch": 0.009584,
+      "grad_norm": 1.5390625,
+      "learning_rate": 9.984193548387097e-05,
+      "loss": 0.2572,
+      "step": 599
+    },
+    {
+      "epoch": 0.0096,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.984032258064516e-05,
+      "loss": 0.2171,
+      "step": 600
+    },
+    {
+      "epoch": 0.009616,
+      "grad_norm": 1.484375,
+      "learning_rate": 9.983870967741936e-05,
+      "loss": 0.1931,
+      "step": 601
+    },
+    {
+      "epoch": 0.009632,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.983709677419354e-05,
+      "loss": 0.1991,
+      "step": 602
+    },
+    {
+      "epoch": 0.009648,
+      "grad_norm": 1.3515625,
+      "learning_rate": 9.983548387096774e-05,
+      "loss": 0.2195,
+      "step": 603
+    },
+    {
+      "epoch": 0.009664,
+      "grad_norm": 1.8359375,
+      "learning_rate": 9.983387096774194e-05,
+      "loss": 0.1975,
+      "step": 604
+    },
+    {
+      "epoch": 0.00968,
+      "grad_norm": 1.625,
+      "learning_rate": 9.983225806451614e-05,
+      "loss": 0.2477,
+      "step": 605
+    },
+    {
+      "epoch": 0.009696,
+      "grad_norm": 1.9609375,
+      "learning_rate": 9.983064516129033e-05,
+      "loss": 0.239,
+      "step": 606
+    },
+    {
+      "epoch": 0.009712,
+      "grad_norm": 1.8671875,
+      "learning_rate": 9.982903225806453e-05,
+      "loss": 0.2338,
+      "step": 607
+    },
+    {
+      "epoch": 0.009728,
+      "grad_norm": 1.3359375,
+      "learning_rate": 9.982741935483871e-05,
+      "loss": 0.249,
+      "step": 608
+    },
+    {
+      "epoch": 0.009744,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.982580645161291e-05,
+      "loss": 0.213,
+      "step": 609
+    },
+    {
+      "epoch": 0.00976,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.98241935483871e-05,
+      "loss": 0.2249,
+      "step": 610
+    },
+    {
+      "epoch": 0.009776,
+      "grad_norm": 1.359375,
+      "learning_rate": 9.982258064516129e-05,
+      "loss": 0.2171,
+      "step": 611
+    },
+    {
+      "epoch": 0.009792,
+      "grad_norm": 1.40625,
+      "learning_rate": 9.982096774193549e-05,
+      "loss": 0.2302,
+      "step": 612
+    },
+    {
+      "epoch": 0.009808,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.981935483870967e-05,
+      "loss": 0.211,
+      "step": 613
+    },
+    {
+      "epoch": 0.009824,
+      "grad_norm": 1.3359375,
+      "learning_rate": 9.981774193548387e-05,
+      "loss": 0.2323,
+      "step": 614
+    },
+    {
+      "epoch": 0.00984,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.981612903225807e-05,
+      "loss": 0.2652,
+      "step": 615
+    },
+    {
+      "epoch": 0.009856,
+      "grad_norm": 1.3984375,
+      "learning_rate": 9.981451612903227e-05,
+      "loss": 0.2248,
+      "step": 616
+    },
+    {
+      "epoch": 0.009872,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.981290322580646e-05,
+      "loss": 0.2206,
+      "step": 617
+    },
+    {
+      "epoch": 0.009888,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.981129032258066e-05,
+      "loss": 0.2036,
+      "step": 618
+    },
+    {
+      "epoch": 0.009904,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.980967741935484e-05,
+      "loss": 0.2556,
+      "step": 619
+    },
+    {
+      "epoch": 0.00992,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.980806451612904e-05,
+      "loss": 0.23,
+      "step": 620
+    },
+    {
+      "epoch": 0.009936,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.980645161290323e-05,
+      "loss": 0.1749,
+      "step": 621
+    },
+    {
+      "epoch": 0.009952,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.980483870967743e-05,
+      "loss": 0.2035,
+      "step": 622
+    },
+    {
+      "epoch": 0.009968,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.980322580645161e-05,
+      "loss": 0.2004,
+      "step": 623
+    },
+    {
+      "epoch": 0.009984,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.980161290322581e-05,
+      "loss": 0.2038,
+      "step": 624
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.98e-05,
+      "loss": 0.2396,
+      "step": 625
+    },
+    {
+      "epoch": 0.010016,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.97983870967742e-05,
+      "loss": 0.246,
+      "step": 626
+    },
+    {
+      "epoch": 0.010032,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.97967741935484e-05,
+      "loss": 0.2192,
+      "step": 627
+    },
+    {
+      "epoch": 0.010048,
+      "grad_norm": 1.0,
+      "learning_rate": 9.979516129032259e-05,
+      "loss": 0.1944,
+      "step": 628
+    },
+    {
+      "epoch": 0.010064,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.979354838709678e-05,
+      "loss": 0.239,
+      "step": 629
+    },
+    {
+      "epoch": 0.01008,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.979193548387097e-05,
+      "loss": 0.2011,
+      "step": 630
+    },
+    {
+      "epoch": 0.010096,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.979032258064517e-05,
+      "loss": 0.2139,
+      "step": 631
+    },
+    {
+      "epoch": 0.010112,
+      "grad_norm": 1.421875,
+      "learning_rate": 9.978870967741936e-05,
+      "loss": 0.2209,
+      "step": 632
+    },
+    {
+      "epoch": 0.010128,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.978709677419356e-05,
+      "loss": 0.1927,
+      "step": 633
+    },
+    {
+      "epoch": 0.010144,
+      "grad_norm": 1.25,
+      "learning_rate": 9.978548387096774e-05,
+      "loss": 0.2266,
+      "step": 634
+    },
+    {
+      "epoch": 0.01016,
+      "grad_norm": 1.75,
+      "learning_rate": 9.978387096774194e-05,
+      "loss": 0.2366,
+      "step": 635
+    },
+    {
+      "epoch": 0.010176,
+      "grad_norm": 1.96875,
+      "learning_rate": 9.978225806451613e-05,
+      "loss": 0.2254,
+      "step": 636
+    },
+    {
+      "epoch": 0.010192,
+      "grad_norm": 1.3984375,
+      "learning_rate": 9.978064516129033e-05,
+      "loss": 0.181,
+      "step": 637
+    },
+    {
+      "epoch": 0.010208,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.977903225806451e-05,
+      "loss": 0.2215,
+      "step": 638
+    },
+    {
+      "epoch": 0.010224,
+      "grad_norm": 1.4765625,
+      "learning_rate": 9.977741935483871e-05,
+      "loss": 0.2034,
+      "step": 639
+    },
+    {
+      "epoch": 0.01024,
+      "grad_norm": 1.8515625,
+      "learning_rate": 9.977580645161291e-05,
+      "loss": 0.218,
+      "step": 640
+    },
+    {
+      "epoch": 0.010256,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.977419354838711e-05,
+      "loss": 0.2423,
+      "step": 641
+    },
+    {
+      "epoch": 0.010272,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.97725806451613e-05,
+      "loss": 0.2492,
+      "step": 642
+    },
+    {
+      "epoch": 0.010288,
+      "grad_norm": 1.9765625,
+      "learning_rate": 9.977096774193548e-05,
+      "loss": 0.2564,
+      "step": 643
+    },
+    {
+      "epoch": 0.010304,
+      "grad_norm": 1.7578125,
+      "learning_rate": 9.976935483870968e-05,
+      "loss": 0.2402,
+      "step": 644
+    },
+    {
+      "epoch": 0.01032,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.976774193548387e-05,
+      "loss": 0.2073,
+      "step": 645
+    },
+    {
+      "epoch": 0.010336,
+      "grad_norm": 1.9609375,
+      "learning_rate": 9.976612903225807e-05,
+      "loss": 0.2732,
+      "step": 646
+    },
+    {
+      "epoch": 0.010352,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.976451612903226e-05,
+      "loss": 0.2381,
+      "step": 647
+    },
+    {
+      "epoch": 0.010368,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.976290322580646e-05,
+      "loss": 0.2117,
+      "step": 648
+    },
+    {
+      "epoch": 0.010384,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.976129032258064e-05,
+      "loss": 0.1856,
+      "step": 649
+    },
+    {
+      "epoch": 0.0104,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.975967741935484e-05,
+      "loss": 0.1958,
+      "step": 650
+    },
+    {
+      "epoch": 0.010416,
+      "grad_norm": 1.5390625,
+      "learning_rate": 9.975806451612904e-05,
+      "loss": 0.2336,
+      "step": 651
+    },
+    {
+      "epoch": 0.010432,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.975645161290324e-05,
+      "loss": 0.1968,
+      "step": 652
+    },
+    {
+      "epoch": 0.010448,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.975483870967743e-05,
+      "loss": 0.2473,
+      "step": 653
+    },
+    {
+      "epoch": 0.010464,
+      "grad_norm": 1.296875,
+      "learning_rate": 9.975322580645163e-05,
+      "loss": 0.2321,
+      "step": 654
+    },
+    {
+      "epoch": 0.01048,
+      "grad_norm": 1.796875,
+      "learning_rate": 9.975161290322581e-05,
+      "loss": 0.2386,
+      "step": 655
+    },
+    {
+      "epoch": 0.010496,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.975000000000001e-05,
+      "loss": 0.2141,
+      "step": 656
+    },
+    {
+      "epoch": 0.010512,
+      "grad_norm": 2.0,
+      "learning_rate": 9.97483870967742e-05,
+      "loss": 0.1694,
+      "step": 657
+    },
+    {
+      "epoch": 0.010528,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.974677419354838e-05,
+      "loss": 0.2006,
+      "step": 658
+    },
+    {
+      "epoch": 0.010544,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.974516129032258e-05,
+      "loss": 0.163,
+      "step": 659
+    },
+    {
+      "epoch": 0.01056,
+      "grad_norm": 1.75,
+      "learning_rate": 9.974354838709677e-05,
+      "loss": 0.2819,
+      "step": 660
+    },
+    {
+      "epoch": 0.010576,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.974193548387097e-05,
+      "loss": 0.204,
+      "step": 661
+    },
+    {
+      "epoch": 0.010592,
+      "grad_norm": 1.8515625,
+      "learning_rate": 9.974032258064516e-05,
+      "loss": 0.2165,
+      "step": 662
+    },
+    {
+      "epoch": 0.010608,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.973870967741936e-05,
+      "loss": 0.2271,
+      "step": 663
+    },
+    {
+      "epoch": 0.010624,
+      "grad_norm": 1.78125,
+      "learning_rate": 9.973709677419356e-05,
+      "loss": 0.2063,
+      "step": 664
+    },
+    {
+      "epoch": 0.01064,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.973548387096775e-05,
+      "loss": 0.2049,
+      "step": 665
+    },
+    {
+      "epoch": 0.010656,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.973387096774194e-05,
+      "loss": 0.2072,
+      "step": 666
+    },
+    {
+      "epoch": 0.010672,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.973225806451614e-05,
+      "loss": 0.2217,
+      "step": 667
+    },
+    {
+      "epoch": 0.010688,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.973064516129033e-05,
+      "loss": 0.2347,
+      "step": 668
+    },
+    {
+      "epoch": 0.010704,
+      "grad_norm": 1.9296875,
+      "learning_rate": 9.972903225806453e-05,
+      "loss": 0.2551,
+      "step": 669
+    },
+    {
+      "epoch": 0.01072,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.972741935483871e-05,
+      "loss": 0.197,
+      "step": 670
+    },
+    {
+      "epoch": 0.010736,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.972580645161291e-05,
+      "loss": 0.2067,
+      "step": 671
+    },
+    {
+      "epoch": 0.010752,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.97241935483871e-05,
+      "loss": 0.1859,
+      "step": 672
+    },
+    {
+      "epoch": 0.010768,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.972258064516128e-05,
+      "loss": 0.2165,
+      "step": 673
+    },
+    {
+      "epoch": 0.010784,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.972096774193548e-05,
+      "loss": 0.2142,
+      "step": 674
+    },
+    {
+      "epoch": 0.0108,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.971935483870968e-05,
+      "loss": 0.2233,
+      "step": 675
+    },
+    {
+      "epoch": 0.010816,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.971774193548388e-05,
+      "loss": 0.1952,
+      "step": 676
+    },
+    {
+      "epoch": 0.010832,
+      "grad_norm": 1.7890625,
+      "learning_rate": 9.971612903225807e-05,
+      "loss": 0.3098,
+      "step": 677
+    },
+    {
+      "epoch": 0.010848,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.971451612903227e-05,
+      "loss": 0.2258,
+      "step": 678
+    },
+    {
+      "epoch": 0.010864,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.971290322580645e-05,
+      "loss": 0.2495,
+      "step": 679
+    },
+    {
+      "epoch": 0.01088,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.971129032258065e-05,
+      "loss": 0.2017,
+      "step": 680
+    },
+    {
+      "epoch": 0.010896,
+      "grad_norm": 1.328125,
+      "learning_rate": 9.970967741935484e-05,
+      "loss": 0.2053,
+      "step": 681
+    },
+    {
+      "epoch": 0.010912,
+      "grad_norm": 1.328125,
+      "learning_rate": 9.970806451612904e-05,
+      "loss": 0.1789,
+      "step": 682
+    },
+    {
+      "epoch": 0.010928,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.970645161290323e-05,
+      "loss": 0.2078,
+      "step": 683
+    },
+    {
+      "epoch": 0.010944,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.970483870967743e-05,
+      "loss": 0.2125,
+      "step": 684
+    },
+    {
+      "epoch": 0.01096,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.970322580645161e-05,
+      "loss": 0.2103,
+      "step": 685
+    },
+    {
+      "epoch": 0.010976,
+      "grad_norm": 1.3359375,
+      "learning_rate": 9.970161290322581e-05,
+      "loss": 0.244,
+      "step": 686
+    },
+    {
+      "epoch": 0.010992,
+      "grad_norm": 1.4296875,
+      "learning_rate": 9.970000000000001e-05,
+      "loss": 0.225,
+      "step": 687
+    },
+    {
+      "epoch": 0.011008,
+      "grad_norm": 1.3515625,
+      "learning_rate": 9.96983870967742e-05,
+      "loss": 0.2329,
+      "step": 688
+    },
+    {
+      "epoch": 0.011024,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.96967741935484e-05,
+      "loss": 0.2012,
+      "step": 689
+    },
+    {
+      "epoch": 0.01104,
+      "grad_norm": 1.390625,
+      "learning_rate": 9.969516129032258e-05,
+      "loss": 0.2037,
+      "step": 690
+    },
+    {
+      "epoch": 0.011056,
+      "grad_norm": 1.8125,
+      "learning_rate": 9.969354838709678e-05,
+      "loss": 0.2204,
+      "step": 691
+    },
+    {
+      "epoch": 0.011072,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.969193548387097e-05,
+      "loss": 0.2104,
+      "step": 692
+    },
+    {
+      "epoch": 0.011088,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.969032258064517e-05,
+      "loss": 0.1956,
+      "step": 693
+    },
+    {
+      "epoch": 0.011104,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.968870967741935e-05,
+      "loss": 0.2028,
+      "step": 694
+    },
+    {
+      "epoch": 0.01112,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.968709677419355e-05,
+      "loss": 0.2225,
+      "step": 695
+    },
+    {
+      "epoch": 0.011136,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.968548387096774e-05,
+      "loss": 0.2074,
+      "step": 696
+    },
+    {
+      "epoch": 0.011152,
+      "grad_norm": 1.7734375,
+      "learning_rate": 9.968387096774194e-05,
+      "loss": 0.2424,
+      "step": 697
+    },
+    {
+      "epoch": 0.011168,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.968225806451613e-05,
+      "loss": 0.2151,
+      "step": 698
+    },
+    {
+      "epoch": 0.011184,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.968064516129033e-05,
+      "loss": 0.2348,
+      "step": 699
+    },
+    {
+      "epoch": 0.0112,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.967903225806452e-05,
+      "loss": 0.2842,
+      "step": 700
+    },
+    {
+      "epoch": 0.011216,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.967741935483872e-05,
+      "loss": 0.2215,
+      "step": 701
+    },
+    {
+      "epoch": 0.011232,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.967580645161291e-05,
+      "loss": 0.2159,
+      "step": 702
+    },
+    {
+      "epoch": 0.011248,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.967419354838711e-05,
+      "loss": 0.2333,
+      "step": 703
+    },
+    {
+      "epoch": 0.011264,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.96725806451613e-05,
+      "loss": 0.2268,
+      "step": 704
+    },
+    {
+      "epoch": 0.01128,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.967096774193548e-05,
+      "loss": 0.2241,
+      "step": 705
+    },
+    {
+      "epoch": 0.011296,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.966935483870968e-05,
+      "loss": 0.2038,
+      "step": 706
+    },
+    {
+      "epoch": 0.011312,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.966774193548387e-05,
+      "loss": 0.2241,
+      "step": 707
+    },
+    {
+      "epoch": 0.011328,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.966612903225807e-05,
+      "loss": 0.197,
+      "step": 708
+    },
+    {
+      "epoch": 0.011344,
+      "grad_norm": 1.5078125,
+      "learning_rate": 9.966451612903225e-05,
+      "loss": 0.2804,
+      "step": 709
+    },
+    {
+      "epoch": 0.01136,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.966290322580645e-05,
+      "loss": 0.1791,
+      "step": 710
+    },
+    {
+      "epoch": 0.011376,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.966129032258065e-05,
+      "loss": 0.1964,
+      "step": 711
+    },
+    {
+      "epoch": 0.011392,
+      "grad_norm": 1.8203125,
+      "learning_rate": 9.965967741935485e-05,
+      "loss": 0.1916,
+      "step": 712
+    },
+    {
+      "epoch": 0.011408,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.965806451612904e-05,
+      "loss": 0.1687,
+      "step": 713
+    },
+    {
+      "epoch": 0.011424,
+      "grad_norm": 1.3515625,
+      "learning_rate": 9.965645161290324e-05,
+      "loss": 0.1872,
+      "step": 714
+    },
+    {
+      "epoch": 0.01144,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.965483870967742e-05,
+      "loss": 0.152,
+      "step": 715
+    },
+    {
+      "epoch": 0.011456,
+      "grad_norm": 1.34375,
+      "learning_rate": 9.965322580645162e-05,
+      "loss": 0.2456,
+      "step": 716
+    },
+    {
+      "epoch": 0.011472,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.965161290322581e-05,
+      "loss": 0.2287,
+      "step": 717
+    },
+    {
+      "epoch": 0.011488,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.965000000000001e-05,
+      "loss": 0.2455,
+      "step": 718
+    },
+    {
+      "epoch": 0.011504,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.96483870967742e-05,
+      "loss": 0.1931,
+      "step": 719
+    },
+    {
+      "epoch": 0.01152,
+      "grad_norm": 1.46875,
+      "learning_rate": 9.964677419354838e-05,
+      "loss": 0.2275,
+      "step": 720
+    },
+    {
+      "epoch": 0.011536,
+      "grad_norm": 1.3984375,
+      "learning_rate": 9.964516129032258e-05,
+      "loss": 0.2642,
+      "step": 721
+    },
+    {
+      "epoch": 0.011552,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.964354838709678e-05,
+      "loss": 0.2104,
+      "step": 722
+    },
+    {
+      "epoch": 0.011568,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.964193548387097e-05,
+      "loss": 0.2519,
+      "step": 723
+    },
+    {
+      "epoch": 0.011584,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.964032258064517e-05,
+      "loss": 0.2009,
+      "step": 724
+    },
+    {
+      "epoch": 0.0116,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.963870967741937e-05,
+      "loss": 0.2427,
+      "step": 725
+    },
+    {
+      "epoch": 0.011616,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.963709677419355e-05,
+      "loss": 0.2292,
+      "step": 726
+    },
+    {
+      "epoch": 0.011632,
+      "grad_norm": 1.46875,
+      "learning_rate": 9.963548387096775e-05,
+      "loss": 0.2484,
+      "step": 727
+    },
+    {
+      "epoch": 0.011648,
+      "grad_norm": 1.40625,
+      "learning_rate": 9.963387096774194e-05,
+      "loss": 0.217,
+      "step": 728
+    },
+    {
+      "epoch": 0.011664,
+      "grad_norm": 1.7578125,
+      "learning_rate": 9.963225806451614e-05,
+      "loss": 0.2343,
+      "step": 729
+    },
+    {
+      "epoch": 0.01168,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.963064516129032e-05,
+      "loss": 0.2305,
+      "step": 730
+    },
+    {
+      "epoch": 0.011696,
+      "grad_norm": 1.3828125,
+      "learning_rate": 9.962903225806452e-05,
+      "loss": 0.2008,
+      "step": 731
+    },
+    {
+      "epoch": 0.011712,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.962741935483871e-05,
+      "loss": 0.237,
+      "step": 732
+    },
+    {
+      "epoch": 0.011728,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.962580645161291e-05,
+      "loss": 0.1682,
+      "step": 733
+    },
+    {
+      "epoch": 0.011744,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.96241935483871e-05,
+      "loss": 0.2004,
+      "step": 734
+    },
+    {
+      "epoch": 0.01176,
+      "grad_norm": 1.296875,
+      "learning_rate": 9.96225806451613e-05,
+      "loss": 0.2307,
+      "step": 735
+    },
+    {
+      "epoch": 0.011776,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.96209677419355e-05,
+      "loss": 0.2032,
+      "step": 736
+    },
+    {
+      "epoch": 0.011792,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.961935483870968e-05,
+      "loss": 0.2416,
+      "step": 737
+    },
+    {
+      "epoch": 0.011808,
+      "grad_norm": 1.8359375,
+      "learning_rate": 9.961774193548388e-05,
+      "loss": 0.1944,
+      "step": 738
+    },
+    {
+      "epoch": 0.011824,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.961612903225807e-05,
+      "loss": 0.1941,
+      "step": 739
+    },
+    {
+      "epoch": 0.01184,
+      "grad_norm": 1.5078125,
+      "learning_rate": 9.961451612903227e-05,
+      "loss": 0.2097,
+      "step": 740
+    },
+    {
+      "epoch": 0.011856,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.961290322580645e-05,
+      "loss": 0.2272,
+      "step": 741
+    },
+    {
+      "epoch": 0.011872,
+      "grad_norm": 1.3515625,
+      "learning_rate": 9.961129032258065e-05,
+      "loss": 0.1824,
+      "step": 742
+    },
+    {
+      "epoch": 0.011888,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.960967741935484e-05,
+      "loss": 0.2028,
+      "step": 743
+    },
+    {
+      "epoch": 0.011904,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.960806451612904e-05,
+      "loss": 0.1955,
+      "step": 744
+    },
+    {
+      "epoch": 0.01192,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.960645161290322e-05,
+      "loss": 0.2003,
+      "step": 745
+    },
+    {
+      "epoch": 0.011936,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.960483870967742e-05,
+      "loss": 0.2222,
+      "step": 746
+    },
+    {
+      "epoch": 0.011952,
+      "grad_norm": 5.375,
+      "learning_rate": 9.960322580645162e-05,
+      "loss": 0.2473,
+      "step": 747
+    },
+    {
+      "epoch": 0.011968,
+      "grad_norm": 1.296875,
+      "learning_rate": 9.960161290322582e-05,
+      "loss": 0.1969,
+      "step": 748
+    },
+    {
+      "epoch": 0.011984,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.960000000000001e-05,
+      "loss": 0.2114,
+      "step": 749
+    },
+    {
+      "epoch": 0.012,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.95983870967742e-05,
+      "loss": 0.2058,
+      "step": 750
+    },
+    {
+      "epoch": 0.012016,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.95967741935484e-05,
+      "loss": 0.2034,
+      "step": 751
+    },
+    {
+      "epoch": 0.012032,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.959516129032258e-05,
+      "loss": 0.1769,
+      "step": 752
+    },
+    {
+      "epoch": 0.012048,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.959354838709678e-05,
+      "loss": 0.2398,
+      "step": 753
+    },
+    {
+      "epoch": 0.012064,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.959193548387097e-05,
+      "loss": 0.2288,
+      "step": 754
+    },
+    {
+      "epoch": 0.01208,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.959032258064517e-05,
+      "loss": 0.1635,
+      "step": 755
+    },
+    {
+      "epoch": 0.012096,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.958870967741935e-05,
+      "loss": 0.1848,
+      "step": 756
+    },
+    {
+      "epoch": 0.012112,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.958709677419355e-05,
+      "loss": 0.1702,
+      "step": 757
+    },
+    {
+      "epoch": 0.012128,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.958548387096774e-05,
+      "loss": 0.2262,
+      "step": 758
+    },
+    {
+      "epoch": 0.012144,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.958387096774194e-05,
+      "loss": 0.1678,
+      "step": 759
+    },
+    {
+      "epoch": 0.01216,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.958225806451614e-05,
+      "loss": 0.1731,
+      "step": 760
+    },
+    {
+      "epoch": 0.012176,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.958064516129034e-05,
+      "loss": 0.2346,
+      "step": 761
+    },
+    {
+      "epoch": 0.012192,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.957903225806452e-05,
+      "loss": 0.2022,
+      "step": 762
+    },
+    {
+      "epoch": 0.012208,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.957741935483872e-05,
+      "loss": 0.2034,
+      "step": 763
+    },
+    {
+      "epoch": 0.012224,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.957580645161291e-05,
+      "loss": 0.1924,
+      "step": 764
+    },
+    {
+      "epoch": 0.01224,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.957419354838711e-05,
+      "loss": 0.1805,
+      "step": 765
+    },
+    {
+      "epoch": 0.012256,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.95725806451613e-05,
+      "loss": 0.2055,
+      "step": 766
+    },
+    {
+      "epoch": 0.012272,
+      "grad_norm": 1.3359375,
+      "learning_rate": 9.957096774193548e-05,
+      "loss": 0.2685,
+      "step": 767
+    },
+    {
+      "epoch": 0.012288,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.956935483870968e-05,
+      "loss": 0.2385,
+      "step": 768
+    },
+    {
+      "epoch": 0.012304,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.956774193548387e-05,
+      "loss": 0.1792,
+      "step": 769
+    },
+    {
+      "epoch": 0.01232,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.956612903225807e-05,
+      "loss": 0.2171,
+      "step": 770
+    },
+    {
+      "epoch": 0.012336,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.956451612903226e-05,
+      "loss": 0.2407,
+      "step": 771
+    },
+    {
+      "epoch": 0.012352,
+      "grad_norm": 1.9375,
+      "learning_rate": 9.956290322580646e-05,
+      "loss": 0.19,
+      "step": 772
+    },
+    {
+      "epoch": 0.012368,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.956129032258065e-05,
+      "loss": 0.2377,
+      "step": 773
+    },
+    {
+      "epoch": 0.012384,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.955967741935485e-05,
+      "loss": 0.1635,
+      "step": 774
+    },
+    {
+      "epoch": 0.0124,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.955806451612904e-05,
+      "loss": 0.2026,
+      "step": 775
+    },
+    {
+      "epoch": 0.012416,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.955645161290324e-05,
+      "loss": 0.2005,
+      "step": 776
+    },
+    {
+      "epoch": 0.012432,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.955483870967742e-05,
+      "loss": 0.1962,
+      "step": 777
+    },
+    {
+      "epoch": 0.012448,
+      "grad_norm": 1.125,
+      "learning_rate": 9.955322580645162e-05,
+      "loss": 0.2015,
+      "step": 778
+    },
+    {
+      "epoch": 0.012464,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.955161290322581e-05,
+      "loss": 0.1914,
+      "step": 779
+    },
+    {
+      "epoch": 0.01248,
+      "grad_norm": 1.4296875,
+      "learning_rate": 9.955000000000001e-05,
+      "loss": 0.2031,
+      "step": 780
+    },
+    {
+      "epoch": 0.012496,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.95483870967742e-05,
+      "loss": 0.2236,
+      "step": 781
+    },
+    {
+      "epoch": 0.012512,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.954677419354839e-05,
+      "loss": 0.1848,
+      "step": 782
+    },
+    {
+      "epoch": 0.012528,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.954516129032259e-05,
+      "loss": 0.1812,
+      "step": 783
+    },
+    {
+      "epoch": 0.012544,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.954354838709678e-05,
+      "loss": 0.2319,
+      "step": 784
+    },
+    {
+      "epoch": 0.01256,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.954193548387098e-05,
+      "loss": 0.2012,
+      "step": 785
+    },
+    {
+      "epoch": 0.012576,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.954032258064516e-05,
+      "loss": 0.192,
+      "step": 786
+    },
+    {
+      "epoch": 0.012592,
+      "grad_norm": 1.34375,
+      "learning_rate": 9.953870967741936e-05,
+      "loss": 0.2039,
+      "step": 787
+    },
+    {
+      "epoch": 0.012608,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.953709677419355e-05,
+      "loss": 0.1609,
+      "step": 788
+    },
+    {
+      "epoch": 0.012624,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.953548387096775e-05,
+      "loss": 0.1954,
+      "step": 789
+    },
+    {
+      "epoch": 0.01264,
+      "grad_norm": 1.484375,
+      "learning_rate": 9.953387096774194e-05,
+      "loss": 0.223,
+      "step": 790
+    },
+    {
+      "epoch": 0.012656,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.953225806451614e-05,
+      "loss": 0.2148,
+      "step": 791
+    },
+    {
+      "epoch": 0.012672,
+      "grad_norm": 1.75,
+      "learning_rate": 9.953064516129032e-05,
+      "loss": 0.2116,
+      "step": 792
+    },
+    {
+      "epoch": 0.012688,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.952903225806452e-05,
+      "loss": 0.2004,
+      "step": 793
+    },
+    {
+      "epoch": 0.012704,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.952741935483871e-05,
+      "loss": 0.2442,
+      "step": 794
+    },
+    {
+      "epoch": 0.01272,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.952580645161291e-05,
+      "loss": 0.2269,
+      "step": 795
+    },
+    {
+      "epoch": 0.012736,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.95241935483871e-05,
+      "loss": 0.1999,
+      "step": 796
+    },
+    {
+      "epoch": 0.012752,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.952258064516129e-05,
+      "loss": 0.1885,
+      "step": 797
+    },
+    {
+      "epoch": 0.012768,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.952096774193549e-05,
+      "loss": 0.2183,
+      "step": 798
+    },
+    {
+      "epoch": 0.012784,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.951935483870968e-05,
+      "loss": 0.2101,
+      "step": 799
+    },
+    {
+      "epoch": 0.0128,
+      "grad_norm": 1.296875,
+      "learning_rate": 9.951774193548388e-05,
+      "loss": 0.2546,
+      "step": 800
+    },
+    {
+      "epoch": 0.012816,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.951612903225806e-05,
+      "loss": 0.2341,
+      "step": 801
+    },
+    {
+      "epoch": 0.012832,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.951451612903226e-05,
+      "loss": 0.2112,
+      "step": 802
+    },
+    {
+      "epoch": 0.012848,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.951290322580645e-05,
+      "loss": 0.2143,
+      "step": 803
+    },
+    {
+      "epoch": 0.012864,
+      "grad_norm": 1.4296875,
+      "learning_rate": 9.951129032258065e-05,
+      "loss": 0.2114,
+      "step": 804
+    },
+    {
+      "epoch": 0.01288,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.950967741935484e-05,
+      "loss": 0.1925,
+      "step": 805
+    },
+    {
+      "epoch": 0.012896,
+      "grad_norm": 1.375,
+      "learning_rate": 9.950806451612904e-05,
+      "loss": 0.2346,
+      "step": 806
+    },
+    {
+      "epoch": 0.012912,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.950645161290323e-05,
+      "loss": 0.2254,
+      "step": 807
+    },
+    {
+      "epoch": 0.012928,
+      "grad_norm": 1.3359375,
+      "learning_rate": 9.950483870967743e-05,
+      "loss": 0.2141,
+      "step": 808
+    },
+    {
+      "epoch": 0.012944,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.950322580645162e-05,
+      "loss": 0.2551,
+      "step": 809
+    },
+    {
+      "epoch": 0.01296,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.950161290322582e-05,
+      "loss": 0.1671,
+      "step": 810
+    },
+    {
+      "epoch": 0.012976,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.95e-05,
+      "loss": 0.2356,
+      "step": 811
+    },
+    {
+      "epoch": 0.012992,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.94983870967742e-05,
+      "loss": 0.21,
+      "step": 812
+    },
+    {
+      "epoch": 0.013008,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.949677419354839e-05,
+      "loss": 0.1723,
+      "step": 813
+    },
+    {
+      "epoch": 0.013024,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.949516129032258e-05,
+      "loss": 0.1829,
+      "step": 814
+    },
+    {
+      "epoch": 0.01304,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.949354838709678e-05,
+      "loss": 0.2432,
+      "step": 815
+    },
+    {
+      "epoch": 0.013056,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.949193548387096e-05,
+      "loss": 0.2011,
+      "step": 816
+    },
+    {
+      "epoch": 0.013072,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.949032258064516e-05,
+      "loss": 0.173,
+      "step": 817
+    },
+    {
+      "epoch": 0.013088,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.948870967741935e-05,
+      "loss": 0.1763,
+      "step": 818
+    },
+    {
+      "epoch": 0.013104,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.948709677419355e-05,
+      "loss": 0.2085,
+      "step": 819
+    },
+    {
+      "epoch": 0.01312,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.948548387096775e-05,
+      "loss": 0.1955,
+      "step": 820
+    },
+    {
+      "epoch": 0.013136,
+      "grad_norm": 1.3359375,
+      "learning_rate": 9.948387096774195e-05,
+      "loss": 0.2483,
+      "step": 821
+    },
+    {
+      "epoch": 0.013152,
+      "grad_norm": 1.3359375,
+      "learning_rate": 9.948225806451613e-05,
+      "loss": 0.2284,
+      "step": 822
+    },
+    {
+      "epoch": 0.013168,
+      "grad_norm": 1.4765625,
+      "learning_rate": 9.948064516129033e-05,
+      "loss": 0.2112,
+      "step": 823
+    },
+    {
+      "epoch": 0.013184,
+      "grad_norm": 1.4296875,
+      "learning_rate": 9.947903225806452e-05,
+      "loss": 0.2141,
+      "step": 824
+    },
+    {
+      "epoch": 0.0132,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.947741935483872e-05,
+      "loss": 0.2904,
+      "step": 825
+    },
+    {
+      "epoch": 0.013216,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.94758064516129e-05,
+      "loss": 0.2136,
+      "step": 826
+    },
+    {
+      "epoch": 0.013232,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.94741935483871e-05,
+      "loss": 0.2146,
+      "step": 827
+    },
+    {
+      "epoch": 0.013248,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.947258064516129e-05,
+      "loss": 0.2263,
+      "step": 828
+    },
+    {
+      "epoch": 0.013264,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.947096774193548e-05,
+      "loss": 0.2406,
+      "step": 829
+    },
+    {
+      "epoch": 0.01328,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.946935483870968e-05,
+      "loss": 0.2023,
+      "step": 830
+    },
+    {
+      "epoch": 0.013296,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.946774193548388e-05,
+      "loss": 0.2262,
+      "step": 831
+    },
+    {
+      "epoch": 0.013312,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.946612903225808e-05,
+      "loss": 0.1889,
+      "step": 832
+    },
+    {
+      "epoch": 0.013328,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.946451612903226e-05,
+      "loss": 0.2241,
+      "step": 833
+    },
+    {
+      "epoch": 0.013344,
+      "grad_norm": 1.40625,
+      "learning_rate": 9.946290322580646e-05,
+      "loss": 0.2066,
+      "step": 834
+    },
+    {
+      "epoch": 0.01336,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.946129032258065e-05,
+      "loss": 0.2462,
+      "step": 835
+    },
+    {
+      "epoch": 0.013376,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.945967741935485e-05,
+      "loss": 0.1734,
+      "step": 836
+    },
+    {
+      "epoch": 0.013392,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.945806451612903e-05,
+      "loss": 0.2366,
+      "step": 837
+    },
+    {
+      "epoch": 0.013408,
+      "grad_norm": 1.40625,
+      "learning_rate": 9.945645161290323e-05,
+      "loss": 0.2175,
+      "step": 838
+    },
+    {
+      "epoch": 0.013424,
+      "grad_norm": 1.4921875,
+      "learning_rate": 9.945483870967742e-05,
+      "loss": 0.217,
+      "step": 839
+    },
+    {
+      "epoch": 0.01344,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.945322580645162e-05,
+      "loss": 0.2056,
+      "step": 840
+    },
+    {
+      "epoch": 0.013456,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.94516129032258e-05,
+      "loss": 0.1785,
+      "step": 841
+    },
+    {
+      "epoch": 0.013472,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.945e-05,
+      "loss": 0.2163,
+      "step": 842
+    },
+    {
+      "epoch": 0.013488,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.94483870967742e-05,
+      "loss": 0.2325,
+      "step": 843
+    },
+    {
+      "epoch": 0.013504,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.944677419354839e-05,
+      "loss": 0.2077,
+      "step": 844
+    },
+    {
+      "epoch": 0.01352,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.944516129032259e-05,
+      "loss": 0.2435,
+      "step": 845
+    },
+    {
+      "epoch": 0.013536,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.944354838709678e-05,
+      "loss": 0.1962,
+      "step": 846
+    },
+    {
+      "epoch": 0.013552,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.944193548387098e-05,
+      "loss": 0.2145,
+      "step": 847
+    },
+    {
+      "epoch": 0.013568,
+      "grad_norm": 1.46875,
+      "learning_rate": 9.944032258064516e-05,
+      "loss": 0.201,
+      "step": 848
+    },
+    {
+      "epoch": 0.013584,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.943870967741936e-05,
+      "loss": 0.19,
+      "step": 849
+    },
+    {
+      "epoch": 0.0136,
+      "grad_norm": 1.3828125,
+      "learning_rate": 9.943709677419355e-05,
+      "loss": 0.2108,
+      "step": 850
+    },
+    {
+      "epoch": 0.013616,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.943548387096775e-05,
+      "loss": 0.1802,
+      "step": 851
+    },
+    {
+      "epoch": 0.013632,
+      "grad_norm": 1.484375,
+      "learning_rate": 9.943387096774193e-05,
+      "loss": 0.2137,
+      "step": 852
+    },
+    {
+      "epoch": 0.013648,
+      "grad_norm": 1.3125,
+      "learning_rate": 9.943225806451613e-05,
+      "loss": 0.203,
+      "step": 853
+    },
+    {
+      "epoch": 0.013664,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.943064516129032e-05,
+      "loss": 0.1695,
+      "step": 854
+    },
+    {
+      "epoch": 0.01368,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.942903225806452e-05,
+      "loss": 0.2387,
+      "step": 855
+    },
+    {
+      "epoch": 0.013696,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.942741935483872e-05,
+      "loss": 0.1774,
+      "step": 856
+    },
+    {
+      "epoch": 0.013712,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.942580645161292e-05,
+      "loss": 0.2013,
+      "step": 857
+    },
+    {
+      "epoch": 0.013728,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.94241935483871e-05,
+      "loss": 0.2341,
+      "step": 858
+    },
+    {
+      "epoch": 0.013744,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.942258064516129e-05,
+      "loss": 0.1948,
+      "step": 859
+    },
+    {
+      "epoch": 0.01376,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.942096774193549e-05,
+      "loss": 0.1863,
+      "step": 860
+    },
+    {
+      "epoch": 0.013776,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.941935483870968e-05,
+      "loss": 0.2684,
+      "step": 861
+    },
+    {
+      "epoch": 0.013792,
+      "grad_norm": 1.4453125,
+      "learning_rate": 9.941774193548388e-05,
+      "loss": 0.2227,
+      "step": 862
+    },
+    {
+      "epoch": 0.013808,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.941612903225806e-05,
+      "loss": 0.2022,
+      "step": 863
+    },
+    {
+      "epoch": 0.013824,
+      "grad_norm": 1.515625,
+      "learning_rate": 9.941451612903226e-05,
+      "loss": 0.2286,
+      "step": 864
+    },
+    {
+      "epoch": 0.01384,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.941290322580645e-05,
+      "loss": 0.2393,
+      "step": 865
+    },
+    {
+      "epoch": 0.013856,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.941129032258065e-05,
+      "loss": 0.2288,
+      "step": 866
+    },
+    {
+      "epoch": 0.013872,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.940967741935485e-05,
+      "loss": 0.1901,
+      "step": 867
+    },
+    {
+      "epoch": 0.013888,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.940806451612905e-05,
+      "loss": 0.1744,
+      "step": 868
+    },
+    {
+      "epoch": 0.013904,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.940645161290323e-05,
+      "loss": 0.2246,
+      "step": 869
+    },
+    {
+      "epoch": 0.01392,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.940483870967743e-05,
+      "loss": 0.2137,
+      "step": 870
+    },
+    {
+      "epoch": 0.013936,
+      "grad_norm": 1.359375,
+      "learning_rate": 9.940322580645162e-05,
+      "loss": 0.27,
+      "step": 871
+    },
+    {
+      "epoch": 0.013952,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.940161290322582e-05,
+      "loss": 0.1714,
+      "step": 872
+    },
+    {
+      "epoch": 0.013968,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.94e-05,
+      "loss": 0.183,
+      "step": 873
+    },
+    {
+      "epoch": 0.013984,
+      "grad_norm": 1.3515625,
+      "learning_rate": 9.93983870967742e-05,
+      "loss": 0.2285,
+      "step": 874
+    },
+    {
+      "epoch": 0.014,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.939677419354839e-05,
+      "loss": 0.2273,
+      "step": 875
+    },
+    {
+      "epoch": 0.014016,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.939516129032258e-05,
+      "loss": 0.2104,
+      "step": 876
+    },
+    {
+      "epoch": 0.014032,
+      "grad_norm": 1.421875,
+      "learning_rate": 9.939354838709678e-05,
+      "loss": 0.2322,
+      "step": 877
+    },
+    {
+      "epoch": 0.014048,
+      "grad_norm": 1.40625,
+      "learning_rate": 9.939193548387097e-05,
+      "loss": 0.2317,
+      "step": 878
+    },
+    {
+      "epoch": 0.014064,
+      "grad_norm": 1.46875,
+      "learning_rate": 9.939032258064517e-05,
+      "loss": 0.2283,
+      "step": 879
+    },
+    {
+      "epoch": 0.01408,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.938870967741936e-05,
+      "loss": 0.2393,
+      "step": 880
+    },
+    {
+      "epoch": 0.014096,
+      "grad_norm": 2.125,
+      "learning_rate": 9.938709677419356e-05,
+      "loss": 0.22,
+      "step": 881
+    },
+    {
+      "epoch": 0.014112,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.938548387096775e-05,
+      "loss": 0.2907,
+      "step": 882
+    },
+    {
+      "epoch": 0.014128,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.938387096774195e-05,
+      "loss": 0.1981,
+      "step": 883
+    },
+    {
+      "epoch": 0.014144,
+      "grad_norm": 1.4453125,
+      "learning_rate": 9.938225806451613e-05,
+      "loss": 0.2293,
+      "step": 884
+    },
+    {
+      "epoch": 0.01416,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.938064516129033e-05,
+      "loss": 0.2157,
+      "step": 885
+    },
+    {
+      "epoch": 0.014176,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.937903225806452e-05,
+      "loss": 0.183,
+      "step": 886
+    },
+    {
+      "epoch": 0.014192,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.937741935483872e-05,
+      "loss": 0.1932,
+      "step": 887
+    },
+    {
+      "epoch": 0.014208,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.93758064516129e-05,
+      "loss": 0.2034,
+      "step": 888
+    },
+    {
+      "epoch": 0.014224,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.93741935483871e-05,
+      "loss": 0.17,
+      "step": 889
+    },
+    {
+      "epoch": 0.01424,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.937258064516129e-05,
+      "loss": 0.1972,
+      "step": 890
+    },
+    {
+      "epoch": 0.014256,
+      "grad_norm": 1.765625,
+      "learning_rate": 9.937096774193549e-05,
+      "loss": 0.166,
+      "step": 891
+    },
+    {
+      "epoch": 0.014272,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.936935483870969e-05,
+      "loss": 0.2194,
+      "step": 892
+    },
+    {
+      "epoch": 0.014288,
+      "grad_norm": 1.453125,
+      "learning_rate": 9.936774193548387e-05,
+      "loss": 0.2623,
+      "step": 893
+    },
+    {
+      "epoch": 0.014304,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.936612903225807e-05,
+      "loss": 0.1837,
+      "step": 894
+    },
+    {
+      "epoch": 0.01432,
+      "grad_norm": 1.3828125,
+      "learning_rate": 9.936451612903226e-05,
+      "loss": 0.2214,
+      "step": 895
+    },
+    {
+      "epoch": 0.014336,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.936290322580646e-05,
+      "loss": 0.1982,
+      "step": 896
+    },
+    {
+      "epoch": 0.014352,
+      "grad_norm": 1.5,
+      "learning_rate": 9.936129032258065e-05,
+      "loss": 0.2465,
+      "step": 897
+    },
+    {
+      "epoch": 0.014368,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.935967741935485e-05,
+      "loss": 0.1675,
+      "step": 898
+    },
+    {
+      "epoch": 0.014384,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.935806451612903e-05,
+      "loss": 0.1721,
+      "step": 899
+    },
+    {
+      "epoch": 0.0144,
+      "grad_norm": 1.359375,
+      "learning_rate": 9.935645161290323e-05,
+      "loss": 0.2078,
+      "step": 900
+    },
+    {
+      "epoch": 0.014416,
+      "grad_norm": 1.125,
+      "learning_rate": 9.935483870967742e-05,
+      "loss": 0.1993,
+      "step": 901
+    },
+    {
+      "epoch": 0.014432,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.935322580645162e-05,
+      "loss": 0.2175,
+      "step": 902
+    },
+    {
+      "epoch": 0.014448,
+      "grad_norm": 1.25,
+      "learning_rate": 9.935161290322582e-05,
+      "loss": 0.1971,
+      "step": 903
+    },
+    {
+      "epoch": 0.014464,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.935000000000002e-05,
+      "loss": 0.2346,
+      "step": 904
+    },
+    {
+      "epoch": 0.01448,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.93483870967742e-05,
+      "loss": 0.17,
+      "step": 905
+    },
+    {
+      "epoch": 0.014496,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.934677419354839e-05,
+      "loss": 0.1834,
+      "step": 906
+    },
+    {
+      "epoch": 0.014512,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.934516129032259e-05,
+      "loss": 0.2397,
+      "step": 907
+    },
+    {
+      "epoch": 0.014528,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.934354838709677e-05,
+      "loss": 0.1874,
+      "step": 908
+    },
+    {
+      "epoch": 0.014544,
+      "grad_norm": 1.5,
+      "learning_rate": 9.934193548387097e-05,
+      "loss": 0.198,
+      "step": 909
+    },
+    {
+      "epoch": 0.01456,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.934032258064516e-05,
+      "loss": 0.2313,
+      "step": 910
+    },
+    {
+      "epoch": 0.014576,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.933870967741936e-05,
+      "loss": 0.2063,
+      "step": 911
+    },
+    {
+      "epoch": 0.014592,
+      "grad_norm": 1.0,
+      "learning_rate": 9.933709677419355e-05,
+      "loss": 0.2003,
+      "step": 912
+    },
+    {
+      "epoch": 0.014608,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.933548387096775e-05,
+      "loss": 0.1959,
+      "step": 913
+    },
+    {
+      "epoch": 0.014624,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.933387096774193e-05,
+      "loss": 0.2202,
+      "step": 914
+    },
+    {
+      "epoch": 0.01464,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.933225806451613e-05,
+      "loss": 0.2125,
+      "step": 915
+    },
+    {
+      "epoch": 0.014656,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.933064516129033e-05,
+      "loss": 0.2219,
+      "step": 916
+    },
+    {
+      "epoch": 0.014672,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.932903225806453e-05,
+      "loss": 0.2201,
+      "step": 917
+    },
+    {
+      "epoch": 0.014688,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.932741935483872e-05,
+      "loss": 0.195,
+      "step": 918
+    },
+    {
+      "epoch": 0.014704,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.932580645161292e-05,
+      "loss": 0.2514,
+      "step": 919
+    },
+    {
+      "epoch": 0.01472,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.93241935483871e-05,
+      "loss": 0.1727,
+      "step": 920
+    },
+    {
+      "epoch": 0.014736,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.93225806451613e-05,
+      "loss": 0.248,
+      "step": 921
+    },
+    {
+      "epoch": 0.014752,
+      "grad_norm": 1.3515625,
+      "learning_rate": 9.932096774193549e-05,
+      "loss": 0.2325,
+      "step": 922
+    },
+    {
+      "epoch": 0.014768,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.931935483870967e-05,
+      "loss": 0.2411,
+      "step": 923
+    },
+    {
+      "epoch": 0.014784,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.931774193548387e-05,
+      "loss": 0.2039,
+      "step": 924
+    },
+    {
+      "epoch": 0.0148,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.931612903225806e-05,
+      "loss": 0.1797,
+      "step": 925
+    },
+    {
+      "epoch": 0.014816,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.931451612903226e-05,
+      "loss": 0.1941,
+      "step": 926
+    },
+    {
+      "epoch": 0.014832,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.931290322580646e-05,
+      "loss": 0.1961,
+      "step": 927
+    },
+    {
+      "epoch": 0.014848,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.931129032258066e-05,
+      "loss": 0.2475,
+      "step": 928
+    },
+    {
+      "epoch": 0.014864,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.930967741935484e-05,
+      "loss": 0.2143,
+      "step": 929
+    },
+    {
+      "epoch": 0.01488,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.930806451612904e-05,
+      "loss": 0.2077,
+      "step": 930
+    },
+    {
+      "epoch": 0.014896,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.930645161290323e-05,
+      "loss": 0.2135,
+      "step": 931
+    },
+    {
+      "epoch": 0.014912,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.930483870967743e-05,
+      "loss": 0.2256,
+      "step": 932
+    },
+    {
+      "epoch": 0.014928,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.930322580645162e-05,
+      "loss": 0.2481,
+      "step": 933
+    },
+    {
+      "epoch": 0.014944,
+      "grad_norm": 1.421875,
+      "learning_rate": 9.930161290322582e-05,
+      "loss": 0.2516,
+      "step": 934
+    },
+    {
+      "epoch": 0.01496,
+      "grad_norm": 1.421875,
+      "learning_rate": 9.93e-05,
+      "loss": 0.2264,
+      "step": 935
+    },
+    {
+      "epoch": 0.014976,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.92983870967742e-05,
+      "loss": 0.2307,
+      "step": 936
+    },
+    {
+      "epoch": 0.014992,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.929677419354839e-05,
+      "loss": 0.2463,
+      "step": 937
+    },
+    {
+      "epoch": 0.015008,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.929516129032259e-05,
+      "loss": 0.2543,
+      "step": 938
+    },
+    {
+      "epoch": 0.015024,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.929354838709679e-05,
+      "loss": 0.2362,
+      "step": 939
+    },
+    {
+      "epoch": 0.01504,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.929193548387097e-05,
+      "loss": 0.1823,
+      "step": 940
+    },
+    {
+      "epoch": 0.015056,
+      "grad_norm": 1.125,
+      "learning_rate": 9.929032258064517e-05,
+      "loss": 0.2382,
+      "step": 941
+    },
+    {
+      "epoch": 0.015072,
+      "grad_norm": 1.390625,
+      "learning_rate": 9.928870967741936e-05,
+      "loss": 0.2167,
+      "step": 942
+    },
+    {
+      "epoch": 0.015088,
+      "grad_norm": 1.25,
+      "learning_rate": 9.928709677419356e-05,
+      "loss": 0.1978,
+      "step": 943
+    },
+    {
+      "epoch": 0.015104,
+      "grad_norm": 1.3984375,
+      "learning_rate": 9.928548387096774e-05,
+      "loss": 0.2223,
+      "step": 944
+    },
+    {
+      "epoch": 0.01512,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.928387096774194e-05,
+      "loss": 0.1772,
+      "step": 945
+    },
+    {
+      "epoch": 0.015136,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.928225806451613e-05,
+      "loss": 0.2323,
+      "step": 946
+    },
+    {
+      "epoch": 0.015152,
+      "grad_norm": 1.4765625,
+      "learning_rate": 9.928064516129033e-05,
+      "loss": 0.293,
+      "step": 947
+    },
+    {
+      "epoch": 0.015168,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.927903225806452e-05,
+      "loss": 0.1987,
+      "step": 948
+    },
+    {
+      "epoch": 0.015184,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.927741935483871e-05,
+      "loss": 0.204,
+      "step": 949
+    },
+    {
+      "epoch": 0.0152,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.92758064516129e-05,
+      "loss": 0.1946,
+      "step": 950
+    },
+    {
+      "epoch": 0.015216,
+      "grad_norm": 1.5,
+      "learning_rate": 9.92741935483871e-05,
+      "loss": 0.1775,
+      "step": 951
+    },
+    {
+      "epoch": 0.015232,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.92725806451613e-05,
+      "loss": 0.2289,
+      "step": 952
+    },
+    {
+      "epoch": 0.015248,
+      "grad_norm": 1.75,
+      "learning_rate": 9.927096774193549e-05,
+      "loss": 0.194,
+      "step": 953
+    },
+    {
+      "epoch": 0.015264,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.926935483870969e-05,
+      "loss": 0.1976,
+      "step": 954
+    },
+    {
+      "epoch": 0.01528,
+      "grad_norm": 1.5,
+      "learning_rate": 9.926774193548387e-05,
+      "loss": 0.1999,
+      "step": 955
+    },
+    {
+      "epoch": 0.015296,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.926612903225807e-05,
+      "loss": 0.2344,
+      "step": 956
+    },
+    {
+      "epoch": 0.015312,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.926451612903226e-05,
+      "loss": 0.2161,
+      "step": 957
+    },
+    {
+      "epoch": 0.015328,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.926290322580646e-05,
+      "loss": 0.2132,
+      "step": 958
+    },
+    {
+      "epoch": 0.015344,
+      "grad_norm": 1.453125,
+      "learning_rate": 9.926129032258064e-05,
+      "loss": 0.2381,
+      "step": 959
+    },
+    {
+      "epoch": 0.01536,
+      "grad_norm": 1.4921875,
+      "learning_rate": 9.925967741935484e-05,
+      "loss": 0.2242,
+      "step": 960
+    },
+    {
+      "epoch": 0.015376,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.925806451612903e-05,
+      "loss": 0.2199,
+      "step": 961
+    },
+    {
+      "epoch": 0.015392,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.925645161290323e-05,
+      "loss": 0.2266,
+      "step": 962
+    },
+    {
+      "epoch": 0.015408,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.925483870967743e-05,
+      "loss": 0.232,
+      "step": 963
+    },
+    {
+      "epoch": 0.015424,
+      "grad_norm": 1.3828125,
+      "learning_rate": 9.925322580645163e-05,
+      "loss": 0.223,
+      "step": 964
+    },
+    {
+      "epoch": 0.01544,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.925161290322581e-05,
+      "loss": 0.2572,
+      "step": 965
+    },
+    {
+      "epoch": 0.015456,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.925000000000001e-05,
+      "loss": 0.2041,
+      "step": 966
+    },
+    {
+      "epoch": 0.015472,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.92483870967742e-05,
+      "loss": 0.1934,
+      "step": 967
+    },
+    {
+      "epoch": 0.015488,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.924677419354839e-05,
+      "loss": 0.2442,
+      "step": 968
+    },
+    {
+      "epoch": 0.015504,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.924516129032259e-05,
+      "loss": 0.2174,
+      "step": 969
+    },
+    {
+      "epoch": 0.01552,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.924354838709677e-05,
+      "loss": 0.1831,
+      "step": 970
+    },
+    {
+      "epoch": 0.015536,
+      "grad_norm": 1.5078125,
+      "learning_rate": 9.924193548387097e-05,
+      "loss": 0.2353,
+      "step": 971
+    },
+    {
+      "epoch": 0.015552,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.924032258064516e-05,
+      "loss": 0.2031,
+      "step": 972
+    },
+    {
+      "epoch": 0.015568,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.923870967741936e-05,
+      "loss": 0.229,
+      "step": 973
+    },
+    {
+      "epoch": 0.015584,
+      "grad_norm": 1.40625,
+      "learning_rate": 9.923709677419356e-05,
+      "loss": 0.2247,
+      "step": 974
+    },
+    {
+      "epoch": 0.0156,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.923548387096774e-05,
+      "loss": 0.2354,
+      "step": 975
+    },
+    {
+      "epoch": 0.015616,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.923387096774194e-05,
+      "loss": 0.1915,
+      "step": 976
+    },
+    {
+      "epoch": 0.015632,
+      "grad_norm": 1.5,
+      "learning_rate": 9.923225806451614e-05,
+      "loss": 0.2383,
+      "step": 977
+    },
+    {
+      "epoch": 0.015648,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.923064516129033e-05,
+      "loss": 0.2187,
+      "step": 978
+    },
+    {
+      "epoch": 0.015664,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.922903225806453e-05,
+      "loss": 0.207,
+      "step": 979
+    },
+    {
+      "epoch": 0.01568,
+      "grad_norm": 1.8984375,
+      "learning_rate": 9.922741935483871e-05,
+      "loss": 0.2065,
+      "step": 980
+    },
+    {
+      "epoch": 0.015696,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.922580645161291e-05,
+      "loss": 0.21,
+      "step": 981
+    },
+    {
+      "epoch": 0.015712,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.92241935483871e-05,
+      "loss": 0.1415,
+      "step": 982
+    },
+    {
+      "epoch": 0.015728,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.92225806451613e-05,
+      "loss": 0.1859,
+      "step": 983
+    },
+    {
+      "epoch": 0.015744,
+      "grad_norm": 1.25,
+      "learning_rate": 9.922096774193549e-05,
+      "loss": 0.2135,
+      "step": 984
+    },
+    {
+      "epoch": 0.01576,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.921935483870967e-05,
+      "loss": 0.2106,
+      "step": 985
+    },
+    {
+      "epoch": 0.015776,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.921774193548387e-05,
+      "loss": 0.2612,
+      "step": 986
+    },
+    {
+      "epoch": 0.015792,
+      "grad_norm": 1.8515625,
+      "learning_rate": 9.921612903225807e-05,
+      "loss": 0.3134,
+      "step": 987
+    },
+    {
+      "epoch": 0.015808,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.921451612903227e-05,
+      "loss": 0.2682,
+      "step": 988
+    },
+    {
+      "epoch": 0.015824,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.921290322580646e-05,
+      "loss": 0.1606,
+      "step": 989
+    },
+    {
+      "epoch": 0.01584,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.921129032258066e-05,
+      "loss": 0.2534,
+      "step": 990
+    },
+    {
+      "epoch": 0.015856,
+      "grad_norm": 1.4921875,
+      "learning_rate": 9.920967741935484e-05,
+      "loss": 0.1944,
+      "step": 991
+    },
+    {
+      "epoch": 0.015872,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.920806451612904e-05,
+      "loss": 0.1889,
+      "step": 992
+    },
+    {
+      "epoch": 0.015888,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.920645161290323e-05,
+      "loss": 0.2028,
+      "step": 993
+    },
+    {
+      "epoch": 0.015904,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.920483870967743e-05,
+      "loss": 0.2315,
+      "step": 994
+    },
+    {
+      "epoch": 0.01592,
+      "grad_norm": 1.4140625,
+      "learning_rate": 9.920322580645161e-05,
+      "loss": 0.242,
+      "step": 995
+    },
+    {
+      "epoch": 0.015936,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.920161290322581e-05,
+      "loss": 0.2328,
+      "step": 996
+    },
+    {
+      "epoch": 0.015952,
+      "grad_norm": 1.4453125,
+      "learning_rate": 9.92e-05,
+      "loss": 0.2631,
+      "step": 997
+    },
+    {
+      "epoch": 0.015968,
+      "grad_norm": 1.4375,
+      "learning_rate": 9.91983870967742e-05,
+      "loss": 0.2386,
+      "step": 998
+    },
+    {
+      "epoch": 0.015984,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.91967741935484e-05,
+      "loss": 0.1959,
+      "step": 999
+    },
+    {
+      "epoch": 0.016,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.919516129032258e-05,
+      "loss": 0.1662,
+      "step": 1000
+    },
+    {
+      "epoch": 0.016016,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.919354838709678e-05,
+      "loss": 0.2002,
+      "step": 1001
+    },
+    {
+      "epoch": 0.016032,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.919193548387097e-05,
+      "loss": 0.207,
+      "step": 1002
+    },
+    {
+      "epoch": 0.016048,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.919032258064517e-05,
+      "loss": 0.2162,
+      "step": 1003
+    },
+    {
+      "epoch": 0.016064,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.918870967741936e-05,
+      "loss": 0.1887,
+      "step": 1004
+    },
+    {
+      "epoch": 0.01608,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.918709677419356e-05,
+      "loss": 0.2322,
+      "step": 1005
+    },
+    {
+      "epoch": 0.016096,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.918548387096774e-05,
+      "loss": 0.1847,
+      "step": 1006
+    },
+    {
+      "epoch": 0.016112,
+      "grad_norm": 1.7578125,
+      "learning_rate": 9.918387096774194e-05,
+      "loss": 0.2638,
+      "step": 1007
+    },
+    {
+      "epoch": 0.016128,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.918225806451613e-05,
+      "loss": 0.2171,
+      "step": 1008
+    },
+    {
+      "epoch": 0.016144,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.918064516129033e-05,
+      "loss": 0.2025,
+      "step": 1009
+    },
+    {
+      "epoch": 0.01616,
+      "grad_norm": 1.0,
+      "learning_rate": 9.917903225806451e-05,
+      "loss": 0.2127,
+      "step": 1010
+    },
+    {
+      "epoch": 0.016176,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.917741935483871e-05,
+      "loss": 0.2048,
+      "step": 1011
+    },
+    {
+      "epoch": 0.016192,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.917580645161291e-05,
+      "loss": 0.2043,
+      "step": 1012
+    },
+    {
+      "epoch": 0.016208,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.917419354838711e-05,
+      "loss": 0.2233,
+      "step": 1013
+    },
+    {
+      "epoch": 0.016224,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.91725806451613e-05,
+      "loss": 0.2175,
+      "step": 1014
+    },
+    {
+      "epoch": 0.01624,
+      "grad_norm": 1.3515625,
+      "learning_rate": 9.917096774193548e-05,
+      "loss": 0.2313,
+      "step": 1015
+    },
+    {
+      "epoch": 0.016256,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.916935483870968e-05,
+      "loss": 0.2434,
+      "step": 1016
+    },
+    {
+      "epoch": 0.016272,
+      "grad_norm": 1.125,
+      "learning_rate": 9.916774193548387e-05,
+      "loss": 0.2499,
+      "step": 1017
+    },
+    {
+      "epoch": 0.016288,
+      "grad_norm": 1.96875,
+      "learning_rate": 9.916612903225807e-05,
+      "loss": 0.2834,
+      "step": 1018
+    },
+    {
+      "epoch": 0.016304,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.916451612903226e-05,
+      "loss": 0.1885,
+      "step": 1019
+    },
+    {
+      "epoch": 0.01632,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.916290322580645e-05,
+      "loss": 0.2378,
+      "step": 1020
+    },
+    {
+      "epoch": 0.016336,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.916129032258064e-05,
+      "loss": 0.2285,
+      "step": 1021
+    },
+    {
+      "epoch": 0.016352,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.915967741935484e-05,
+      "loss": 0.2521,
+      "step": 1022
+    },
+    {
+      "epoch": 0.016368,
+      "grad_norm": 1.5,
+      "learning_rate": 9.915806451612904e-05,
+      "loss": 0.246,
+      "step": 1023
+    },
+    {
+      "epoch": 0.016384,
+      "grad_norm": 1.984375,
+      "learning_rate": 9.915645161290324e-05,
+      "loss": 0.1861,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0164,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.915483870967743e-05,
+      "loss": 0.2249,
+      "step": 1025
+    },
+    {
+      "epoch": 0.016416,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.915322580645163e-05,
+      "loss": 0.2032,
+      "step": 1026
+    },
+    {
+      "epoch": 0.016432,
+      "grad_norm": 1.9296875,
+      "learning_rate": 9.915161290322581e-05,
+      "loss": 0.235,
+      "step": 1027
+    },
+    {
+      "epoch": 0.016448,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.915000000000001e-05,
+      "loss": 0.2254,
+      "step": 1028
+    },
+    {
+      "epoch": 0.016464,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.91483870967742e-05,
+      "loss": 0.2247,
+      "step": 1029
+    },
+    {
+      "epoch": 0.01648,
+      "grad_norm": 1.859375,
+      "learning_rate": 9.91467741935484e-05,
+      "loss": 0.2385,
+      "step": 1030
+    },
+    {
+      "epoch": 0.016496,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.914516129032258e-05,
+      "loss": 0.1808,
+      "step": 1031
+    },
+    {
+      "epoch": 0.016512,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.914354838709677e-05,
+      "loss": 0.1819,
+      "step": 1032
+    },
+    {
+      "epoch": 0.016528,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.914193548387097e-05,
+      "loss": 0.1744,
+      "step": 1033
+    },
+    {
+      "epoch": 0.016544,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.914032258064517e-05,
+      "loss": 0.1631,
+      "step": 1034
+    },
+    {
+      "epoch": 0.01656,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.913870967741937e-05,
+      "loss": 0.1968,
+      "step": 1035
+    },
+    {
+      "epoch": 0.016576,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.913709677419355e-05,
+      "loss": 0.18,
+      "step": 1036
+    },
+    {
+      "epoch": 0.016592,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.913548387096775e-05,
+      "loss": 0.1678,
+      "step": 1037
+    },
+    {
+      "epoch": 0.016608,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.913387096774194e-05,
+      "loss": 0.2357,
+      "step": 1038
+    },
+    {
+      "epoch": 0.016624,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.913225806451614e-05,
+      "loss": 0.1982,
+      "step": 1039
+    },
+    {
+      "epoch": 0.01664,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.913064516129033e-05,
+      "loss": 0.2487,
+      "step": 1040
+    },
+    {
+      "epoch": 0.016656,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.912903225806453e-05,
+      "loss": 0.1986,
+      "step": 1041
+    },
+    {
+      "epoch": 0.016672,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.912741935483871e-05,
+      "loss": 0.1869,
+      "step": 1042
+    },
+    {
+      "epoch": 0.016688,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.912580645161291e-05,
+      "loss": 0.2055,
+      "step": 1043
+    },
+    {
+      "epoch": 0.016704,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.91241935483871e-05,
+      "loss": 0.201,
+      "step": 1044
+    },
+    {
+      "epoch": 0.01672,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.91225806451613e-05,
+      "loss": 0.2129,
+      "step": 1045
+    },
+    {
+      "epoch": 0.016736,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.912096774193548e-05,
+      "loss": 0.22,
+      "step": 1046
+    },
+    {
+      "epoch": 0.016752,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.911935483870968e-05,
+      "loss": 0.2436,
+      "step": 1047
+    },
+    {
+      "epoch": 0.016768,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.911774193548388e-05,
+      "loss": 0.2204,
+      "step": 1048
+    },
+    {
+      "epoch": 0.016784,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.911612903225807e-05,
+      "loss": 0.2015,
+      "step": 1049
+    },
+    {
+      "epoch": 0.0168,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.911451612903227e-05,
+      "loss": 0.2226,
+      "step": 1050
+    },
+    {
+      "epoch": 0.016816,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.911290322580645e-05,
+      "loss": 0.158,
+      "step": 1051
+    },
+    {
+      "epoch": 0.016832,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.911129032258065e-05,
+      "loss": 0.1885,
+      "step": 1052
+    },
+    {
+      "epoch": 0.016848,
+      "grad_norm": 1.515625,
+      "learning_rate": 9.910967741935484e-05,
+      "loss": 0.2088,
+      "step": 1053
+    },
+    {
+      "epoch": 0.016864,
+      "grad_norm": 1.390625,
+      "learning_rate": 9.910806451612904e-05,
+      "loss": 0.18,
+      "step": 1054
+    },
+    {
+      "epoch": 0.01688,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.910645161290323e-05,
+      "loss": 0.1843,
+      "step": 1055
+    },
+    {
+      "epoch": 0.016896,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.910483870967742e-05,
+      "loss": 0.2017,
+      "step": 1056
+    },
+    {
+      "epoch": 0.016912,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.910322580645161e-05,
+      "loss": 0.2178,
+      "step": 1057
+    },
+    {
+      "epoch": 0.016928,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.910161290322581e-05,
+      "loss": 0.2007,
+      "step": 1058
+    },
+    {
+      "epoch": 0.016944,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.910000000000001e-05,
+      "loss": 0.2185,
+      "step": 1059
+    },
+    {
+      "epoch": 0.01696,
+      "grad_norm": 1.328125,
+      "learning_rate": 9.909838709677421e-05,
+      "loss": 0.216,
+      "step": 1060
+    },
+    {
+      "epoch": 0.016976,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.90967741935484e-05,
+      "loss": 0.2053,
+      "step": 1061
+    },
+    {
+      "epoch": 0.016992,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.909516129032258e-05,
+      "loss": 0.1728,
+      "step": 1062
+    },
+    {
+      "epoch": 0.017008,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.909354838709678e-05,
+      "loss": 0.2058,
+      "step": 1063
+    },
+    {
+      "epoch": 0.017024,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.909193548387097e-05,
+      "loss": 0.1749,
+      "step": 1064
+    },
+    {
+      "epoch": 0.01704,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.909032258064517e-05,
+      "loss": 0.1672,
+      "step": 1065
+    },
+    {
+      "epoch": 0.017056,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.908870967741935e-05,
+      "loss": 0.2164,
+      "step": 1066
+    },
+    {
+      "epoch": 0.017072,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.908709677419355e-05,
+      "loss": 0.212,
+      "step": 1067
+    },
+    {
+      "epoch": 0.017088,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.908548387096774e-05,
+      "loss": 0.2312,
+      "step": 1068
+    },
+    {
+      "epoch": 0.017104,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.908387096774194e-05,
+      "loss": 0.1643,
+      "step": 1069
+    },
+    {
+      "epoch": 0.01712,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.908225806451612e-05,
+      "loss": 0.1843,
+      "step": 1070
+    },
+    {
+      "epoch": 0.017136,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.908064516129032e-05,
+      "loss": 0.2651,
+      "step": 1071
+    },
+    {
+      "epoch": 0.017152,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.907903225806452e-05,
+      "loss": 0.1808,
+      "step": 1072
+    },
+    {
+      "epoch": 0.017168,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.907741935483872e-05,
+      "loss": 0.1896,
+      "step": 1073
+    },
+    {
+      "epoch": 0.017184,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.907580645161291e-05,
+      "loss": 0.1553,
+      "step": 1074
+    },
+    {
+      "epoch": 0.0172,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.907419354838711e-05,
+      "loss": 0.208,
+      "step": 1075
+    },
+    {
+      "epoch": 0.017216,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.90725806451613e-05,
+      "loss": 0.2051,
+      "step": 1076
+    },
+    {
+      "epoch": 0.017232,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.907096774193548e-05,
+      "loss": 0.2097,
+      "step": 1077
+    },
+    {
+      "epoch": 0.017248,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.906935483870968e-05,
+      "loss": 0.2535,
+      "step": 1078
+    },
+    {
+      "epoch": 0.017264,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.906774193548387e-05,
+      "loss": 0.2243,
+      "step": 1079
+    },
+    {
+      "epoch": 0.01728,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.906612903225807e-05,
+      "loss": 0.2259,
+      "step": 1080
+    },
+    {
+      "epoch": 0.017296,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.906451612903225e-05,
+      "loss": 0.207,
+      "step": 1081
+    },
+    {
+      "epoch": 0.017312,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.906290322580645e-05,
+      "loss": 0.1751,
+      "step": 1082
+    },
+    {
+      "epoch": 0.017328,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.906129032258065e-05,
+      "loss": 0.2246,
+      "step": 1083
+    },
+    {
+      "epoch": 0.017344,
+      "grad_norm": 1.390625,
+      "learning_rate": 9.905967741935485e-05,
+      "loss": 0.2195,
+      "step": 1084
+    },
+    {
+      "epoch": 0.01736,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.905806451612904e-05,
+      "loss": 0.1691,
+      "step": 1085
+    },
+    {
+      "epoch": 0.017376,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.905645161290324e-05,
+      "loss": 0.2114,
+      "step": 1086
+    },
+    {
+      "epoch": 0.017392,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.905483870967742e-05,
+      "loss": 0.1908,
+      "step": 1087
+    },
+    {
+      "epoch": 0.017408,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.905322580645162e-05,
+      "loss": 0.1846,
+      "step": 1088
+    },
+    {
+      "epoch": 0.017424,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.905161290322581e-05,
+      "loss": 0.2338,
+      "step": 1089
+    },
+    {
+      "epoch": 0.01744,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.905000000000001e-05,
+      "loss": 0.2479,
+      "step": 1090
+    },
+    {
+      "epoch": 0.017456,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.90483870967742e-05,
+      "loss": 0.2415,
+      "step": 1091
+    },
+    {
+      "epoch": 0.017472,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.90467741935484e-05,
+      "loss": 0.1839,
+      "step": 1092
+    },
+    {
+      "epoch": 0.017488,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.904516129032258e-05,
+      "loss": 0.2018,
+      "step": 1093
+    },
+    {
+      "epoch": 0.017504,
+      "grad_norm": 1.3125,
+      "learning_rate": 9.904354838709678e-05,
+      "loss": 0.2404,
+      "step": 1094
+    },
+    {
+      "epoch": 0.01752,
+      "grad_norm": 1.359375,
+      "learning_rate": 9.904193548387098e-05,
+      "loss": 0.1787,
+      "step": 1095
+    },
+    {
+      "epoch": 0.017536,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.904032258064517e-05,
+      "loss": 0.1907,
+      "step": 1096
+    },
+    {
+      "epoch": 0.017552,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.903870967741937e-05,
+      "loss": 0.2577,
+      "step": 1097
+    },
+    {
+      "epoch": 0.017568,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.903709677419355e-05,
+      "loss": 0.2167,
+      "step": 1098
+    },
+    {
+      "epoch": 0.017584,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.903548387096775e-05,
+      "loss": 0.1922,
+      "step": 1099
+    },
+    {
+      "epoch": 0.0176,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.903387096774194e-05,
+      "loss": 0.1884,
+      "step": 1100
+    },
+    {
+      "epoch": 0.017616,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.903225806451614e-05,
+      "loss": 0.1863,
+      "step": 1101
+    },
+    {
+      "epoch": 0.017632,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.903064516129032e-05,
+      "loss": 0.2127,
+      "step": 1102
+    },
+    {
+      "epoch": 0.017648,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.902903225806452e-05,
+      "loss": 0.2418,
+      "step": 1103
+    },
+    {
+      "epoch": 0.017664,
+      "grad_norm": 1.5078125,
+      "learning_rate": 9.902741935483871e-05,
+      "loss": 0.2109,
+      "step": 1104
+    },
+    {
+      "epoch": 0.01768,
+      "grad_norm": 1.4765625,
+      "learning_rate": 9.902580645161291e-05,
+      "loss": 0.2571,
+      "step": 1105
+    },
+    {
+      "epoch": 0.017696,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.90241935483871e-05,
+      "loss": 0.1902,
+      "step": 1106
+    },
+    {
+      "epoch": 0.017712,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.90225806451613e-05,
+      "loss": 0.1861,
+      "step": 1107
+    },
+    {
+      "epoch": 0.017728,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.90209677419355e-05,
+      "loss": 0.2162,
+      "step": 1108
+    },
+    {
+      "epoch": 0.017744,
+      "grad_norm": 1.25,
+      "learning_rate": 9.901935483870968e-05,
+      "loss": 0.206,
+      "step": 1109
+    },
+    {
+      "epoch": 0.01776,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.901774193548388e-05,
+      "loss": 0.1942,
+      "step": 1110
+    },
+    {
+      "epoch": 0.017776,
+      "grad_norm": 1.125,
+      "learning_rate": 9.901612903225807e-05,
+      "loss": 0.248,
+      "step": 1111
+    },
+    {
+      "epoch": 0.017792,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.901451612903227e-05,
+      "loss": 0.1916,
+      "step": 1112
+    },
+    {
+      "epoch": 0.017808,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.901290322580645e-05,
+      "loss": 0.2186,
+      "step": 1113
+    },
+    {
+      "epoch": 0.017824,
+      "grad_norm": 1.796875,
+      "learning_rate": 9.901129032258065e-05,
+      "loss": 0.222,
+      "step": 1114
+    },
+    {
+      "epoch": 0.01784,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.900967741935484e-05,
+      "loss": 0.216,
+      "step": 1115
+    },
+    {
+      "epoch": 0.017856,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.900806451612904e-05,
+      "loss": 0.2351,
+      "step": 1116
+    },
+    {
+      "epoch": 0.017872,
+      "grad_norm": 1.4140625,
+      "learning_rate": 9.900645161290322e-05,
+      "loss": 0.2087,
+      "step": 1117
+    },
+    {
+      "epoch": 0.017888,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.900483870967742e-05,
+      "loss": 0.1896,
+      "step": 1118
+    },
+    {
+      "epoch": 0.017904,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.900322580645162e-05,
+      "loss": 0.2058,
+      "step": 1119
+    },
+    {
+      "epoch": 0.01792,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.900161290322582e-05,
+      "loss": 0.2242,
+      "step": 1120
+    },
+    {
+      "epoch": 0.017936,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.900000000000001e-05,
+      "loss": 0.2271,
+      "step": 1121
+    },
+    {
+      "epoch": 0.017952,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.899838709677421e-05,
+      "loss": 0.2165,
+      "step": 1122
+    },
+    {
+      "epoch": 0.017968,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.89967741935484e-05,
+      "loss": 0.1905,
+      "step": 1123
+    },
+    {
+      "epoch": 0.017984,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.899516129032258e-05,
+      "loss": 0.1815,
+      "step": 1124
+    },
+    {
+      "epoch": 0.018,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.899354838709678e-05,
+      "loss": 0.2033,
+      "step": 1125
+    },
+    {
+      "epoch": 0.018016,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.899193548387097e-05,
+      "loss": 0.162,
+      "step": 1126
+    },
+    {
+      "epoch": 0.018032,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.899032258064516e-05,
+      "loss": 0.1874,
+      "step": 1127
+    },
+    {
+      "epoch": 0.018048,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.898870967741935e-05,
+      "loss": 0.1711,
+      "step": 1128
+    },
+    {
+      "epoch": 0.018064,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.898709677419355e-05,
+      "loss": 0.223,
+      "step": 1129
+    },
+    {
+      "epoch": 0.01808,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.898548387096775e-05,
+      "loss": 0.2065,
+      "step": 1130
+    },
+    {
+      "epoch": 0.018096,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.898387096774195e-05,
+      "loss": 0.2231,
+      "step": 1131
+    },
+    {
+      "epoch": 0.018112,
+      "grad_norm": 1.0,
+      "learning_rate": 9.898225806451614e-05,
+      "loss": 0.1947,
+      "step": 1132
+    },
+    {
+      "epoch": 0.018128,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.898064516129034e-05,
+      "loss": 0.1846,
+      "step": 1133
+    },
+    {
+      "epoch": 0.018144,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.897903225806452e-05,
+      "loss": 0.2082,
+      "step": 1134
+    },
+    {
+      "epoch": 0.01816,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.897741935483872e-05,
+      "loss": 0.2343,
+      "step": 1135
+    },
+    {
+      "epoch": 0.018176,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.897580645161291e-05,
+      "loss": 0.2419,
+      "step": 1136
+    },
+    {
+      "epoch": 0.018192,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.897419354838711e-05,
+      "loss": 0.2,
+      "step": 1137
+    },
+    {
+      "epoch": 0.018208,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.897258064516129e-05,
+      "loss": 0.2,
+      "step": 1138
+    },
+    {
+      "epoch": 0.018224,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.897096774193549e-05,
+      "loss": 0.2162,
+      "step": 1139
+    },
+    {
+      "epoch": 0.01824,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.896935483870968e-05,
+      "loss": 0.1667,
+      "step": 1140
+    },
+    {
+      "epoch": 0.018256,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.896774193548386e-05,
+      "loss": 0.2513,
+      "step": 1141
+    },
+    {
+      "epoch": 0.018272,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.896612903225806e-05,
+      "loss": 0.2293,
+      "step": 1142
+    },
+    {
+      "epoch": 0.018288,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.896451612903226e-05,
+      "loss": 0.2194,
+      "step": 1143
+    },
+    {
+      "epoch": 0.018304,
+      "grad_norm": 1.390625,
+      "learning_rate": 9.896290322580646e-05,
+      "loss": 0.1985,
+      "step": 1144
+    },
+    {
+      "epoch": 0.01832,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.896129032258065e-05,
+      "loss": 0.2246,
+      "step": 1145
+    },
+    {
+      "epoch": 0.018336,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.895967741935485e-05,
+      "loss": 0.2066,
+      "step": 1146
+    },
+    {
+      "epoch": 0.018352,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.895806451612904e-05,
+      "loss": 0.2621,
+      "step": 1147
+    },
+    {
+      "epoch": 0.018368,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.895645161290323e-05,
+      "loss": 0.1865,
+      "step": 1148
+    },
+    {
+      "epoch": 0.018384,
+      "grad_norm": 1.3515625,
+      "learning_rate": 9.895483870967742e-05,
+      "loss": 0.3197,
+      "step": 1149
+    },
+    {
+      "epoch": 0.0184,
+      "grad_norm": 1.5234375,
+      "learning_rate": 9.895322580645162e-05,
+      "loss": 0.2556,
+      "step": 1150
+    },
+    {
+      "epoch": 0.018416,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.89516129032258e-05,
+      "loss": 0.244,
+      "step": 1151
+    },
+    {
+      "epoch": 0.018432,
+      "grad_norm": 1.4375,
+      "learning_rate": 9.895e-05,
+      "loss": 0.2353,
+      "step": 1152
+    },
+    {
+      "epoch": 0.018448,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.894838709677419e-05,
+      "loss": 0.2094,
+      "step": 1153
+    },
+    {
+      "epoch": 0.018464,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.894677419354839e-05,
+      "loss": 0.2157,
+      "step": 1154
+    },
+    {
+      "epoch": 0.01848,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.894516129032259e-05,
+      "loss": 0.2361,
+      "step": 1155
+    },
+    {
+      "epoch": 0.018496,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.894354838709678e-05,
+      "loss": 0.1895,
+      "step": 1156
+    },
+    {
+      "epoch": 0.018512,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.894193548387098e-05,
+      "loss": 0.2365,
+      "step": 1157
+    },
+    {
+      "epoch": 0.018528,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.894032258064516e-05,
+      "loss": 0.2125,
+      "step": 1158
+    },
+    {
+      "epoch": 0.018544,
+      "grad_norm": 1.3125,
+      "learning_rate": 9.893870967741936e-05,
+      "loss": 0.2057,
+      "step": 1159
+    },
+    {
+      "epoch": 0.01856,
+      "grad_norm": 1.40625,
+      "learning_rate": 9.893709677419355e-05,
+      "loss": 0.2017,
+      "step": 1160
+    },
+    {
+      "epoch": 0.018576,
+      "grad_norm": 1.3515625,
+      "learning_rate": 9.893548387096775e-05,
+      "loss": 0.2241,
+      "step": 1161
+    },
+    {
+      "epoch": 0.018592,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.893387096774193e-05,
+      "loss": 0.2166,
+      "step": 1162
+    },
+    {
+      "epoch": 0.018608,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.893225806451613e-05,
+      "loss": 0.1854,
+      "step": 1163
+    },
+    {
+      "epoch": 0.018624,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.893064516129032e-05,
+      "loss": 0.2416,
+      "step": 1164
+    },
+    {
+      "epoch": 0.01864,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.892903225806452e-05,
+      "loss": 0.2084,
+      "step": 1165
+    },
+    {
+      "epoch": 0.018656,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.89274193548387e-05,
+      "loss": 0.2347,
+      "step": 1166
+    },
+    {
+      "epoch": 0.018672,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.89258064516129e-05,
+      "loss": 0.1768,
+      "step": 1167
+    },
+    {
+      "epoch": 0.018688,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.89241935483871e-05,
+      "loss": 0.2025,
+      "step": 1168
+    },
+    {
+      "epoch": 0.018704,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.89225806451613e-05,
+      "loss": 0.1906,
+      "step": 1169
+    },
+    {
+      "epoch": 0.01872,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.892096774193549e-05,
+      "loss": 0.2282,
+      "step": 1170
+    },
+    {
+      "epoch": 0.018736,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.891935483870968e-05,
+      "loss": 0.2181,
+      "step": 1171
+    },
+    {
+      "epoch": 0.018752,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.891774193548388e-05,
+      "loss": 0.2009,
+      "step": 1172
+    },
+    {
+      "epoch": 0.018768,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.891612903225806e-05,
+      "loss": 0.1886,
+      "step": 1173
+    },
+    {
+      "epoch": 0.018784,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.891451612903226e-05,
+      "loss": 0.1975,
+      "step": 1174
+    },
+    {
+      "epoch": 0.0188,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.891290322580645e-05,
+      "loss": 0.1795,
+      "step": 1175
+    },
+    {
+      "epoch": 0.018816,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.891129032258065e-05,
+      "loss": 0.2008,
+      "step": 1176
+    },
+    {
+      "epoch": 0.018832,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.890967741935483e-05,
+      "loss": 0.1956,
+      "step": 1177
+    },
+    {
+      "epoch": 0.018848,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.890806451612903e-05,
+      "loss": 0.1974,
+      "step": 1178
+    },
+    {
+      "epoch": 0.018864,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.890645161290323e-05,
+      "loss": 0.2266,
+      "step": 1179
+    },
+    {
+      "epoch": 0.01888,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.890483870967743e-05,
+      "loss": 0.2059,
+      "step": 1180
+    },
+    {
+      "epoch": 0.018896,
+      "grad_norm": 1.453125,
+      "learning_rate": 9.890322580645162e-05,
+      "loss": 0.2293,
+      "step": 1181
+    },
+    {
+      "epoch": 0.018912,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.890161290322582e-05,
+      "loss": 0.2416,
+      "step": 1182
+    },
+    {
+      "epoch": 0.018928,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.89e-05,
+      "loss": 0.2368,
+      "step": 1183
+    },
+    {
+      "epoch": 0.018944,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.88983870967742e-05,
+      "loss": 0.1421,
+      "step": 1184
+    },
+    {
+      "epoch": 0.01896,
+      "grad_norm": 1.625,
+      "learning_rate": 9.889677419354839e-05,
+      "loss": 0.2286,
+      "step": 1185
+    },
+    {
+      "epoch": 0.018976,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.889516129032258e-05,
+      "loss": 0.239,
+      "step": 1186
+    },
+    {
+      "epoch": 0.018992,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.889354838709678e-05,
+      "loss": 0.1986,
+      "step": 1187
+    },
+    {
+      "epoch": 0.019008,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.889193548387096e-05,
+      "loss": 0.1791,
+      "step": 1188
+    },
+    {
+      "epoch": 0.019024,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.889032258064516e-05,
+      "loss": 0.2323,
+      "step": 1189
+    },
+    {
+      "epoch": 0.01904,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.888870967741936e-05,
+      "loss": 0.2232,
+      "step": 1190
+    },
+    {
+      "epoch": 0.019056,
+      "grad_norm": 1.3828125,
+      "learning_rate": 9.888709677419356e-05,
+      "loss": 0.198,
+      "step": 1191
+    },
+    {
+      "epoch": 0.019072,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.888548387096775e-05,
+      "loss": 0.2354,
+      "step": 1192
+    },
+    {
+      "epoch": 0.019088,
+      "grad_norm": 1.4921875,
+      "learning_rate": 9.888387096774195e-05,
+      "loss": 0.1872,
+      "step": 1193
+    },
+    {
+      "epoch": 0.019104,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.888225806451613e-05,
+      "loss": 0.1889,
+      "step": 1194
+    },
+    {
+      "epoch": 0.01912,
+      "grad_norm": 1.328125,
+      "learning_rate": 9.888064516129033e-05,
+      "loss": 0.1736,
+      "step": 1195
+    },
+    {
+      "epoch": 0.019136,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.887903225806452e-05,
+      "loss": 0.2175,
+      "step": 1196
+    },
+    {
+      "epoch": 0.019152,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.887741935483872e-05,
+      "loss": 0.2099,
+      "step": 1197
+    },
+    {
+      "epoch": 0.019168,
+      "grad_norm": 1.7578125,
+      "learning_rate": 9.88758064516129e-05,
+      "loss": 0.2211,
+      "step": 1198
+    },
+    {
+      "epoch": 0.019184,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.88741935483871e-05,
+      "loss": 0.1707,
+      "step": 1199
+    },
+    {
+      "epoch": 0.0192,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.887258064516129e-05,
+      "loss": 0.1989,
+      "step": 1200
+    },
+    {
+      "epoch": 0.019216,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.887096774193549e-05,
+      "loss": 0.1845,
+      "step": 1201
+    },
+    {
+      "epoch": 0.019232,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.886935483870968e-05,
+      "loss": 0.2616,
+      "step": 1202
+    },
+    {
+      "epoch": 0.019248,
+      "grad_norm": 1.875,
+      "learning_rate": 9.886774193548388e-05,
+      "loss": 0.2296,
+      "step": 1203
+    },
+    {
+      "epoch": 0.019264,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.886612903225808e-05,
+      "loss": 0.2217,
+      "step": 1204
+    },
+    {
+      "epoch": 0.01928,
+      "grad_norm": 1.7734375,
+      "learning_rate": 9.886451612903226e-05,
+      "loss": 0.2633,
+      "step": 1205
+    },
+    {
+      "epoch": 0.019296,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.886290322580646e-05,
+      "loss": 0.194,
+      "step": 1206
+    },
+    {
+      "epoch": 0.019312,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.886129032258065e-05,
+      "loss": 0.1932,
+      "step": 1207
+    },
+    {
+      "epoch": 0.019328,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.885967741935485e-05,
+      "loss": 0.1534,
+      "step": 1208
+    },
+    {
+      "epoch": 0.019344,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.885806451612903e-05,
+      "loss": 0.1715,
+      "step": 1209
+    },
+    {
+      "epoch": 0.01936,
+      "grad_norm": 1.328125,
+      "learning_rate": 9.885645161290323e-05,
+      "loss": 0.1853,
+      "step": 1210
+    },
+    {
+      "epoch": 0.019376,
+      "grad_norm": 10.875,
+      "learning_rate": 9.885483870967742e-05,
+      "loss": 0.2495,
+      "step": 1211
+    },
+    {
+      "epoch": 0.019392,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.885322580645162e-05,
+      "loss": 0.3524,
+      "step": 1212
+    },
+    {
+      "epoch": 0.019408,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.88516129032258e-05,
+      "loss": 0.2729,
+      "step": 1213
+    },
+    {
+      "epoch": 0.019424,
+      "grad_norm": 3.671875,
+      "learning_rate": 9.885e-05,
+      "loss": 0.3281,
+      "step": 1214
+    },
+    {
+      "epoch": 0.01944,
+      "grad_norm": 2.25,
+      "learning_rate": 9.88483870967742e-05,
+      "loss": 0.2858,
+      "step": 1215
+    },
+    {
+      "epoch": 0.019456,
+      "grad_norm": 1.3359375,
+      "learning_rate": 9.88467741935484e-05,
+      "loss": 0.218,
+      "step": 1216
+    },
+    {
+      "epoch": 0.019472,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.884516129032259e-05,
+      "loss": 0.2035,
+      "step": 1217
+    },
+    {
+      "epoch": 0.019488,
+      "grad_norm": 1.375,
+      "learning_rate": 9.884354838709678e-05,
+      "loss": 0.2551,
+      "step": 1218
+    },
+    {
+      "epoch": 0.019504,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.884193548387097e-05,
+      "loss": 0.2445,
+      "step": 1219
+    },
+    {
+      "epoch": 0.01952,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.884032258064516e-05,
+      "loss": 0.2545,
+      "step": 1220
+    },
+    {
+      "epoch": 0.019536,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.883870967741936e-05,
+      "loss": 0.2329,
+      "step": 1221
+    },
+    {
+      "epoch": 0.019552,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.883709677419355e-05,
+      "loss": 0.2113,
+      "step": 1222
+    },
+    {
+      "epoch": 0.019568,
+      "grad_norm": 1.3125,
+      "learning_rate": 9.883548387096775e-05,
+      "loss": 0.2868,
+      "step": 1223
+    },
+    {
+      "epoch": 0.019584,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.883387096774193e-05,
+      "loss": 0.2239,
+      "step": 1224
+    },
+    {
+      "epoch": 0.0196,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.883225806451613e-05,
+      "loss": 0.1683,
+      "step": 1225
+    },
+    {
+      "epoch": 0.019616,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.883064516129033e-05,
+      "loss": 0.2237,
+      "step": 1226
+    },
+    {
+      "epoch": 0.019632,
+      "grad_norm": 9.1875,
+      "learning_rate": 9.882903225806452e-05,
+      "loss": 0.2141,
+      "step": 1227
+    },
+    {
+      "epoch": 0.019648,
+      "grad_norm": 1.8359375,
+      "learning_rate": 9.882741935483872e-05,
+      "loss": 0.2238,
+      "step": 1228
+    },
+    {
+      "epoch": 0.019664,
+      "grad_norm": 1.5234375,
+      "learning_rate": 9.882580645161292e-05,
+      "loss": 0.2205,
+      "step": 1229
+    },
+    {
+      "epoch": 0.01968,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.88241935483871e-05,
+      "loss": 0.2032,
+      "step": 1230
+    },
+    {
+      "epoch": 0.019696,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.88225806451613e-05,
+      "loss": 0.2567,
+      "step": 1231
+    },
+    {
+      "epoch": 0.019712,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.882096774193549e-05,
+      "loss": 0.2728,
+      "step": 1232
+    },
+    {
+      "epoch": 0.019728,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.881935483870967e-05,
+      "loss": 0.2114,
+      "step": 1233
+    },
+    {
+      "epoch": 0.019744,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.881774193548387e-05,
+      "loss": 0.2291,
+      "step": 1234
+    },
+    {
+      "epoch": 0.01976,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.881612903225806e-05,
+      "loss": 0.2492,
+      "step": 1235
+    },
+    {
+      "epoch": 0.019776,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.881451612903226e-05,
+      "loss": 0.2151,
+      "step": 1236
+    },
+    {
+      "epoch": 0.019792,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.881290322580645e-05,
+      "loss": 0.2046,
+      "step": 1237
+    },
+    {
+      "epoch": 0.019808,
+      "grad_norm": 1.25,
+      "learning_rate": 9.881129032258065e-05,
+      "loss": 0.246,
+      "step": 1238
+    },
+    {
+      "epoch": 0.019824,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.880967741935485e-05,
+      "loss": 0.1915,
+      "step": 1239
+    },
+    {
+      "epoch": 0.01984,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.880806451612905e-05,
+      "loss": 0.208,
+      "step": 1240
+    },
+    {
+      "epoch": 0.019856,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.880645161290323e-05,
+      "loss": 0.2242,
+      "step": 1241
+    },
+    {
+      "epoch": 0.019872,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.880483870967743e-05,
+      "loss": 0.2281,
+      "step": 1242
+    },
+    {
+      "epoch": 0.019888,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.880322580645162e-05,
+      "loss": 0.2424,
+      "step": 1243
+    },
+    {
+      "epoch": 0.019904,
+      "grad_norm": 1.34375,
+      "learning_rate": 9.880161290322582e-05,
+      "loss": 0.2099,
+      "step": 1244
+    },
+    {
+      "epoch": 0.01992,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.88e-05,
+      "loss": 0.195,
+      "step": 1245
+    },
+    {
+      "epoch": 0.019936,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.87983870967742e-05,
+      "loss": 0.1791,
+      "step": 1246
+    },
+    {
+      "epoch": 0.019952,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.879677419354839e-05,
+      "loss": 0.2025,
+      "step": 1247
+    },
+    {
+      "epoch": 0.019968,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.879516129032257e-05,
+      "loss": 0.229,
+      "step": 1248
+    },
+    {
+      "epoch": 0.019984,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.879354838709677e-05,
+      "loss": 0.2816,
+      "step": 1249
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.879193548387097e-05,
+      "loss": 0.2522,
+      "step": 1250
+    },
+    {
+      "epoch": 0.020016,
+      "grad_norm": 1.4921875,
+      "learning_rate": 9.879032258064517e-05,
+      "loss": 0.1856,
+      "step": 1251
+    },
+    {
+      "epoch": 0.020032,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.878870967741936e-05,
+      "loss": 0.1926,
+      "step": 1252
+    },
+    {
+      "epoch": 0.020048,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.878709677419356e-05,
+      "loss": 0.1858,
+      "step": 1253
+    },
+    {
+      "epoch": 0.020064,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.878548387096775e-05,
+      "loss": 0.2152,
+      "step": 1254
+    },
+    {
+      "epoch": 0.02008,
+      "grad_norm": 1.5703125,
+      "learning_rate": 9.878387096774194e-05,
+      "loss": 0.2127,
+      "step": 1255
+    },
+    {
+      "epoch": 0.020096,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.878225806451613e-05,
+      "loss": 0.2178,
+      "step": 1256
+    },
+    {
+      "epoch": 0.020112,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.878064516129033e-05,
+      "loss": 0.2336,
+      "step": 1257
+    },
+    {
+      "epoch": 0.020128,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.877903225806452e-05,
+      "loss": 0.1861,
+      "step": 1258
+    },
+    {
+      "epoch": 0.020144,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.877741935483872e-05,
+      "loss": 0.2286,
+      "step": 1259
+    },
+    {
+      "epoch": 0.02016,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.87758064516129e-05,
+      "loss": 0.194,
+      "step": 1260
+    },
+    {
+      "epoch": 0.020176,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.87741935483871e-05,
+      "loss": 0.1988,
+      "step": 1261
+    },
+    {
+      "epoch": 0.020192,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.877258064516129e-05,
+      "loss": 0.2325,
+      "step": 1262
+    },
+    {
+      "epoch": 0.020208,
+      "grad_norm": 1.359375,
+      "learning_rate": 9.877096774193549e-05,
+      "loss": 0.2329,
+      "step": 1263
+    },
+    {
+      "epoch": 0.020224,
+      "grad_norm": 1.296875,
+      "learning_rate": 9.876935483870969e-05,
+      "loss": 0.26,
+      "step": 1264
+    },
+    {
+      "epoch": 0.02024,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.876774193548387e-05,
+      "loss": 0.2088,
+      "step": 1265
+    },
+    {
+      "epoch": 0.020256,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.876612903225807e-05,
+      "loss": 0.2417,
+      "step": 1266
+    },
+    {
+      "epoch": 0.020272,
+      "grad_norm": 1.5390625,
+      "learning_rate": 9.876451612903226e-05,
+      "loss": 0.1728,
+      "step": 1267
+    },
+    {
+      "epoch": 0.020288,
+      "grad_norm": 1.4296875,
+      "learning_rate": 9.876290322580646e-05,
+      "loss": 0.2169,
+      "step": 1268
+    },
+    {
+      "epoch": 0.020304,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.876129032258064e-05,
+      "loss": 0.2079,
+      "step": 1269
+    },
+    {
+      "epoch": 0.02032,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.875967741935484e-05,
+      "loss": 0.2018,
+      "step": 1270
+    },
+    {
+      "epoch": 0.020336,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.875806451612903e-05,
+      "loss": 0.2072,
+      "step": 1271
+    },
+    {
+      "epoch": 0.020352,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.875645161290323e-05,
+      "loss": 0.2467,
+      "step": 1272
+    },
+    {
+      "epoch": 0.020368,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.875483870967742e-05,
+      "loss": 0.188,
+      "step": 1273
+    },
+    {
+      "epoch": 0.020384,
+      "grad_norm": 2.875,
+      "learning_rate": 9.875322580645162e-05,
+      "loss": 0.2778,
+      "step": 1274
+    },
+    {
+      "epoch": 0.0204,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.875161290322582e-05,
+      "loss": 0.208,
+      "step": 1275
+    },
+    {
+      "epoch": 0.020416,
+      "grad_norm": 2.0,
+      "learning_rate": 9.875000000000002e-05,
+      "loss": 0.2241,
+      "step": 1276
+    },
+    {
+      "epoch": 0.020432,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.87483870967742e-05,
+      "loss": 0.1965,
+      "step": 1277
+    },
+    {
+      "epoch": 0.020448,
+      "grad_norm": 1.8515625,
+      "learning_rate": 9.87467741935484e-05,
+      "loss": 0.2645,
+      "step": 1278
+    },
+    {
+      "epoch": 0.020464,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.874516129032259e-05,
+      "loss": 0.1568,
+      "step": 1279
+    },
+    {
+      "epoch": 0.02048,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.874354838709677e-05,
+      "loss": 0.2537,
+      "step": 1280
+    },
+    {
+      "epoch": 0.020496,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.874193548387097e-05,
+      "loss": 0.2086,
+      "step": 1281
+    },
+    {
+      "epoch": 0.020512,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.874032258064516e-05,
+      "loss": 0.2338,
+      "step": 1282
+    },
+    {
+      "epoch": 0.020528,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.873870967741936e-05,
+      "loss": 0.2799,
+      "step": 1283
+    },
+    {
+      "epoch": 0.020544,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.873709677419354e-05,
+      "loss": 0.1875,
+      "step": 1284
+    },
+    {
+      "epoch": 0.02056,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.873548387096774e-05,
+      "loss": 0.2126,
+      "step": 1285
+    },
+    {
+      "epoch": 0.020576,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.873387096774194e-05,
+      "loss": 0.1947,
+      "step": 1286
+    },
+    {
+      "epoch": 0.020592,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.873225806451614e-05,
+      "loss": 0.2274,
+      "step": 1287
+    },
+    {
+      "epoch": 0.020608,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.873064516129033e-05,
+      "loss": 0.1968,
+      "step": 1288
+    },
+    {
+      "epoch": 0.020624,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.872903225806453e-05,
+      "loss": 0.1658,
+      "step": 1289
+    },
+    {
+      "epoch": 0.02064,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.872741935483872e-05,
+      "loss": 0.1785,
+      "step": 1290
+    },
+    {
+      "epoch": 0.020656,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.872580645161291e-05,
+      "loss": 0.2014,
+      "step": 1291
+    },
+    {
+      "epoch": 0.020672,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.87241935483871e-05,
+      "loss": 0.2351,
+      "step": 1292
+    },
+    {
+      "epoch": 0.020688,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.87225806451613e-05,
+      "loss": 0.1873,
+      "step": 1293
+    },
+    {
+      "epoch": 0.020704,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.872096774193549e-05,
+      "loss": 0.1996,
+      "step": 1294
+    },
+    {
+      "epoch": 0.02072,
+      "grad_norm": 1.8203125,
+      "learning_rate": 9.871935483870967e-05,
+      "loss": 0.2115,
+      "step": 1295
+    },
+    {
+      "epoch": 0.020736,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.871774193548387e-05,
+      "loss": 0.1733,
+      "step": 1296
+    },
+    {
+      "epoch": 0.020752,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.871612903225806e-05,
+      "loss": 0.2223,
+      "step": 1297
+    },
+    {
+      "epoch": 0.020768,
+      "grad_norm": 1.5,
+      "learning_rate": 9.871451612903226e-05,
+      "loss": 0.2067,
+      "step": 1298
+    },
+    {
+      "epoch": 0.020784,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.871290322580646e-05,
+      "loss": 0.2126,
+      "step": 1299
+    },
+    {
+      "epoch": 0.0208,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.871129032258066e-05,
+      "loss": 0.2219,
+      "step": 1300
+    },
+    {
+      "epoch": 0.020816,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.870967741935484e-05,
+      "loss": 0.2112,
+      "step": 1301
+    },
+    {
+      "epoch": 0.020832,
+      "grad_norm": 1.390625,
+      "learning_rate": 9.870806451612904e-05,
+      "loss": 0.2363,
+      "step": 1302
+    },
+    {
+      "epoch": 0.020848,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.870645161290323e-05,
+      "loss": 0.2462,
+      "step": 1303
+    },
+    {
+      "epoch": 0.020864,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.870483870967743e-05,
+      "loss": 0.18,
+      "step": 1304
+    },
+    {
+      "epoch": 0.02088,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.870322580645161e-05,
+      "loss": 0.2025,
+      "step": 1305
+    },
+    {
+      "epoch": 0.020896,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.870161290322581e-05,
+      "loss": 0.2045,
+      "step": 1306
+    },
+    {
+      "epoch": 0.020912,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.87e-05,
+      "loss": 0.1742,
+      "step": 1307
+    },
+    {
+      "epoch": 0.020928,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.86983870967742e-05,
+      "loss": 0.1951,
+      "step": 1308
+    },
+    {
+      "epoch": 0.020944,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.869677419354839e-05,
+      "loss": 0.2063,
+      "step": 1309
+    },
+    {
+      "epoch": 0.02096,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.869516129032259e-05,
+      "loss": 0.1812,
+      "step": 1310
+    },
+    {
+      "epoch": 0.020976,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.869354838709679e-05,
+      "loss": 0.1968,
+      "step": 1311
+    },
+    {
+      "epoch": 0.020992,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.869193548387097e-05,
+      "loss": 0.1855,
+      "step": 1312
+    },
+    {
+      "epoch": 0.021008,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.869032258064517e-05,
+      "loss": 0.2178,
+      "step": 1313
+    },
+    {
+      "epoch": 0.021024,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.868870967741936e-05,
+      "loss": 0.275,
+      "step": 1314
+    },
+    {
+      "epoch": 0.02104,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.868709677419356e-05,
+      "loss": 0.2058,
+      "step": 1315
+    },
+    {
+      "epoch": 0.021056,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.868548387096774e-05,
+      "loss": 0.1859,
+      "step": 1316
+    },
+    {
+      "epoch": 0.021072,
+      "grad_norm": 1.3359375,
+      "learning_rate": 9.868387096774194e-05,
+      "loss": 0.2666,
+      "step": 1317
+    },
+    {
+      "epoch": 0.021088,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.868225806451613e-05,
+      "loss": 0.2267,
+      "step": 1318
+    },
+    {
+      "epoch": 0.021104,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.868064516129033e-05,
+      "loss": 0.2132,
+      "step": 1319
+    },
+    {
+      "epoch": 0.02112,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.867903225806451e-05,
+      "loss": 0.2255,
+      "step": 1320
+    },
+    {
+      "epoch": 0.021136,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.867741935483871e-05,
+      "loss": 0.1854,
+      "step": 1321
+    },
+    {
+      "epoch": 0.021152,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.86758064516129e-05,
+      "loss": 0.1962,
+      "step": 1322
+    },
+    {
+      "epoch": 0.021168,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.86741935483871e-05,
+      "loss": 0.2275,
+      "step": 1323
+    },
+    {
+      "epoch": 0.021184,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.86725806451613e-05,
+      "loss": 0.2414,
+      "step": 1324
+    },
+    {
+      "epoch": 0.0212,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.86709677419355e-05,
+      "loss": 0.1907,
+      "step": 1325
+    },
+    {
+      "epoch": 0.021216,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.866935483870968e-05,
+      "loss": 0.2165,
+      "step": 1326
+    },
+    {
+      "epoch": 0.021232,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.866774193548387e-05,
+      "loss": 0.2103,
+      "step": 1327
+    },
+    {
+      "epoch": 0.021248,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.866612903225807e-05,
+      "loss": 0.2099,
+      "step": 1328
+    },
+    {
+      "epoch": 0.021264,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.866451612903226e-05,
+      "loss": 0.2203,
+      "step": 1329
+    },
+    {
+      "epoch": 0.02128,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.866290322580646e-05,
+      "loss": 0.2416,
+      "step": 1330
+    },
+    {
+      "epoch": 0.021296,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.866129032258064e-05,
+      "loss": 0.174,
+      "step": 1331
+    },
+    {
+      "epoch": 0.021312,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.865967741935484e-05,
+      "loss": 0.2119,
+      "step": 1332
+    },
+    {
+      "epoch": 0.021328,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.865806451612903e-05,
+      "loss": 0.1738,
+      "step": 1333
+    },
+    {
+      "epoch": 0.021344,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.865645161290323e-05,
+      "loss": 0.1781,
+      "step": 1334
+    },
+    {
+      "epoch": 0.02136,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.865483870967743e-05,
+      "loss": 0.277,
+      "step": 1335
+    },
+    {
+      "epoch": 0.021376,
+      "grad_norm": 1.484375,
+      "learning_rate": 9.865322580645163e-05,
+      "loss": 0.1941,
+      "step": 1336
+    },
+    {
+      "epoch": 0.021392,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.865161290322581e-05,
+      "loss": 0.2036,
+      "step": 1337
+    },
+    {
+      "epoch": 0.021408,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.865000000000001e-05,
+      "loss": 0.2316,
+      "step": 1338
+    },
+    {
+      "epoch": 0.021424,
+      "grad_norm": 1.4609375,
+      "learning_rate": 9.86483870967742e-05,
+      "loss": 0.2361,
+      "step": 1339
+    },
+    {
+      "epoch": 0.02144,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.86467741935484e-05,
+      "loss": 0.1855,
+      "step": 1340
+    },
+    {
+      "epoch": 0.021456,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.864516129032258e-05,
+      "loss": 0.2048,
+      "step": 1341
+    },
+    {
+      "epoch": 0.021472,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.864354838709677e-05,
+      "loss": 0.2819,
+      "step": 1342
+    },
+    {
+      "epoch": 0.021488,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.864193548387097e-05,
+      "loss": 0.2102,
+      "step": 1343
+    },
+    {
+      "epoch": 0.021504,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.864032258064516e-05,
+      "loss": 0.2203,
+      "step": 1344
+    },
+    {
+      "epoch": 0.02152,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.863870967741936e-05,
+      "loss": 0.1989,
+      "step": 1345
+    },
+    {
+      "epoch": 0.021536,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.863709677419356e-05,
+      "loss": 0.2025,
+      "step": 1346
+    },
+    {
+      "epoch": 0.021552,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.863548387096776e-05,
+      "loss": 0.2192,
+      "step": 1347
+    },
+    {
+      "epoch": 0.021568,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.863387096774194e-05,
+      "loss": 0.1933,
+      "step": 1348
+    },
+    {
+      "epoch": 0.021584,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.863225806451614e-05,
+      "loss": 0.1964,
+      "step": 1349
+    },
+    {
+      "epoch": 0.0216,
+      "grad_norm": 1.390625,
+      "learning_rate": 9.863064516129033e-05,
+      "loss": 0.2593,
+      "step": 1350
+    },
+    {
+      "epoch": 0.021616,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.862903225806453e-05,
+      "loss": 0.2505,
+      "step": 1351
+    },
+    {
+      "epoch": 0.021632,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.862741935483871e-05,
+      "loss": 0.2206,
+      "step": 1352
+    },
+    {
+      "epoch": 0.021648,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.862580645161291e-05,
+      "loss": 0.2178,
+      "step": 1353
+    },
+    {
+      "epoch": 0.021664,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.86241935483871e-05,
+      "loss": 0.2162,
+      "step": 1354
+    },
+    {
+      "epoch": 0.02168,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.86225806451613e-05,
+      "loss": 0.2629,
+      "step": 1355
+    },
+    {
+      "epoch": 0.021696,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.862096774193548e-05,
+      "loss": 0.229,
+      "step": 1356
+    },
+    {
+      "epoch": 0.021712,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.861935483870967e-05,
+      "loss": 0.2015,
+      "step": 1357
+    },
+    {
+      "epoch": 0.021728,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.861774193548387e-05,
+      "loss": 0.1739,
+      "step": 1358
+    },
+    {
+      "epoch": 0.021744,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.861612903225807e-05,
+      "loss": 0.2015,
+      "step": 1359
+    },
+    {
+      "epoch": 0.02176,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.861451612903227e-05,
+      "loss": 0.2102,
+      "step": 1360
+    },
+    {
+      "epoch": 0.021776,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.861290322580646e-05,
+      "loss": 0.2698,
+      "step": 1361
+    },
+    {
+      "epoch": 0.021792,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.861129032258065e-05,
+      "loss": 0.1994,
+      "step": 1362
+    },
+    {
+      "epoch": 0.021808,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.860967741935484e-05,
+      "loss": 0.1495,
+      "step": 1363
+    },
+    {
+      "epoch": 0.021824,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.860806451612904e-05,
+      "loss": 0.1874,
+      "step": 1364
+    },
+    {
+      "epoch": 0.02184,
+      "grad_norm": 1.4140625,
+      "learning_rate": 9.860645161290323e-05,
+      "loss": 0.2492,
+      "step": 1365
+    },
+    {
+      "epoch": 0.021856,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.860483870967743e-05,
+      "loss": 0.2164,
+      "step": 1366
+    },
+    {
+      "epoch": 0.021872,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.860322580645161e-05,
+      "loss": 0.1751,
+      "step": 1367
+    },
+    {
+      "epoch": 0.021888,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.860161290322581e-05,
+      "loss": 0.1855,
+      "step": 1368
+    },
+    {
+      "epoch": 0.021904,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.86e-05,
+      "loss": 0.1613,
+      "step": 1369
+    },
+    {
+      "epoch": 0.02192,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.85983870967742e-05,
+      "loss": 0.187,
+      "step": 1370
+    },
+    {
+      "epoch": 0.021936,
+      "grad_norm": 1.359375,
+      "learning_rate": 9.85967741935484e-05,
+      "loss": 0.2754,
+      "step": 1371
+    },
+    {
+      "epoch": 0.021952,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.85951612903226e-05,
+      "loss": 0.1967,
+      "step": 1372
+    },
+    {
+      "epoch": 0.021968,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.859354838709678e-05,
+      "loss": 0.1903,
+      "step": 1373
+    },
+    {
+      "epoch": 0.021984,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.859193548387097e-05,
+      "loss": 0.1872,
+      "step": 1374
+    },
+    {
+      "epoch": 0.022,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.859032258064517e-05,
+      "loss": 0.2426,
+      "step": 1375
+    },
+    {
+      "epoch": 0.022016,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.858870967741935e-05,
+      "loss": 0.1942,
+      "step": 1376
+    },
+    {
+      "epoch": 0.022032,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.858709677419355e-05,
+      "loss": 0.2074,
+      "step": 1377
+    },
+    {
+      "epoch": 0.022048,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.858548387096774e-05,
+      "loss": 0.2281,
+      "step": 1378
+    },
+    {
+      "epoch": 0.022064,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.858387096774194e-05,
+      "loss": 0.1981,
+      "step": 1379
+    },
+    {
+      "epoch": 0.02208,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.858225806451613e-05,
+      "loss": 0.2193,
+      "step": 1380
+    },
+    {
+      "epoch": 0.022096,
+      "grad_norm": 1.9296875,
+      "learning_rate": 9.858064516129033e-05,
+      "loss": 0.2399,
+      "step": 1381
+    },
+    {
+      "epoch": 0.022112,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.857903225806453e-05,
+      "loss": 0.1736,
+      "step": 1382
+    },
+    {
+      "epoch": 0.022128,
+      "grad_norm": 1.5390625,
+      "learning_rate": 9.857741935483872e-05,
+      "loss": 0.2203,
+      "step": 1383
+    },
+    {
+      "epoch": 0.022144,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.857580645161291e-05,
+      "loss": 0.205,
+      "step": 1384
+    },
+    {
+      "epoch": 0.02216,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.857419354838711e-05,
+      "loss": 0.1782,
+      "step": 1385
+    },
+    {
+      "epoch": 0.022176,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.85725806451613e-05,
+      "loss": 0.1651,
+      "step": 1386
+    },
+    {
+      "epoch": 0.022192,
+      "grad_norm": 1.40625,
+      "learning_rate": 9.85709677419355e-05,
+      "loss": 0.2106,
+      "step": 1387
+    },
+    {
+      "epoch": 0.022208,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.856935483870968e-05,
+      "loss": 0.1944,
+      "step": 1388
+    },
+    {
+      "epoch": 0.022224,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.856774193548387e-05,
+      "loss": 0.1721,
+      "step": 1389
+    },
+    {
+      "epoch": 0.02224,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.856612903225807e-05,
+      "loss": 0.2068,
+      "step": 1390
+    },
+    {
+      "epoch": 0.022256,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.856451612903225e-05,
+      "loss": 0.1947,
+      "step": 1391
+    },
+    {
+      "epoch": 0.022272,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.856290322580645e-05,
+      "loss": 0.212,
+      "step": 1392
+    },
+    {
+      "epoch": 0.022288,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.856129032258064e-05,
+      "loss": 0.2038,
+      "step": 1393
+    },
+    {
+      "epoch": 0.022304,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.855967741935484e-05,
+      "loss": 0.213,
+      "step": 1394
+    },
+    {
+      "epoch": 0.02232,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.855806451612904e-05,
+      "loss": 0.2634,
+      "step": 1395
+    },
+    {
+      "epoch": 0.022336,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.855645161290324e-05,
+      "loss": 0.2154,
+      "step": 1396
+    },
+    {
+      "epoch": 0.022352,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.855483870967742e-05,
+      "loss": 0.2054,
+      "step": 1397
+    },
+    {
+      "epoch": 0.022368,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.855322580645162e-05,
+      "loss": 0.2183,
+      "step": 1398
+    },
+    {
+      "epoch": 0.022384,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.855161290322581e-05,
+      "loss": 0.1865,
+      "step": 1399
+    },
+    {
+      "epoch": 0.0224,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.855000000000001e-05,
+      "loss": 0.1985,
+      "step": 1400
+    },
+    {
+      "epoch": 0.022416,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.85483870967742e-05,
+      "loss": 0.1847,
+      "step": 1401
+    },
+    {
+      "epoch": 0.022432,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.85467741935484e-05,
+      "loss": 0.2083,
+      "step": 1402
+    },
+    {
+      "epoch": 0.022448,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.854516129032258e-05,
+      "loss": 0.1997,
+      "step": 1403
+    },
+    {
+      "epoch": 0.022464,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.854354838709677e-05,
+      "loss": 0.2236,
+      "step": 1404
+    },
+    {
+      "epoch": 0.02248,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.854193548387097e-05,
+      "loss": 0.1798,
+      "step": 1405
+    },
+    {
+      "epoch": 0.022496,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.854032258064517e-05,
+      "loss": 0.1333,
+      "step": 1406
+    },
+    {
+      "epoch": 0.022512,
+      "grad_norm": 1.328125,
+      "learning_rate": 9.853870967741937e-05,
+      "loss": 0.2366,
+      "step": 1407
+    },
+    {
+      "epoch": 0.022528,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.853709677419355e-05,
+      "loss": 0.2134,
+      "step": 1408
+    },
+    {
+      "epoch": 0.022544,
+      "grad_norm": 1.5078125,
+      "learning_rate": 9.853548387096775e-05,
+      "loss": 0.2551,
+      "step": 1409
+    },
+    {
+      "epoch": 0.02256,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.853387096774194e-05,
+      "loss": 0.2118,
+      "step": 1410
+    },
+    {
+      "epoch": 0.022576,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.853225806451614e-05,
+      "loss": 0.2274,
+      "step": 1411
+    },
+    {
+      "epoch": 0.022592,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.853064516129032e-05,
+      "loss": 0.1802,
+      "step": 1412
+    },
+    {
+      "epoch": 0.022608,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.852903225806452e-05,
+      "loss": 0.1987,
+      "step": 1413
+    },
+    {
+      "epoch": 0.022624,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.852741935483871e-05,
+      "loss": 0.2385,
+      "step": 1414
+    },
+    {
+      "epoch": 0.02264,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.852580645161291e-05,
+      "loss": 0.2018,
+      "step": 1415
+    },
+    {
+      "epoch": 0.022656,
+      "grad_norm": 1.5,
+      "learning_rate": 9.85241935483871e-05,
+      "loss": 0.2589,
+      "step": 1416
+    },
+    {
+      "epoch": 0.022672,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.85225806451613e-05,
+      "loss": 0.2131,
+      "step": 1417
+    },
+    {
+      "epoch": 0.022688,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.852096774193548e-05,
+      "loss": 0.2475,
+      "step": 1418
+    },
+    {
+      "epoch": 0.022704,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.851935483870968e-05,
+      "loss": 0.2284,
+      "step": 1419
+    },
+    {
+      "epoch": 0.02272,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.851774193548388e-05,
+      "loss": 0.201,
+      "step": 1420
+    },
+    {
+      "epoch": 0.022736,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.851612903225807e-05,
+      "loss": 0.2425,
+      "step": 1421
+    },
+    {
+      "epoch": 0.022752,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.851451612903227e-05,
+      "loss": 0.22,
+      "step": 1422
+    },
+    {
+      "epoch": 0.022768,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.851290322580645e-05,
+      "loss": 0.2073,
+      "step": 1423
+    },
+    {
+      "epoch": 0.022784,
+      "grad_norm": 1.4296875,
+      "learning_rate": 9.851129032258065e-05,
+      "loss": 0.1931,
+      "step": 1424
+    },
+    {
+      "epoch": 0.0228,
+      "grad_norm": 1.3359375,
+      "learning_rate": 9.850967741935484e-05,
+      "loss": 0.2086,
+      "step": 1425
+    },
+    {
+      "epoch": 0.022816,
+      "grad_norm": 1.40625,
+      "learning_rate": 9.850806451612904e-05,
+      "loss": 0.211,
+      "step": 1426
+    },
+    {
+      "epoch": 0.022832,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.850645161290322e-05,
+      "loss": 0.1723,
+      "step": 1427
+    },
+    {
+      "epoch": 0.022848,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.850483870967742e-05,
+      "loss": 0.1533,
+      "step": 1428
+    },
+    {
+      "epoch": 0.022864,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.850322580645161e-05,
+      "loss": 0.1922,
+      "step": 1429
+    },
+    {
+      "epoch": 0.02288,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.850161290322581e-05,
+      "loss": 0.2259,
+      "step": 1430
+    },
+    {
+      "epoch": 0.022896,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.850000000000001e-05,
+      "loss": 0.1602,
+      "step": 1431
+    },
+    {
+      "epoch": 0.022912,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.849838709677421e-05,
+      "loss": 0.2284,
+      "step": 1432
+    },
+    {
+      "epoch": 0.022928,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.84967741935484e-05,
+      "loss": 0.2414,
+      "step": 1433
+    },
+    {
+      "epoch": 0.022944,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.84951612903226e-05,
+      "loss": 0.2138,
+      "step": 1434
+    },
+    {
+      "epoch": 0.02296,
+      "grad_norm": 1.34375,
+      "learning_rate": 9.849354838709678e-05,
+      "loss": 0.2305,
+      "step": 1435
+    },
+    {
+      "epoch": 0.022976,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.849193548387097e-05,
+      "loss": 0.1618,
+      "step": 1436
+    },
+    {
+      "epoch": 0.022992,
+      "grad_norm": 1.84375,
+      "learning_rate": 9.849032258064517e-05,
+      "loss": 0.2218,
+      "step": 1437
+    },
+    {
+      "epoch": 0.023008,
+      "grad_norm": 1.0,
+      "learning_rate": 9.848870967741935e-05,
+      "loss": 0.2086,
+      "step": 1438
+    },
+    {
+      "epoch": 0.023024,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.848709677419355e-05,
+      "loss": 0.2001,
+      "step": 1439
+    },
+    {
+      "epoch": 0.02304,
+      "grad_norm": 1.4140625,
+      "learning_rate": 9.848548387096774e-05,
+      "loss": 0.2285,
+      "step": 1440
+    },
+    {
+      "epoch": 0.023056,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.848387096774194e-05,
+      "loss": 0.2566,
+      "step": 1441
+    },
+    {
+      "epoch": 0.023072,
+      "grad_norm": 1.75,
+      "learning_rate": 9.848225806451614e-05,
+      "loss": 0.2286,
+      "step": 1442
+    },
+    {
+      "epoch": 0.023088,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.848064516129034e-05,
+      "loss": 0.2157,
+      "step": 1443
+    },
+    {
+      "epoch": 0.023104,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.847903225806452e-05,
+      "loss": 0.1822,
+      "step": 1444
+    },
+    {
+      "epoch": 0.02312,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.847741935483872e-05,
+      "loss": 0.1956,
+      "step": 1445
+    },
+    {
+      "epoch": 0.023136,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.847580645161291e-05,
+      "loss": 0.1766,
+      "step": 1446
+    },
+    {
+      "epoch": 0.023152,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.847419354838711e-05,
+      "loss": 0.1824,
+      "step": 1447
+    },
+    {
+      "epoch": 0.023168,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.84725806451613e-05,
+      "loss": 0.2189,
+      "step": 1448
+    },
+    {
+      "epoch": 0.023184,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.84709677419355e-05,
+      "loss": 0.1939,
+      "step": 1449
+    },
+    {
+      "epoch": 0.0232,
+      "grad_norm": 1.3984375,
+      "learning_rate": 9.846935483870968e-05,
+      "loss": 0.1788,
+      "step": 1450
+    },
+    {
+      "epoch": 0.023216,
+      "grad_norm": 1.0,
+      "learning_rate": 9.846774193548387e-05,
+      "loss": 0.2174,
+      "step": 1451
+    },
+    {
+      "epoch": 0.023232,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.846612903225807e-05,
+      "loss": 0.209,
+      "step": 1452
+    },
+    {
+      "epoch": 0.023248,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.846451612903225e-05,
+      "loss": 0.1543,
+      "step": 1453
+    },
+    {
+      "epoch": 0.023264,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.846290322580645e-05,
+      "loss": 0.2573,
+      "step": 1454
+    },
+    {
+      "epoch": 0.02328,
+      "grad_norm": 1.3125,
+      "learning_rate": 9.846129032258065e-05,
+      "loss": 0.1883,
+      "step": 1455
+    },
+    {
+      "epoch": 0.023296,
+      "grad_norm": 1.3828125,
+      "learning_rate": 9.845967741935485e-05,
+      "loss": 0.2028,
+      "step": 1456
+    },
+    {
+      "epoch": 0.023312,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.845806451612904e-05,
+      "loss": 0.1891,
+      "step": 1457
+    },
+    {
+      "epoch": 0.023328,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.845645161290324e-05,
+      "loss": 0.2167,
+      "step": 1458
+    },
+    {
+      "epoch": 0.023344,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.845483870967742e-05,
+      "loss": 0.2111,
+      "step": 1459
+    },
+    {
+      "epoch": 0.02336,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.845322580645162e-05,
+      "loss": 0.2093,
+      "step": 1460
+    },
+    {
+      "epoch": 0.023376,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.845161290322581e-05,
+      "loss": 0.159,
+      "step": 1461
+    },
+    {
+      "epoch": 0.023392,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.845000000000001e-05,
+      "loss": 0.1799,
+      "step": 1462
+    },
+    {
+      "epoch": 0.023408,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.84483870967742e-05,
+      "loss": 0.2259,
+      "step": 1463
+    },
+    {
+      "epoch": 0.023424,
+      "grad_norm": 1.5390625,
+      "learning_rate": 9.84467741935484e-05,
+      "loss": 0.1856,
+      "step": 1464
+    },
+    {
+      "epoch": 0.02344,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.844516129032258e-05,
+      "loss": 0.1861,
+      "step": 1465
+    },
+    {
+      "epoch": 0.023456,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.844354838709678e-05,
+      "loss": 0.1733,
+      "step": 1466
+    },
+    {
+      "epoch": 0.023472,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.844193548387098e-05,
+      "loss": 0.2427,
+      "step": 1467
+    },
+    {
+      "epoch": 0.023488,
+      "grad_norm": 1.3359375,
+      "learning_rate": 9.844032258064516e-05,
+      "loss": 0.2074,
+      "step": 1468
+    },
+    {
+      "epoch": 0.023504,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.843870967741936e-05,
+      "loss": 0.1367,
+      "step": 1469
+    },
+    {
+      "epoch": 0.02352,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.843709677419355e-05,
+      "loss": 0.2064,
+      "step": 1470
+    },
+    {
+      "epoch": 0.023536,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.843548387096775e-05,
+      "loss": 0.2231,
+      "step": 1471
+    },
+    {
+      "epoch": 0.023552,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.843387096774194e-05,
+      "loss": 0.1915,
+      "step": 1472
+    },
+    {
+      "epoch": 0.023568,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.843225806451614e-05,
+      "loss": 0.2051,
+      "step": 1473
+    },
+    {
+      "epoch": 0.023584,
+      "grad_norm": 1.296875,
+      "learning_rate": 9.843064516129032e-05,
+      "loss": 0.2322,
+      "step": 1474
+    },
+    {
+      "epoch": 0.0236,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.842903225806452e-05,
+      "loss": 0.1961,
+      "step": 1475
+    },
+    {
+      "epoch": 0.023616,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.842741935483871e-05,
+      "loss": 0.247,
+      "step": 1476
+    },
+    {
+      "epoch": 0.023632,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.842580645161291e-05,
+      "loss": 0.1914,
+      "step": 1477
+    },
+    {
+      "epoch": 0.023648,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.842419354838711e-05,
+      "loss": 0.1545,
+      "step": 1478
+    },
+    {
+      "epoch": 0.023664,
+      "grad_norm": 1.375,
+      "learning_rate": 9.842258064516129e-05,
+      "loss": 0.18,
+      "step": 1479
+    },
+    {
+      "epoch": 0.02368,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.842096774193549e-05,
+      "loss": 0.1917,
+      "step": 1480
+    },
+    {
+      "epoch": 0.023696,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.841935483870969e-05,
+      "loss": 0.188,
+      "step": 1481
+    },
+    {
+      "epoch": 0.023712,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.841774193548388e-05,
+      "loss": 0.2182,
+      "step": 1482
+    },
+    {
+      "epoch": 0.023728,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.841612903225806e-05,
+      "loss": 0.2059,
+      "step": 1483
+    },
+    {
+      "epoch": 0.023744,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.841451612903226e-05,
+      "loss": 0.2171,
+      "step": 1484
+    },
+    {
+      "epoch": 0.02376,
+      "grad_norm": 1.4375,
+      "learning_rate": 9.841290322580645e-05,
+      "loss": 0.2821,
+      "step": 1485
+    },
+    {
+      "epoch": 0.023776,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.841129032258065e-05,
+      "loss": 0.1813,
+      "step": 1486
+    },
+    {
+      "epoch": 0.023792,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.840967741935484e-05,
+      "loss": 0.2118,
+      "step": 1487
+    },
+    {
+      "epoch": 0.023808,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.840806451612904e-05,
+      "loss": 0.1805,
+      "step": 1488
+    },
+    {
+      "epoch": 0.023824,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.840645161290322e-05,
+      "loss": 0.1822,
+      "step": 1489
+    },
+    {
+      "epoch": 0.02384,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.840483870967742e-05,
+      "loss": 0.1432,
+      "step": 1490
+    },
+    {
+      "epoch": 0.023856,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.840322580645162e-05,
+      "loss": 0.2123,
+      "step": 1491
+    },
+    {
+      "epoch": 0.023872,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.840161290322582e-05,
+      "loss": 0.1769,
+      "step": 1492
+    },
+    {
+      "epoch": 0.023888,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.84e-05,
+      "loss": 0.2663,
+      "step": 1493
+    },
+    {
+      "epoch": 0.023904,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.83983870967742e-05,
+      "loss": 0.2297,
+      "step": 1494
+    },
+    {
+      "epoch": 0.02392,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.839677419354839e-05,
+      "loss": 0.2046,
+      "step": 1495
+    },
+    {
+      "epoch": 0.023936,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.839516129032259e-05,
+      "loss": 0.2329,
+      "step": 1496
+    },
+    {
+      "epoch": 0.023952,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.839354838709678e-05,
+      "loss": 0.2264,
+      "step": 1497
+    },
+    {
+      "epoch": 0.023968,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.839193548387096e-05,
+      "loss": 0.2092,
+      "step": 1498
+    },
+    {
+      "epoch": 0.023984,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.839032258064516e-05,
+      "loss": 0.1642,
+      "step": 1499
+    },
+    {
+      "epoch": 0.024,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.838870967741935e-05,
+      "loss": 0.1899,
+      "step": 1500
+    },
+    {
+      "epoch": 0.024016,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.838709677419355e-05,
+      "loss": 0.2023,
+      "step": 1501
+    },
+    {
+      "epoch": 0.024032,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.838548387096775e-05,
+      "loss": 0.2088,
+      "step": 1502
+    },
+    {
+      "epoch": 0.024048,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.838387096774195e-05,
+      "loss": 0.1867,
+      "step": 1503
+    },
+    {
+      "epoch": 0.024064,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.838225806451613e-05,
+      "loss": 0.2339,
+      "step": 1504
+    },
+    {
+      "epoch": 0.02408,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.838064516129033e-05,
+      "loss": 0.2068,
+      "step": 1505
+    },
+    {
+      "epoch": 0.024096,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.837903225806452e-05,
+      "loss": 0.192,
+      "step": 1506
+    },
+    {
+      "epoch": 0.024112,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.837741935483872e-05,
+      "loss": 0.1467,
+      "step": 1507
+    },
+    {
+      "epoch": 0.024128,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.83758064516129e-05,
+      "loss": 0.2131,
+      "step": 1508
+    },
+    {
+      "epoch": 0.024144,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.83741935483871e-05,
+      "loss": 0.146,
+      "step": 1509
+    },
+    {
+      "epoch": 0.02416,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.837258064516129e-05,
+      "loss": 0.151,
+      "step": 1510
+    },
+    {
+      "epoch": 0.024176,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.837096774193549e-05,
+      "loss": 0.2315,
+      "step": 1511
+    },
+    {
+      "epoch": 0.024192,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.836935483870968e-05,
+      "loss": 0.2117,
+      "step": 1512
+    },
+    {
+      "epoch": 0.024208,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.836774193548386e-05,
+      "loss": 0.2022,
+      "step": 1513
+    },
+    {
+      "epoch": 0.024224,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.836612903225806e-05,
+      "loss": 0.169,
+      "step": 1514
+    },
+    {
+      "epoch": 0.02424,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.836451612903226e-05,
+      "loss": 0.2377,
+      "step": 1515
+    },
+    {
+      "epoch": 0.024256,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.836290322580646e-05,
+      "loss": 0.214,
+      "step": 1516
+    },
+    {
+      "epoch": 0.024272,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.836129032258065e-05,
+      "loss": 0.1934,
+      "step": 1517
+    },
+    {
+      "epoch": 0.024288,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.835967741935485e-05,
+      "loss": 0.1949,
+      "step": 1518
+    },
+    {
+      "epoch": 0.024304,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.835806451612903e-05,
+      "loss": 0.1909,
+      "step": 1519
+    },
+    {
+      "epoch": 0.02432,
+      "grad_norm": 1.3125,
+      "learning_rate": 9.835645161290323e-05,
+      "loss": 0.2023,
+      "step": 1520
+    },
+    {
+      "epoch": 0.024336,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.835483870967742e-05,
+      "loss": 0.172,
+      "step": 1521
+    },
+    {
+      "epoch": 0.024352,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.835322580645162e-05,
+      "loss": 0.2664,
+      "step": 1522
+    },
+    {
+      "epoch": 0.024368,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.83516129032258e-05,
+      "loss": 0.2331,
+      "step": 1523
+    },
+    {
+      "epoch": 0.024384,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.835e-05,
+      "loss": 0.1848,
+      "step": 1524
+    },
+    {
+      "epoch": 0.0244,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.834838709677419e-05,
+      "loss": 0.1807,
+      "step": 1525
+    },
+    {
+      "epoch": 0.024416,
+      "grad_norm": 1.4921875,
+      "learning_rate": 9.834677419354839e-05,
+      "loss": 0.2056,
+      "step": 1526
+    },
+    {
+      "epoch": 0.024432,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.834516129032259e-05,
+      "loss": 0.1935,
+      "step": 1527
+    },
+    {
+      "epoch": 0.024448,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.834354838709679e-05,
+      "loss": 0.2479,
+      "step": 1528
+    },
+    {
+      "epoch": 0.024464,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.834193548387098e-05,
+      "loss": 0.2438,
+      "step": 1529
+    },
+    {
+      "epoch": 0.02448,
+      "grad_norm": 1.40625,
+      "learning_rate": 9.834032258064516e-05,
+      "loss": 0.226,
+      "step": 1530
+    },
+    {
+      "epoch": 0.024496,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.833870967741936e-05,
+      "loss": 0.1974,
+      "step": 1531
+    },
+    {
+      "epoch": 0.024512,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.833709677419355e-05,
+      "loss": 0.1752,
+      "step": 1532
+    },
+    {
+      "epoch": 0.024528,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.833548387096775e-05,
+      "loss": 0.2073,
+      "step": 1533
+    },
+    {
+      "epoch": 0.024544,
+      "grad_norm": 1.40625,
+      "learning_rate": 9.833387096774193e-05,
+      "loss": 0.2098,
+      "step": 1534
+    },
+    {
+      "epoch": 0.02456,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.833225806451613e-05,
+      "loss": 0.176,
+      "step": 1535
+    },
+    {
+      "epoch": 0.024576,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.833064516129032e-05,
+      "loss": 0.1937,
+      "step": 1536
+    },
+    {
+      "epoch": 0.024592,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.832903225806452e-05,
+      "loss": 0.2549,
+      "step": 1537
+    },
+    {
+      "epoch": 0.024608,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.832741935483872e-05,
+      "loss": 0.2285,
+      "step": 1538
+    },
+    {
+      "epoch": 0.024624,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.832580645161292e-05,
+      "loss": 0.1968,
+      "step": 1539
+    },
+    {
+      "epoch": 0.02464,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.83241935483871e-05,
+      "loss": 0.2239,
+      "step": 1540
+    },
+    {
+      "epoch": 0.024656,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.83225806451613e-05,
+      "loss": 0.2023,
+      "step": 1541
+    },
+    {
+      "epoch": 0.024672,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.832096774193549e-05,
+      "loss": 0.229,
+      "step": 1542
+    },
+    {
+      "epoch": 0.024688,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.831935483870969e-05,
+      "loss": 0.1827,
+      "step": 1543
+    },
+    {
+      "epoch": 0.024704,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.831774193548388e-05,
+      "loss": 0.1608,
+      "step": 1544
+    },
+    {
+      "epoch": 0.02472,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.831612903225806e-05,
+      "loss": 0.1986,
+      "step": 1545
+    },
+    {
+      "epoch": 0.024736,
+      "grad_norm": 1.5078125,
+      "learning_rate": 9.831451612903226e-05,
+      "loss": 0.1639,
+      "step": 1546
+    },
+    {
+      "epoch": 0.024752,
+      "grad_norm": 1.4609375,
+      "learning_rate": 9.831290322580645e-05,
+      "loss": 0.2293,
+      "step": 1547
+    },
+    {
+      "epoch": 0.024768,
+      "grad_norm": 1.765625,
+      "learning_rate": 9.831129032258065e-05,
+      "loss": 0.2052,
+      "step": 1548
+    },
+    {
+      "epoch": 0.024784,
+      "grad_norm": 1.3359375,
+      "learning_rate": 9.830967741935483e-05,
+      "loss": 0.2525,
+      "step": 1549
+    },
+    {
+      "epoch": 0.0248,
+      "grad_norm": 1.7734375,
+      "learning_rate": 9.830806451612903e-05,
+      "loss": 0.2397,
+      "step": 1550
+    },
+    {
+      "epoch": 0.024816,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.830645161290323e-05,
+      "loss": 0.1924,
+      "step": 1551
+    },
+    {
+      "epoch": 0.024832,
+      "grad_norm": 1.5234375,
+      "learning_rate": 9.830483870967743e-05,
+      "loss": 0.1936,
+      "step": 1552
+    },
+    {
+      "epoch": 0.024848,
+      "grad_norm": 0.875,
+      "learning_rate": 9.830322580645162e-05,
+      "loss": 0.1843,
+      "step": 1553
+    },
+    {
+      "epoch": 0.024864,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.830161290322582e-05,
+      "loss": 0.1741,
+      "step": 1554
+    },
+    {
+      "epoch": 0.02488,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.83e-05,
+      "loss": 0.1968,
+      "step": 1555
+    },
+    {
+      "epoch": 0.024896,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.82983870967742e-05,
+      "loss": 0.1891,
+      "step": 1556
+    },
+    {
+      "epoch": 0.024912,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.829677419354839e-05,
+      "loss": 0.2462,
+      "step": 1557
+    },
+    {
+      "epoch": 0.024928,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.829516129032259e-05,
+      "loss": 0.2187,
+      "step": 1558
+    },
+    {
+      "epoch": 0.024944,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.829354838709678e-05,
+      "loss": 0.1892,
+      "step": 1559
+    },
+    {
+      "epoch": 0.02496,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.829193548387096e-05,
+      "loss": 0.1972,
+      "step": 1560
+    },
+    {
+      "epoch": 0.024976,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.829032258064516e-05,
+      "loss": 0.1544,
+      "step": 1561
+    },
+    {
+      "epoch": 0.024992,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.828870967741936e-05,
+      "loss": 0.1786,
+      "step": 1562
+    },
+    {
+      "epoch": 0.025008,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.828709677419356e-05,
+      "loss": 0.2252,
+      "step": 1563
+    },
+    {
+      "epoch": 0.025024,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.828548387096775e-05,
+      "loss": 0.1739,
+      "step": 1564
+    },
+    {
+      "epoch": 0.02504,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.828387096774195e-05,
+      "loss": 0.186,
+      "step": 1565
+    },
+    {
+      "epoch": 0.025056,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.828225806451613e-05,
+      "loss": 0.1693,
+      "step": 1566
+    },
+    {
+      "epoch": 0.025072,
+      "grad_norm": 1.453125,
+      "learning_rate": 9.828064516129033e-05,
+      "loss": 0.2222,
+      "step": 1567
+    },
+    {
+      "epoch": 0.025088,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.827903225806452e-05,
+      "loss": 0.1998,
+      "step": 1568
+    },
+    {
+      "epoch": 0.025104,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.827741935483872e-05,
+      "loss": 0.1812,
+      "step": 1569
+    },
+    {
+      "epoch": 0.02512,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.82758064516129e-05,
+      "loss": 0.201,
+      "step": 1570
+    },
+    {
+      "epoch": 0.025136,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.82741935483871e-05,
+      "loss": 0.2096,
+      "step": 1571
+    },
+    {
+      "epoch": 0.025152,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.827258064516129e-05,
+      "loss": 0.1821,
+      "step": 1572
+    },
+    {
+      "epoch": 0.025168,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.827096774193549e-05,
+      "loss": 0.1588,
+      "step": 1573
+    },
+    {
+      "epoch": 0.025184,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.826935483870968e-05,
+      "loss": 0.2086,
+      "step": 1574
+    },
+    {
+      "epoch": 0.0252,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.826774193548387e-05,
+      "loss": 0.1971,
+      "step": 1575
+    },
+    {
+      "epoch": 0.025216,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.826612903225807e-05,
+      "loss": 0.2057,
+      "step": 1576
+    },
+    {
+      "epoch": 0.025232,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.826451612903226e-05,
+      "loss": 0.208,
+      "step": 1577
+    },
+    {
+      "epoch": 0.025248,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.826290322580646e-05,
+      "loss": 0.2286,
+      "step": 1578
+    },
+    {
+      "epoch": 0.025264,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.826129032258065e-05,
+      "loss": 0.2277,
+      "step": 1579
+    },
+    {
+      "epoch": 0.02528,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.825967741935485e-05,
+      "loss": 0.2141,
+      "step": 1580
+    },
+    {
+      "epoch": 0.025296,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.825806451612903e-05,
+      "loss": 0.1927,
+      "step": 1581
+    },
+    {
+      "epoch": 0.025312,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.825645161290323e-05,
+      "loss": 0.1873,
+      "step": 1582
+    },
+    {
+      "epoch": 0.025328,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.825483870967742e-05,
+      "loss": 0.2075,
+      "step": 1583
+    },
+    {
+      "epoch": 0.025344,
+      "grad_norm": 1.3515625,
+      "learning_rate": 9.825322580645162e-05,
+      "loss": 0.2004,
+      "step": 1584
+    },
+    {
+      "epoch": 0.02536,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.82516129032258e-05,
+      "loss": 0.1994,
+      "step": 1585
+    },
+    {
+      "epoch": 0.025376,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.825e-05,
+      "loss": 0.2373,
+      "step": 1586
+    },
+    {
+      "epoch": 0.025392,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.82483870967742e-05,
+      "loss": 0.1983,
+      "step": 1587
+    },
+    {
+      "epoch": 0.025408,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.82467741935484e-05,
+      "loss": 0.2305,
+      "step": 1588
+    },
+    {
+      "epoch": 0.025424,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.824516129032259e-05,
+      "loss": 0.2187,
+      "step": 1589
+    },
+    {
+      "epoch": 0.02544,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.824354838709679e-05,
+      "loss": 0.1961,
+      "step": 1590
+    },
+    {
+      "epoch": 0.025456,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.824193548387097e-05,
+      "loss": 0.1776,
+      "step": 1591
+    },
+    {
+      "epoch": 0.025472,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.824032258064516e-05,
+      "loss": 0.2602,
+      "step": 1592
+    },
+    {
+      "epoch": 0.025488,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.823870967741936e-05,
+      "loss": 0.2294,
+      "step": 1593
+    },
+    {
+      "epoch": 0.025504,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.823709677419355e-05,
+      "loss": 0.1752,
+      "step": 1594
+    },
+    {
+      "epoch": 0.02552,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.823548387096775e-05,
+      "loss": 0.203,
+      "step": 1595
+    },
+    {
+      "epoch": 0.025536,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.823387096774193e-05,
+      "loss": 0.1865,
+      "step": 1596
+    },
+    {
+      "epoch": 0.025552,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.823225806451613e-05,
+      "loss": 0.1838,
+      "step": 1597
+    },
+    {
+      "epoch": 0.025568,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.823064516129033e-05,
+      "loss": 0.1768,
+      "step": 1598
+    },
+    {
+      "epoch": 0.025584,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.822903225806453e-05,
+      "loss": 0.191,
+      "step": 1599
+    },
+    {
+      "epoch": 0.0256,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.822741935483872e-05,
+      "loss": 0.2118,
+      "step": 1600
+    },
+    {
+      "epoch": 0.025616,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.822580645161292e-05,
+      "loss": 0.2252,
+      "step": 1601
+    },
+    {
+      "epoch": 0.025632,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.82241935483871e-05,
+      "loss": 0.1469,
+      "step": 1602
+    },
+    {
+      "epoch": 0.025648,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.82225806451613e-05,
+      "loss": 0.224,
+      "step": 1603
+    },
+    {
+      "epoch": 0.025664,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.822096774193549e-05,
+      "loss": 0.2931,
+      "step": 1604
+    },
+    {
+      "epoch": 0.02568,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.821935483870969e-05,
+      "loss": 0.2206,
+      "step": 1605
+    },
+    {
+      "epoch": 0.025696,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.821774193548387e-05,
+      "loss": 0.1928,
+      "step": 1606
+    },
+    {
+      "epoch": 0.025712,
+      "grad_norm": 1.4921875,
+      "learning_rate": 9.821612903225806e-05,
+      "loss": 0.215,
+      "step": 1607
+    },
+    {
+      "epoch": 0.025728,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.821451612903226e-05,
+      "loss": 0.2043,
+      "step": 1608
+    },
+    {
+      "epoch": 0.025744,
+      "grad_norm": 0.75,
+      "learning_rate": 9.821290322580645e-05,
+      "loss": 0.1405,
+      "step": 1609
+    },
+    {
+      "epoch": 0.02576,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.821129032258064e-05,
+      "loss": 0.217,
+      "step": 1610
+    },
+    {
+      "epoch": 0.025776,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.820967741935484e-05,
+      "loss": 0.1816,
+      "step": 1611
+    },
+    {
+      "epoch": 0.025792,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.820806451612904e-05,
+      "loss": 0.2084,
+      "step": 1612
+    },
+    {
+      "epoch": 0.025808,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.820645161290323e-05,
+      "loss": 0.2183,
+      "step": 1613
+    },
+    {
+      "epoch": 0.025824,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.820483870967743e-05,
+      "loss": 0.1889,
+      "step": 1614
+    },
+    {
+      "epoch": 0.02584,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.820322580645162e-05,
+      "loss": 0.1949,
+      "step": 1615
+    },
+    {
+      "epoch": 0.025856,
+      "grad_norm": 1.828125,
+      "learning_rate": 9.820161290322582e-05,
+      "loss": 0.2486,
+      "step": 1616
+    },
+    {
+      "epoch": 0.025872,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.82e-05,
+      "loss": 0.1835,
+      "step": 1617
+    },
+    {
+      "epoch": 0.025888,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.81983870967742e-05,
+      "loss": 0.225,
+      "step": 1618
+    },
+    {
+      "epoch": 0.025904,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.819677419354839e-05,
+      "loss": 0.2105,
+      "step": 1619
+    },
+    {
+      "epoch": 0.02592,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.819516129032259e-05,
+      "loss": 0.2211,
+      "step": 1620
+    },
+    {
+      "epoch": 0.025936,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.819354838709677e-05,
+      "loss": 0.2004,
+      "step": 1621
+    },
+    {
+      "epoch": 0.025952,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.819193548387097e-05,
+      "loss": 0.1838,
+      "step": 1622
+    },
+    {
+      "epoch": 0.025968,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.819032258064517e-05,
+      "loss": 0.1788,
+      "step": 1623
+    },
+    {
+      "epoch": 0.025984,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.818870967741936e-05,
+      "loss": 0.2066,
+      "step": 1624
+    },
+    {
+      "epoch": 0.026,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.818709677419356e-05,
+      "loss": 0.1999,
+      "step": 1625
+    },
+    {
+      "epoch": 0.026016,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.818548387096774e-05,
+      "loss": 0.2007,
+      "step": 1626
+    },
+    {
+      "epoch": 0.026032,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.818387096774194e-05,
+      "loss": 0.2021,
+      "step": 1627
+    },
+    {
+      "epoch": 0.026048,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.818225806451613e-05,
+      "loss": 0.2059,
+      "step": 1628
+    },
+    {
+      "epoch": 0.026064,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.818064516129033e-05,
+      "loss": 0.2242,
+      "step": 1629
+    },
+    {
+      "epoch": 0.02608,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.817903225806452e-05,
+      "loss": 0.178,
+      "step": 1630
+    },
+    {
+      "epoch": 0.026096,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.817741935483872e-05,
+      "loss": 0.2307,
+      "step": 1631
+    },
+    {
+      "epoch": 0.026112,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.81758064516129e-05,
+      "loss": 0.2026,
+      "step": 1632
+    },
+    {
+      "epoch": 0.026128,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.81741935483871e-05,
+      "loss": 0.2223,
+      "step": 1633
+    },
+    {
+      "epoch": 0.026144,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.81725806451613e-05,
+      "loss": 0.2659,
+      "step": 1634
+    },
+    {
+      "epoch": 0.02616,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.81709677419355e-05,
+      "loss": 0.2013,
+      "step": 1635
+    },
+    {
+      "epoch": 0.026176,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.816935483870969e-05,
+      "loss": 0.2141,
+      "step": 1636
+    },
+    {
+      "epoch": 0.026192,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.816774193548389e-05,
+      "loss": 0.2156,
+      "step": 1637
+    },
+    {
+      "epoch": 0.026208,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.816612903225807e-05,
+      "loss": 0.1857,
+      "step": 1638
+    },
+    {
+      "epoch": 0.026224,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.816451612903226e-05,
+      "loss": 0.19,
+      "step": 1639
+    },
+    {
+      "epoch": 0.02624,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.816290322580646e-05,
+      "loss": 0.1949,
+      "step": 1640
+    },
+    {
+      "epoch": 0.026256,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.816129032258064e-05,
+      "loss": 0.1849,
+      "step": 1641
+    },
+    {
+      "epoch": 0.026272,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.815967741935484e-05,
+      "loss": 0.2305,
+      "step": 1642
+    },
+    {
+      "epoch": 0.026288,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.815806451612903e-05,
+      "loss": 0.2052,
+      "step": 1643
+    },
+    {
+      "epoch": 0.026304,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.815645161290323e-05,
+      "loss": 0.2207,
+      "step": 1644
+    },
+    {
+      "epoch": 0.02632,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.815483870967742e-05,
+      "loss": 0.2172,
+      "step": 1645
+    },
+    {
+      "epoch": 0.026336,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.815322580645161e-05,
+      "loss": 0.2041,
+      "step": 1646
+    },
+    {
+      "epoch": 0.026352,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.815161290322581e-05,
+      "loss": 0.1637,
+      "step": 1647
+    },
+    {
+      "epoch": 0.026368,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.815000000000001e-05,
+      "loss": 0.1426,
+      "step": 1648
+    },
+    {
+      "epoch": 0.026384,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.81483870967742e-05,
+      "loss": 0.2069,
+      "step": 1649
+    },
+    {
+      "epoch": 0.0264,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.81467741935484e-05,
+      "loss": 0.2305,
+      "step": 1650
+    },
+    {
+      "epoch": 0.026416,
+      "grad_norm": 1.34375,
+      "learning_rate": 9.814516129032259e-05,
+      "loss": 0.1939,
+      "step": 1651
+    },
+    {
+      "epoch": 0.026432,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.814354838709679e-05,
+      "loss": 0.1793,
+      "step": 1652
+    },
+    {
+      "epoch": 0.026448,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.814193548387097e-05,
+      "loss": 0.2436,
+      "step": 1653
+    },
+    {
+      "epoch": 0.026464,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.814032258064516e-05,
+      "loss": 0.2075,
+      "step": 1654
+    },
+    {
+      "epoch": 0.02648,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.813870967741936e-05,
+      "loss": 0.2291,
+      "step": 1655
+    },
+    {
+      "epoch": 0.026496,
+      "grad_norm": 1.0,
+      "learning_rate": 9.813709677419354e-05,
+      "loss": 0.1972,
+      "step": 1656
+    },
+    {
+      "epoch": 0.026512,
+      "grad_norm": 1.125,
+      "learning_rate": 9.813548387096774e-05,
+      "loss": 0.2182,
+      "step": 1657
+    },
+    {
+      "epoch": 0.026528,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.813387096774194e-05,
+      "loss": 0.229,
+      "step": 1658
+    },
+    {
+      "epoch": 0.026544,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.813225806451614e-05,
+      "loss": 0.1912,
+      "step": 1659
+    },
+    {
+      "epoch": 0.02656,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.813064516129033e-05,
+      "loss": 0.2155,
+      "step": 1660
+    },
+    {
+      "epoch": 0.026576,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.812903225806453e-05,
+      "loss": 0.2151,
+      "step": 1661
+    },
+    {
+      "epoch": 0.026592,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.812741935483871e-05,
+      "loss": 0.2028,
+      "step": 1662
+    },
+    {
+      "epoch": 0.026608,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.812580645161291e-05,
+      "loss": 0.2139,
+      "step": 1663
+    },
+    {
+      "epoch": 0.026624,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.81241935483871e-05,
+      "loss": 0.1622,
+      "step": 1664
+    },
+    {
+      "epoch": 0.02664,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.81225806451613e-05,
+      "loss": 0.2161,
+      "step": 1665
+    },
+    {
+      "epoch": 0.026656,
+      "grad_norm": 1.46875,
+      "learning_rate": 9.812096774193549e-05,
+      "loss": 0.2086,
+      "step": 1666
+    },
+    {
+      "epoch": 0.026672,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.811935483870969e-05,
+      "loss": 0.1411,
+      "step": 1667
+    },
+    {
+      "epoch": 0.026688,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.811774193548387e-05,
+      "loss": 0.2012,
+      "step": 1668
+    },
+    {
+      "epoch": 0.026704,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.811612903225807e-05,
+      "loss": 0.2619,
+      "step": 1669
+    },
+    {
+      "epoch": 0.02672,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.811451612903226e-05,
+      "loss": 0.2051,
+      "step": 1670
+    },
+    {
+      "epoch": 0.026736,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.811290322580646e-05,
+      "loss": 0.1595,
+      "step": 1671
+    },
+    {
+      "epoch": 0.026752,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.811129032258066e-05,
+      "loss": 0.2004,
+      "step": 1672
+    },
+    {
+      "epoch": 0.026768,
+      "grad_norm": 1.0,
+      "learning_rate": 9.810967741935484e-05,
+      "loss": 0.2232,
+      "step": 1673
+    },
+    {
+      "epoch": 0.026784,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.810806451612904e-05,
+      "loss": 0.1634,
+      "step": 1674
+    },
+    {
+      "epoch": 0.0268,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.810645161290323e-05,
+      "loss": 0.2185,
+      "step": 1675
+    },
+    {
+      "epoch": 0.026816,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.810483870967743e-05,
+      "loss": 0.2003,
+      "step": 1676
+    },
+    {
+      "epoch": 0.026832,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.810322580645161e-05,
+      "loss": 0.1844,
+      "step": 1677
+    },
+    {
+      "epoch": 0.026848,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.810161290322581e-05,
+      "loss": 0.2184,
+      "step": 1678
+    },
+    {
+      "epoch": 0.026864,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.81e-05,
+      "loss": 0.2186,
+      "step": 1679
+    },
+    {
+      "epoch": 0.02688,
+      "grad_norm": 1.4921875,
+      "learning_rate": 9.80983870967742e-05,
+      "loss": 0.2411,
+      "step": 1680
+    },
+    {
+      "epoch": 0.026896,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.809677419354838e-05,
+      "loss": 0.2203,
+      "step": 1681
+    },
+    {
+      "epoch": 0.026912,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.809516129032258e-05,
+      "loss": 0.2144,
+      "step": 1682
+    },
+    {
+      "epoch": 0.026928,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.809354838709678e-05,
+      "loss": 0.1642,
+      "step": 1683
+    },
+    {
+      "epoch": 0.026944,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.809193548387097e-05,
+      "loss": 0.1732,
+      "step": 1684
+    },
+    {
+      "epoch": 0.02696,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.809032258064517e-05,
+      "loss": 0.2217,
+      "step": 1685
+    },
+    {
+      "epoch": 0.026976,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.808870967741936e-05,
+      "loss": 0.2109,
+      "step": 1686
+    },
+    {
+      "epoch": 0.026992,
+      "grad_norm": 1.125,
+      "learning_rate": 9.808709677419356e-05,
+      "loss": 0.2307,
+      "step": 1687
+    },
+    {
+      "epoch": 0.027008,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.808548387096774e-05,
+      "loss": 0.2064,
+      "step": 1688
+    },
+    {
+      "epoch": 0.027024,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.808387096774194e-05,
+      "loss": 0.2248,
+      "step": 1689
+    },
+    {
+      "epoch": 0.02704,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.808225806451613e-05,
+      "loss": 0.2243,
+      "step": 1690
+    },
+    {
+      "epoch": 0.027056,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.808064516129033e-05,
+      "loss": 0.1877,
+      "step": 1691
+    },
+    {
+      "epoch": 0.027072,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.807903225806451e-05,
+      "loss": 0.2001,
+      "step": 1692
+    },
+    {
+      "epoch": 0.027088,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.807741935483871e-05,
+      "loss": 0.2487,
+      "step": 1693
+    },
+    {
+      "epoch": 0.027104,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.807580645161291e-05,
+      "loss": 0.1836,
+      "step": 1694
+    },
+    {
+      "epoch": 0.02712,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.807419354838711e-05,
+      "loss": 0.2171,
+      "step": 1695
+    },
+    {
+      "epoch": 0.027136,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.80725806451613e-05,
+      "loss": 0.2127,
+      "step": 1696
+    },
+    {
+      "epoch": 0.027152,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.80709677419355e-05,
+      "loss": 0.1924,
+      "step": 1697
+    },
+    {
+      "epoch": 0.027168,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.806935483870968e-05,
+      "loss": 0.2063,
+      "step": 1698
+    },
+    {
+      "epoch": 0.027184,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.806774193548388e-05,
+      "loss": 0.2352,
+      "step": 1699
+    },
+    {
+      "epoch": 0.0272,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.806612903225807e-05,
+      "loss": 0.24,
+      "step": 1700
+    },
+    {
+      "epoch": 0.027216,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.806451612903226e-05,
+      "loss": 0.2083,
+      "step": 1701
+    },
+    {
+      "epoch": 0.027232,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.806290322580646e-05,
+      "loss": 0.2088,
+      "step": 1702
+    },
+    {
+      "epoch": 0.027248,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.806129032258064e-05,
+      "loss": 0.1798,
+      "step": 1703
+    },
+    {
+      "epoch": 0.027264,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.805967741935484e-05,
+      "loss": 0.2349,
+      "step": 1704
+    },
+    {
+      "epoch": 0.02728,
+      "grad_norm": 1.25,
+      "learning_rate": 9.805806451612903e-05,
+      "loss": 0.2027,
+      "step": 1705
+    },
+    {
+      "epoch": 0.027296,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.805645161290323e-05,
+      "loss": 0.2283,
+      "step": 1706
+    },
+    {
+      "epoch": 0.027312,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.805483870967743e-05,
+      "loss": 0.1683,
+      "step": 1707
+    },
+    {
+      "epoch": 0.027328,
+      "grad_norm": 1.296875,
+      "learning_rate": 9.805322580645163e-05,
+      "loss": 0.1903,
+      "step": 1708
+    },
+    {
+      "epoch": 0.027344,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.805161290322581e-05,
+      "loss": 0.189,
+      "step": 1709
+    },
+    {
+      "epoch": 0.02736,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.805000000000001e-05,
+      "loss": 0.2322,
+      "step": 1710
+    },
+    {
+      "epoch": 0.027376,
+      "grad_norm": 1.796875,
+      "learning_rate": 9.80483870967742e-05,
+      "loss": 0.2027,
+      "step": 1711
+    },
+    {
+      "epoch": 0.027392,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.80467741935484e-05,
+      "loss": 0.2182,
+      "step": 1712
+    },
+    {
+      "epoch": 0.027408,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.804516129032258e-05,
+      "loss": 0.1903,
+      "step": 1713
+    },
+    {
+      "epoch": 0.027424,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.804354838709678e-05,
+      "loss": 0.2386,
+      "step": 1714
+    },
+    {
+      "epoch": 0.02744,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.804193548387097e-05,
+      "loss": 0.2061,
+      "step": 1715
+    },
+    {
+      "epoch": 0.027456,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.804032258064516e-05,
+      "loss": 0.1928,
+      "step": 1716
+    },
+    {
+      "epoch": 0.027472,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.803870967741935e-05,
+      "loss": 0.2163,
+      "step": 1717
+    },
+    {
+      "epoch": 0.027488,
+      "grad_norm": 1.0,
+      "learning_rate": 9.803709677419355e-05,
+      "loss": 0.1952,
+      "step": 1718
+    },
+    {
+      "epoch": 0.027504,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.803548387096775e-05,
+      "loss": 0.1784,
+      "step": 1719
+    },
+    {
+      "epoch": 0.02752,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.803387096774194e-05,
+      "loss": 0.1993,
+      "step": 1720
+    },
+    {
+      "epoch": 0.027536,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.803225806451614e-05,
+      "loss": 0.1999,
+      "step": 1721
+    },
+    {
+      "epoch": 0.027552,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.803064516129033e-05,
+      "loss": 0.2403,
+      "step": 1722
+    },
+    {
+      "epoch": 0.027568,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.802903225806453e-05,
+      "loss": 0.2293,
+      "step": 1723
+    },
+    {
+      "epoch": 0.027584,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.802741935483871e-05,
+      "loss": 0.1799,
+      "step": 1724
+    },
+    {
+      "epoch": 0.0276,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.802580645161291e-05,
+      "loss": 0.1588,
+      "step": 1725
+    },
+    {
+      "epoch": 0.027616,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.80241935483871e-05,
+      "loss": 0.1818,
+      "step": 1726
+    },
+    {
+      "epoch": 0.027632,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.80225806451613e-05,
+      "loss": 0.2304,
+      "step": 1727
+    },
+    {
+      "epoch": 0.027648,
+      "grad_norm": 1.5390625,
+      "learning_rate": 9.802096774193548e-05,
+      "loss": 0.1952,
+      "step": 1728
+    },
+    {
+      "epoch": 0.027664,
+      "grad_norm": 1.359375,
+      "learning_rate": 9.801935483870968e-05,
+      "loss": 0.2407,
+      "step": 1729
+    },
+    {
+      "epoch": 0.02768,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.801774193548388e-05,
+      "loss": 0.2116,
+      "step": 1730
+    },
+    {
+      "epoch": 0.027696,
+      "grad_norm": 0.875,
+      "learning_rate": 9.801612903225807e-05,
+      "loss": 0.1896,
+      "step": 1731
+    },
+    {
+      "epoch": 0.027712,
+      "grad_norm": 1.3984375,
+      "learning_rate": 9.801451612903227e-05,
+      "loss": 0.2529,
+      "step": 1732
+    },
+    {
+      "epoch": 0.027728,
+      "grad_norm": 0.75,
+      "learning_rate": 9.801290322580645e-05,
+      "loss": 0.1994,
+      "step": 1733
+    },
+    {
+      "epoch": 0.027744,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.801129032258065e-05,
+      "loss": 0.1942,
+      "step": 1734
+    },
+    {
+      "epoch": 0.02776,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.800967741935484e-05,
+      "loss": 0.2087,
+      "step": 1735
+    },
+    {
+      "epoch": 0.027776,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.800806451612904e-05,
+      "loss": 0.1848,
+      "step": 1736
+    },
+    {
+      "epoch": 0.027792,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.800645161290323e-05,
+      "loss": 0.1522,
+      "step": 1737
+    },
+    {
+      "epoch": 0.027808,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.800483870967743e-05,
+      "loss": 0.1989,
+      "step": 1738
+    },
+    {
+      "epoch": 0.027824,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.800322580645161e-05,
+      "loss": 0.2521,
+      "step": 1739
+    },
+    {
+      "epoch": 0.02784,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.800161290322581e-05,
+      "loss": 0.2283,
+      "step": 1740
+    },
+    {
+      "epoch": 0.027856,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.8e-05,
+      "loss": 0.1802,
+      "step": 1741
+    },
+    {
+      "epoch": 0.027872,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.79983870967742e-05,
+      "loss": 0.2152,
+      "step": 1742
+    },
+    {
+      "epoch": 0.027888,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.79967741935484e-05,
+      "loss": 0.2264,
+      "step": 1743
+    },
+    {
+      "epoch": 0.027904,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.79951612903226e-05,
+      "loss": 0.1812,
+      "step": 1744
+    },
+    {
+      "epoch": 0.02792,
+      "grad_norm": 1.328125,
+      "learning_rate": 9.799354838709678e-05,
+      "loss": 0.1807,
+      "step": 1745
+    },
+    {
+      "epoch": 0.027936,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.799193548387097e-05,
+      "loss": 0.1802,
+      "step": 1746
+    },
+    {
+      "epoch": 0.027952,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.799032258064517e-05,
+      "loss": 0.2244,
+      "step": 1747
+    },
+    {
+      "epoch": 0.027968,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.798870967741935e-05,
+      "loss": 0.2025,
+      "step": 1748
+    },
+    {
+      "epoch": 0.027984,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.798709677419355e-05,
+      "loss": 0.2027,
+      "step": 1749
+    },
+    {
+      "epoch": 0.028,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.798548387096774e-05,
+      "loss": 0.1923,
+      "step": 1750
+    },
+    {
+      "epoch": 0.028016,
+      "grad_norm": 1.0,
+      "learning_rate": 9.798387096774194e-05,
+      "loss": 0.1765,
+      "step": 1751
+    },
+    {
+      "epoch": 0.028032,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.798225806451612e-05,
+      "loss": 0.1843,
+      "step": 1752
+    },
+    {
+      "epoch": 0.028048,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.798064516129032e-05,
+      "loss": 0.2261,
+      "step": 1753
+    },
+    {
+      "epoch": 0.028064,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.797903225806452e-05,
+      "loss": 0.1798,
+      "step": 1754
+    },
+    {
+      "epoch": 0.02808,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.797741935483872e-05,
+      "loss": 0.2113,
+      "step": 1755
+    },
+    {
+      "epoch": 0.028096,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.797580645161291e-05,
+      "loss": 0.194,
+      "step": 1756
+    },
+    {
+      "epoch": 0.028112,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.797419354838711e-05,
+      "loss": 0.159,
+      "step": 1757
+    },
+    {
+      "epoch": 0.028128,
+      "grad_norm": 1.125,
+      "learning_rate": 9.79725806451613e-05,
+      "loss": 0.243,
+      "step": 1758
+    },
+    {
+      "epoch": 0.028144,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.79709677419355e-05,
+      "loss": 0.1904,
+      "step": 1759
+    },
+    {
+      "epoch": 0.02816,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.796935483870968e-05,
+      "loss": 0.2139,
+      "step": 1760
+    },
+    {
+      "epoch": 0.028176,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.796774193548388e-05,
+      "loss": 0.2021,
+      "step": 1761
+    },
+    {
+      "epoch": 0.028192,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.796612903225807e-05,
+      "loss": 0.1727,
+      "step": 1762
+    },
+    {
+      "epoch": 0.028208,
+      "grad_norm": 1.328125,
+      "learning_rate": 9.796451612903225e-05,
+      "loss": 0.1774,
+      "step": 1763
+    },
+    {
+      "epoch": 0.028224,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.796290322580645e-05,
+      "loss": 0.2174,
+      "step": 1764
+    },
+    {
+      "epoch": 0.02824,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.796129032258064e-05,
+      "loss": 0.2303,
+      "step": 1765
+    },
+    {
+      "epoch": 0.028256,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.795967741935484e-05,
+      "loss": 0.1828,
+      "step": 1766
+    },
+    {
+      "epoch": 0.028272,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.795806451612904e-05,
+      "loss": 0.2115,
+      "step": 1767
+    },
+    {
+      "epoch": 0.028288,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.795645161290324e-05,
+      "loss": 0.2114,
+      "step": 1768
+    },
+    {
+      "epoch": 0.028304,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.795483870967742e-05,
+      "loss": 0.1923,
+      "step": 1769
+    },
+    {
+      "epoch": 0.02832,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.795322580645162e-05,
+      "loss": 0.2084,
+      "step": 1770
+    },
+    {
+      "epoch": 0.028336,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.795161290322581e-05,
+      "loss": 0.2269,
+      "step": 1771
+    },
+    {
+      "epoch": 0.028352,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.795000000000001e-05,
+      "loss": 0.1934,
+      "step": 1772
+    },
+    {
+      "epoch": 0.028368,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.79483870967742e-05,
+      "loss": 0.164,
+      "step": 1773
+    },
+    {
+      "epoch": 0.028384,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.79467741935484e-05,
+      "loss": 0.2096,
+      "step": 1774
+    },
+    {
+      "epoch": 0.0284,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.794516129032258e-05,
+      "loss": 0.1904,
+      "step": 1775
+    },
+    {
+      "epoch": 0.028416,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.794354838709678e-05,
+      "loss": 0.1813,
+      "step": 1776
+    },
+    {
+      "epoch": 0.028432,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.794193548387097e-05,
+      "loss": 0.2026,
+      "step": 1777
+    },
+    {
+      "epoch": 0.028448,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.794032258064517e-05,
+      "loss": 0.2248,
+      "step": 1778
+    },
+    {
+      "epoch": 0.028464,
+      "grad_norm": 1.0,
+      "learning_rate": 9.793870967741937e-05,
+      "loss": 0.1975,
+      "step": 1779
+    },
+    {
+      "epoch": 0.02848,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.793709677419355e-05,
+      "loss": 0.2108,
+      "step": 1780
+    },
+    {
+      "epoch": 0.028496,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.793548387096775e-05,
+      "loss": 0.1758,
+      "step": 1781
+    },
+    {
+      "epoch": 0.028512,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.793387096774194e-05,
+      "loss": 0.1539,
+      "step": 1782
+    },
+    {
+      "epoch": 0.028528,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.793225806451614e-05,
+      "loss": 0.251,
+      "step": 1783
+    },
+    {
+      "epoch": 0.028544,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.793064516129032e-05,
+      "loss": 0.1547,
+      "step": 1784
+    },
+    {
+      "epoch": 0.02856,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.792903225806452e-05,
+      "loss": 0.1604,
+      "step": 1785
+    },
+    {
+      "epoch": 0.028576,
+      "grad_norm": 0.875,
+      "learning_rate": 9.792741935483871e-05,
+      "loss": 0.2064,
+      "step": 1786
+    },
+    {
+      "epoch": 0.028592,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.792580645161291e-05,
+      "loss": 0.192,
+      "step": 1787
+    },
+    {
+      "epoch": 0.028608,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.79241935483871e-05,
+      "loss": 0.176,
+      "step": 1788
+    },
+    {
+      "epoch": 0.028624,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.79225806451613e-05,
+      "loss": 0.2129,
+      "step": 1789
+    },
+    {
+      "epoch": 0.02864,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.79209677419355e-05,
+      "loss": 0.2106,
+      "step": 1790
+    },
+    {
+      "epoch": 0.028656,
+      "grad_norm": 0.875,
+      "learning_rate": 9.79193548387097e-05,
+      "loss": 0.1872,
+      "step": 1791
+    },
+    {
+      "epoch": 0.028672,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.791774193548388e-05,
+      "loss": 0.191,
+      "step": 1792
+    },
+    {
+      "epoch": 0.028688,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.791612903225807e-05,
+      "loss": 0.1752,
+      "step": 1793
+    },
+    {
+      "epoch": 0.028704,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.791451612903227e-05,
+      "loss": 0.1573,
+      "step": 1794
+    },
+    {
+      "epoch": 0.02872,
+      "grad_norm": 1.7734375,
+      "learning_rate": 9.791290322580645e-05,
+      "loss": 0.2313,
+      "step": 1795
+    },
+    {
+      "epoch": 0.028736,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.791129032258065e-05,
+      "loss": 0.2015,
+      "step": 1796
+    },
+    {
+      "epoch": 0.028752,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.790967741935484e-05,
+      "loss": 0.1847,
+      "step": 1797
+    },
+    {
+      "epoch": 0.028768,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.790806451612904e-05,
+      "loss": 0.1293,
+      "step": 1798
+    },
+    {
+      "epoch": 0.028784,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.790645161290322e-05,
+      "loss": 0.1899,
+      "step": 1799
+    },
+    {
+      "epoch": 0.0288,
+      "grad_norm": 1.125,
+      "learning_rate": 9.790483870967742e-05,
+      "loss": 0.1912,
+      "step": 1800
+    },
+    {
+      "epoch": 0.028816,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.790322580645161e-05,
+      "loss": 0.2247,
+      "step": 1801
+    },
+    {
+      "epoch": 0.028832,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.790161290322581e-05,
+      "loss": 0.2031,
+      "step": 1802
+    },
+    {
+      "epoch": 0.028848,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.790000000000001e-05,
+      "loss": 0.2128,
+      "step": 1803
+    },
+    {
+      "epoch": 0.028864,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.789838709677421e-05,
+      "loss": 0.1995,
+      "step": 1804
+    },
+    {
+      "epoch": 0.02888,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.78967741935484e-05,
+      "loss": 0.2204,
+      "step": 1805
+    },
+    {
+      "epoch": 0.028896,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.78951612903226e-05,
+      "loss": 0.2098,
+      "step": 1806
+    },
+    {
+      "epoch": 0.028912,
+      "grad_norm": 1.4609375,
+      "learning_rate": 9.789354838709678e-05,
+      "loss": 0.2001,
+      "step": 1807
+    },
+    {
+      "epoch": 0.028928,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.789193548387098e-05,
+      "loss": 0.218,
+      "step": 1808
+    },
+    {
+      "epoch": 0.028944,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.789032258064517e-05,
+      "loss": 0.1852,
+      "step": 1809
+    },
+    {
+      "epoch": 0.02896,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.788870967741935e-05,
+      "loss": 0.1926,
+      "step": 1810
+    },
+    {
+      "epoch": 0.028976,
+      "grad_norm": 1.515625,
+      "learning_rate": 9.788709677419355e-05,
+      "loss": 0.1722,
+      "step": 1811
+    },
+    {
+      "epoch": 0.028992,
+      "grad_norm": 1.4765625,
+      "learning_rate": 9.788548387096774e-05,
+      "loss": 0.2001,
+      "step": 1812
+    },
+    {
+      "epoch": 0.029008,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.788387096774194e-05,
+      "loss": 0.1724,
+      "step": 1813
+    },
+    {
+      "epoch": 0.029024,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.788225806451614e-05,
+      "loss": 0.19,
+      "step": 1814
+    },
+    {
+      "epoch": 0.02904,
+      "grad_norm": 1.4140625,
+      "learning_rate": 9.788064516129034e-05,
+      "loss": 0.2299,
+      "step": 1815
+    },
+    {
+      "epoch": 0.029056,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.787903225806452e-05,
+      "loss": 0.1502,
+      "step": 1816
+    },
+    {
+      "epoch": 0.029072,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.787741935483872e-05,
+      "loss": 0.237,
+      "step": 1817
+    },
+    {
+      "epoch": 0.029088,
+      "grad_norm": 1.8828125,
+      "learning_rate": 9.787580645161291e-05,
+      "loss": 0.2401,
+      "step": 1818
+    },
+    {
+      "epoch": 0.029104,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.787419354838711e-05,
+      "loss": 0.1919,
+      "step": 1819
+    },
+    {
+      "epoch": 0.02912,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.78725806451613e-05,
+      "loss": 0.2194,
+      "step": 1820
+    },
+    {
+      "epoch": 0.029136,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.787096774193549e-05,
+      "loss": 0.2104,
+      "step": 1821
+    },
+    {
+      "epoch": 0.029152,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.786935483870968e-05,
+      "loss": 0.1965,
+      "step": 1822
+    },
+    {
+      "epoch": 0.029168,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.786774193548388e-05,
+      "loss": 0.2141,
+      "step": 1823
+    },
+    {
+      "epoch": 0.029184,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.786612903225806e-05,
+      "loss": 0.1884,
+      "step": 1824
+    },
+    {
+      "epoch": 0.0292,
+      "grad_norm": 1.359375,
+      "learning_rate": 9.786451612903226e-05,
+      "loss": 0.2308,
+      "step": 1825
+    },
+    {
+      "epoch": 0.029216,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.786290322580645e-05,
+      "loss": 0.2439,
+      "step": 1826
+    },
+    {
+      "epoch": 0.029232,
+      "grad_norm": 1.375,
+      "learning_rate": 9.786129032258065e-05,
+      "loss": 0.2275,
+      "step": 1827
+    },
+    {
+      "epoch": 0.029248,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.785967741935485e-05,
+      "loss": 0.1776,
+      "step": 1828
+    },
+    {
+      "epoch": 0.029264,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.785806451612904e-05,
+      "loss": 0.2132,
+      "step": 1829
+    },
+    {
+      "epoch": 0.02928,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.785645161290324e-05,
+      "loss": 0.1965,
+      "step": 1830
+    },
+    {
+      "epoch": 0.029296,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.785483870967742e-05,
+      "loss": 0.1916,
+      "step": 1831
+    },
+    {
+      "epoch": 0.029312,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.785322580645162e-05,
+      "loss": 0.2114,
+      "step": 1832
+    },
+    {
+      "epoch": 0.029328,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.785161290322581e-05,
+      "loss": 0.1829,
+      "step": 1833
+    },
+    {
+      "epoch": 0.029344,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.785e-05,
+      "loss": 0.2123,
+      "step": 1834
+    },
+    {
+      "epoch": 0.02936,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.784838709677419e-05,
+      "loss": 0.1967,
+      "step": 1835
+    },
+    {
+      "epoch": 0.029376,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.784677419354839e-05,
+      "loss": 0.2058,
+      "step": 1836
+    },
+    {
+      "epoch": 0.029392,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.784516129032258e-05,
+      "loss": 0.1717,
+      "step": 1837
+    },
+    {
+      "epoch": 0.029408,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.784354838709678e-05,
+      "loss": 0.1544,
+      "step": 1838
+    },
+    {
+      "epoch": 0.029424,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.784193548387098e-05,
+      "loss": 0.1969,
+      "step": 1839
+    },
+    {
+      "epoch": 0.02944,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.784032258064516e-05,
+      "loss": 0.1625,
+      "step": 1840
+    },
+    {
+      "epoch": 0.029456,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.783870967741936e-05,
+      "loss": 0.1776,
+      "step": 1841
+    },
+    {
+      "epoch": 0.029472,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.783709677419355e-05,
+      "loss": 0.1728,
+      "step": 1842
+    },
+    {
+      "epoch": 0.029488,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.783548387096775e-05,
+      "loss": 0.2484,
+      "step": 1843
+    },
+    {
+      "epoch": 0.029504,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.783387096774194e-05,
+      "loss": 0.2005,
+      "step": 1844
+    },
+    {
+      "epoch": 0.02952,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.783225806451613e-05,
+      "loss": 0.1811,
+      "step": 1845
+    },
+    {
+      "epoch": 0.029536,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.783064516129032e-05,
+      "loss": 0.2276,
+      "step": 1846
+    },
+    {
+      "epoch": 0.029552,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.782903225806452e-05,
+      "loss": 0.1961,
+      "step": 1847
+    },
+    {
+      "epoch": 0.029568,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.78274193548387e-05,
+      "loss": 0.1933,
+      "step": 1848
+    },
+    {
+      "epoch": 0.029584,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.78258064516129e-05,
+      "loss": 0.2368,
+      "step": 1849
+    },
+    {
+      "epoch": 0.0296,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.78241935483871e-05,
+      "loss": 0.1928,
+      "step": 1850
+    },
+    {
+      "epoch": 0.029616,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.78225806451613e-05,
+      "loss": 0.195,
+      "step": 1851
+    },
+    {
+      "epoch": 0.029632,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.782096774193549e-05,
+      "loss": 0.2099,
+      "step": 1852
+    },
+    {
+      "epoch": 0.029648,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.781935483870969e-05,
+      "loss": 0.1937,
+      "step": 1853
+    },
+    {
+      "epoch": 0.029664,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.781774193548388e-05,
+      "loss": 0.2397,
+      "step": 1854
+    },
+    {
+      "epoch": 0.02968,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.781612903225806e-05,
+      "loss": 0.1924,
+      "step": 1855
+    },
+    {
+      "epoch": 0.029696,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.781451612903226e-05,
+      "loss": 0.2086,
+      "step": 1856
+    },
+    {
+      "epoch": 0.029712,
+      "grad_norm": 1.125,
+      "learning_rate": 9.781290322580645e-05,
+      "loss": 0.2051,
+      "step": 1857
+    },
+    {
+      "epoch": 0.029728,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.781129032258065e-05,
+      "loss": 0.1971,
+      "step": 1858
+    },
+    {
+      "epoch": 0.029744,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.780967741935483e-05,
+      "loss": 0.1868,
+      "step": 1859
+    },
+    {
+      "epoch": 0.02976,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.780806451612903e-05,
+      "loss": 0.1709,
+      "step": 1860
+    },
+    {
+      "epoch": 0.029776,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.780645161290322e-05,
+      "loss": 0.229,
+      "step": 1861
+    },
+    {
+      "epoch": 0.029792,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.780483870967742e-05,
+      "loss": 0.1855,
+      "step": 1862
+    },
+    {
+      "epoch": 0.029808,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.780322580645162e-05,
+      "loss": 0.1571,
+      "step": 1863
+    },
+    {
+      "epoch": 0.029824,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.780161290322582e-05,
+      "loss": 0.1968,
+      "step": 1864
+    },
+    {
+      "epoch": 0.02984,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.78e-05,
+      "loss": 0.1775,
+      "step": 1865
+    },
+    {
+      "epoch": 0.029856,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.77983870967742e-05,
+      "loss": 0.2487,
+      "step": 1866
+    },
+    {
+      "epoch": 0.029872,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.779677419354839e-05,
+      "loss": 0.1931,
+      "step": 1867
+    },
+    {
+      "epoch": 0.029888,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.779516129032259e-05,
+      "loss": 0.1803,
+      "step": 1868
+    },
+    {
+      "epoch": 0.029904,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.779354838709678e-05,
+      "loss": 0.1581,
+      "step": 1869
+    },
+    {
+      "epoch": 0.02992,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.779193548387098e-05,
+      "loss": 0.168,
+      "step": 1870
+    },
+    {
+      "epoch": 0.029936,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.779032258064516e-05,
+      "loss": 0.2035,
+      "step": 1871
+    },
+    {
+      "epoch": 0.029952,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.778870967741935e-05,
+      "loss": 0.1953,
+      "step": 1872
+    },
+    {
+      "epoch": 0.029968,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.778709677419355e-05,
+      "loss": 0.1685,
+      "step": 1873
+    },
+    {
+      "epoch": 0.029984,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.778548387096775e-05,
+      "loss": 0.2401,
+      "step": 1874
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.778387096774195e-05,
+      "loss": 0.216,
+      "step": 1875
+    },
+    {
+      "epoch": 0.030016,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.778225806451613e-05,
+      "loss": 0.2555,
+      "step": 1876
+    },
+    {
+      "epoch": 0.030032,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.778064516129033e-05,
+      "loss": 0.2078,
+      "step": 1877
+    },
+    {
+      "epoch": 0.030048,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.777903225806452e-05,
+      "loss": 0.2042,
+      "step": 1878
+    },
+    {
+      "epoch": 0.030064,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.777741935483872e-05,
+      "loss": 0.1949,
+      "step": 1879
+    },
+    {
+      "epoch": 0.03008,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.77758064516129e-05,
+      "loss": 0.1803,
+      "step": 1880
+    },
+    {
+      "epoch": 0.030096,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.77741935483871e-05,
+      "loss": 0.1911,
+      "step": 1881
+    },
+    {
+      "epoch": 0.030112,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.777258064516129e-05,
+      "loss": 0.1609,
+      "step": 1882
+    },
+    {
+      "epoch": 0.030128,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.777096774193549e-05,
+      "loss": 0.1881,
+      "step": 1883
+    },
+    {
+      "epoch": 0.030144,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.776935483870968e-05,
+      "loss": 0.1991,
+      "step": 1884
+    },
+    {
+      "epoch": 0.03016,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.776774193548388e-05,
+      "loss": 0.2141,
+      "step": 1885
+    },
+    {
+      "epoch": 0.030176,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.776612903225808e-05,
+      "loss": 0.177,
+      "step": 1886
+    },
+    {
+      "epoch": 0.030192,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.776451612903226e-05,
+      "loss": 0.1866,
+      "step": 1887
+    },
+    {
+      "epoch": 0.030208,
+      "grad_norm": 0.875,
+      "learning_rate": 9.776290322580646e-05,
+      "loss": 0.1801,
+      "step": 1888
+    },
+    {
+      "epoch": 0.030224,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.776129032258065e-05,
+      "loss": 0.1893,
+      "step": 1889
+    },
+    {
+      "epoch": 0.03024,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.775967741935485e-05,
+      "loss": 0.1829,
+      "step": 1890
+    },
+    {
+      "epoch": 0.030256,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.775806451612903e-05,
+      "loss": 0.2079,
+      "step": 1891
+    },
+    {
+      "epoch": 0.030272,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.775645161290323e-05,
+      "loss": 0.2018,
+      "step": 1892
+    },
+    {
+      "epoch": 0.030288,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.775483870967742e-05,
+      "loss": 0.2375,
+      "step": 1893
+    },
+    {
+      "epoch": 0.030304,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.775322580645162e-05,
+      "loss": 0.2,
+      "step": 1894
+    },
+    {
+      "epoch": 0.03032,
+      "grad_norm": 0.75,
+      "learning_rate": 9.77516129032258e-05,
+      "loss": 0.1587,
+      "step": 1895
+    },
+    {
+      "epoch": 0.030336,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.775e-05,
+      "loss": 0.214,
+      "step": 1896
+    },
+    {
+      "epoch": 0.030352,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.774838709677419e-05,
+      "loss": 0.2185,
+      "step": 1897
+    },
+    {
+      "epoch": 0.030368,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.774677419354839e-05,
+      "loss": 0.1759,
+      "step": 1898
+    },
+    {
+      "epoch": 0.030384,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.774516129032259e-05,
+      "loss": 0.2089,
+      "step": 1899
+    },
+    {
+      "epoch": 0.0304,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.774354838709679e-05,
+      "loss": 0.1674,
+      "step": 1900
+    },
+    {
+      "epoch": 0.030416,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.774193548387098e-05,
+      "loss": 0.2352,
+      "step": 1901
+    },
+    {
+      "epoch": 0.030432,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.774032258064516e-05,
+      "loss": 0.1454,
+      "step": 1902
+    },
+    {
+      "epoch": 0.030448,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.773870967741936e-05,
+      "loss": 0.1559,
+      "step": 1903
+    },
+    {
+      "epoch": 0.030464,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.773709677419355e-05,
+      "loss": 0.1634,
+      "step": 1904
+    },
+    {
+      "epoch": 0.03048,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.773548387096775e-05,
+      "loss": 0.2115,
+      "step": 1905
+    },
+    {
+      "epoch": 0.030496,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.773387096774193e-05,
+      "loss": 0.1558,
+      "step": 1906
+    },
+    {
+      "epoch": 0.030512,
+      "grad_norm": 0.75,
+      "learning_rate": 9.773225806451613e-05,
+      "loss": 0.1887,
+      "step": 1907
+    },
+    {
+      "epoch": 0.030528,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.773064516129032e-05,
+      "loss": 0.1702,
+      "step": 1908
+    },
+    {
+      "epoch": 0.030544,
+      "grad_norm": 1.4296875,
+      "learning_rate": 9.772903225806452e-05,
+      "loss": 0.2358,
+      "step": 1909
+    },
+    {
+      "epoch": 0.03056,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.772741935483872e-05,
+      "loss": 0.1717,
+      "step": 1910
+    },
+    {
+      "epoch": 0.030576,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.772580645161292e-05,
+      "loss": 0.1932,
+      "step": 1911
+    },
+    {
+      "epoch": 0.030592,
+      "grad_norm": 0.875,
+      "learning_rate": 9.77241935483871e-05,
+      "loss": 0.2144,
+      "step": 1912
+    },
+    {
+      "epoch": 0.030608,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.77225806451613e-05,
+      "loss": 0.2,
+      "step": 1913
+    },
+    {
+      "epoch": 0.030624,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.772096774193549e-05,
+      "loss": 0.2143,
+      "step": 1914
+    },
+    {
+      "epoch": 0.03064,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.771935483870969e-05,
+      "loss": 0.2166,
+      "step": 1915
+    },
+    {
+      "epoch": 0.030656,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.771774193548387e-05,
+      "loss": 0.2578,
+      "step": 1916
+    },
+    {
+      "epoch": 0.030672,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.771612903225807e-05,
+      "loss": 0.2259,
+      "step": 1917
+    },
+    {
+      "epoch": 0.030688,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.771451612903226e-05,
+      "loss": 0.1907,
+      "step": 1918
+    },
+    {
+      "epoch": 0.030704,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.771290322580645e-05,
+      "loss": 0.2461,
+      "step": 1919
+    },
+    {
+      "epoch": 0.03072,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.771129032258065e-05,
+      "loss": 0.2036,
+      "step": 1920
+    },
+    {
+      "epoch": 0.030736,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.770967741935485e-05,
+      "loss": 0.1889,
+      "step": 1921
+    },
+    {
+      "epoch": 0.030752,
+      "grad_norm": 1.0,
+      "learning_rate": 9.770806451612903e-05,
+      "loss": 0.2045,
+      "step": 1922
+    },
+    {
+      "epoch": 0.030768,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.770645161290323e-05,
+      "loss": 0.1916,
+      "step": 1923
+    },
+    {
+      "epoch": 0.030784,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.770483870967743e-05,
+      "loss": 0.2101,
+      "step": 1924
+    },
+    {
+      "epoch": 0.0308,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.770322580645162e-05,
+      "loss": 0.1791,
+      "step": 1925
+    },
+    {
+      "epoch": 0.030816,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.770161290322582e-05,
+      "loss": 0.2178,
+      "step": 1926
+    },
+    {
+      "epoch": 0.030832,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.77e-05,
+      "loss": 0.1843,
+      "step": 1927
+    },
+    {
+      "epoch": 0.030848,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.76983870967742e-05,
+      "loss": 0.19,
+      "step": 1928
+    },
+    {
+      "epoch": 0.030864,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.769677419354839e-05,
+      "loss": 0.1871,
+      "step": 1929
+    },
+    {
+      "epoch": 0.03088,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.769516129032259e-05,
+      "loss": 0.1834,
+      "step": 1930
+    },
+    {
+      "epoch": 0.030896,
+      "grad_norm": 1.25,
+      "learning_rate": 9.769354838709677e-05,
+      "loss": 0.2097,
+      "step": 1931
+    },
+    {
+      "epoch": 0.030912,
+      "grad_norm": 1.5,
+      "learning_rate": 9.769193548387097e-05,
+      "loss": 0.1889,
+      "step": 1932
+    },
+    {
+      "epoch": 0.030928,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.769032258064516e-05,
+      "loss": 0.2051,
+      "step": 1933
+    },
+    {
+      "epoch": 0.030944,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.768870967741936e-05,
+      "loss": 0.1385,
+      "step": 1934
+    },
+    {
+      "epoch": 0.03096,
+      "grad_norm": 1.375,
+      "learning_rate": 9.768709677419356e-05,
+      "loss": 0.2259,
+      "step": 1935
+    },
+    {
+      "epoch": 0.030976,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.768548387096775e-05,
+      "loss": 0.1777,
+      "step": 1936
+    },
+    {
+      "epoch": 0.030992,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.768387096774195e-05,
+      "loss": 0.2092,
+      "step": 1937
+    },
+    {
+      "epoch": 0.031008,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.768225806451613e-05,
+      "loss": 0.1559,
+      "step": 1938
+    },
+    {
+      "epoch": 0.031024,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.768064516129033e-05,
+      "loss": 0.212,
+      "step": 1939
+    },
+    {
+      "epoch": 0.03104,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.767903225806452e-05,
+      "loss": 0.2009,
+      "step": 1940
+    },
+    {
+      "epoch": 0.031056,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.767741935483872e-05,
+      "loss": 0.1727,
+      "step": 1941
+    },
+    {
+      "epoch": 0.031072,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.76758064516129e-05,
+      "loss": 0.2189,
+      "step": 1942
+    },
+    {
+      "epoch": 0.031088,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.76741935483871e-05,
+      "loss": 0.2364,
+      "step": 1943
+    },
+    {
+      "epoch": 0.031104,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.767258064516129e-05,
+      "loss": 0.1594,
+      "step": 1944
+    },
+    {
+      "epoch": 0.03112,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.767096774193549e-05,
+      "loss": 0.2169,
+      "step": 1945
+    },
+    {
+      "epoch": 0.031136,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.766935483870969e-05,
+      "loss": 0.1528,
+      "step": 1946
+    },
+    {
+      "epoch": 0.031152,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.766774193548389e-05,
+      "loss": 0.1843,
+      "step": 1947
+    },
+    {
+      "epoch": 0.031168,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.766612903225807e-05,
+      "loss": 0.2302,
+      "step": 1948
+    },
+    {
+      "epoch": 0.031184,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.766451612903226e-05,
+      "loss": 0.2144,
+      "step": 1949
+    },
+    {
+      "epoch": 0.0312,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.766290322580646e-05,
+      "loss": 0.1927,
+      "step": 1950
+    },
+    {
+      "epoch": 0.031216,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.766129032258065e-05,
+      "loss": 0.1989,
+      "step": 1951
+    },
+    {
+      "epoch": 0.031232,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.765967741935484e-05,
+      "loss": 0.1613,
+      "step": 1952
+    },
+    {
+      "epoch": 0.031248,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.765806451612903e-05,
+      "loss": 0.1935,
+      "step": 1953
+    },
+    {
+      "epoch": 0.031264,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.765645161290323e-05,
+      "loss": 0.2023,
+      "step": 1954
+    },
+    {
+      "epoch": 0.03128,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.765483870967742e-05,
+      "loss": 0.2453,
+      "step": 1955
+    },
+    {
+      "epoch": 0.031296,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.765322580645162e-05,
+      "loss": 0.1856,
+      "step": 1956
+    },
+    {
+      "epoch": 0.031312,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.76516129032258e-05,
+      "loss": 0.2178,
+      "step": 1957
+    },
+    {
+      "epoch": 0.031328,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.765e-05,
+      "loss": 0.2026,
+      "step": 1958
+    },
+    {
+      "epoch": 0.031344,
+      "grad_norm": 1.875,
+      "learning_rate": 9.76483870967742e-05,
+      "loss": 0.2333,
+      "step": 1959
+    },
+    {
+      "epoch": 0.03136,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.76467741935484e-05,
+      "loss": 0.1828,
+      "step": 1960
+    },
+    {
+      "epoch": 0.031376,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.764516129032259e-05,
+      "loss": 0.2002,
+      "step": 1961
+    },
+    {
+      "epoch": 0.031392,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.764354838709679e-05,
+      "loss": 0.2265,
+      "step": 1962
+    },
+    {
+      "epoch": 0.031408,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.764193548387097e-05,
+      "loss": 0.2125,
+      "step": 1963
+    },
+    {
+      "epoch": 0.031424,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.764032258064516e-05,
+      "loss": 0.2489,
+      "step": 1964
+    },
+    {
+      "epoch": 0.03144,
+      "grad_norm": 0.75,
+      "learning_rate": 9.763870967741936e-05,
+      "loss": 0.1656,
+      "step": 1965
+    },
+    {
+      "epoch": 0.031456,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.763709677419354e-05,
+      "loss": 0.1529,
+      "step": 1966
+    },
+    {
+      "epoch": 0.031472,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.763548387096774e-05,
+      "loss": 0.2022,
+      "step": 1967
+    },
+    {
+      "epoch": 0.031488,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.763387096774193e-05,
+      "loss": 0.2152,
+      "step": 1968
+    },
+    {
+      "epoch": 0.031504,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.763225806451613e-05,
+      "loss": 0.2129,
+      "step": 1969
+    },
+    {
+      "epoch": 0.03152,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.763064516129033e-05,
+      "loss": 0.1971,
+      "step": 1970
+    },
+    {
+      "epoch": 0.031536,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.762903225806453e-05,
+      "loss": 0.2058,
+      "step": 1971
+    },
+    {
+      "epoch": 0.031552,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.762741935483872e-05,
+      "loss": 0.2166,
+      "step": 1972
+    },
+    {
+      "epoch": 0.031568,
+      "grad_norm": 1.4609375,
+      "learning_rate": 9.762580645161292e-05,
+      "loss": 0.2445,
+      "step": 1973
+    },
+    {
+      "epoch": 0.031584,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.76241935483871e-05,
+      "loss": 0.164,
+      "step": 1974
+    },
+    {
+      "epoch": 0.0316,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.76225806451613e-05,
+      "loss": 0.22,
+      "step": 1975
+    },
+    {
+      "epoch": 0.031616,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.762096774193549e-05,
+      "loss": 0.2193,
+      "step": 1976
+    },
+    {
+      "epoch": 0.031632,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.761935483870969e-05,
+      "loss": 0.1975,
+      "step": 1977
+    },
+    {
+      "epoch": 0.031648,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.761774193548387e-05,
+      "loss": 0.1879,
+      "step": 1978
+    },
+    {
+      "epoch": 0.031664,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.761612903225807e-05,
+      "loss": 0.1692,
+      "step": 1979
+    },
+    {
+      "epoch": 0.03168,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.761451612903226e-05,
+      "loss": 0.1629,
+      "step": 1980
+    },
+    {
+      "epoch": 0.031696,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.761290322580646e-05,
+      "loss": 0.226,
+      "step": 1981
+    },
+    {
+      "epoch": 0.031712,
+      "grad_norm": 1.8828125,
+      "learning_rate": 9.761129032258066e-05,
+      "loss": 0.2011,
+      "step": 1982
+    },
+    {
+      "epoch": 0.031728,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.760967741935484e-05,
+      "loss": 0.1988,
+      "step": 1983
+    },
+    {
+      "epoch": 0.031744,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.760806451612904e-05,
+      "loss": 0.2163,
+      "step": 1984
+    },
+    {
+      "epoch": 0.03176,
+      "grad_norm": 1.0,
+      "learning_rate": 9.760645161290323e-05,
+      "loss": 0.1845,
+      "step": 1985
+    },
+    {
+      "epoch": 0.031776,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.760483870967743e-05,
+      "loss": 0.1643,
+      "step": 1986
+    },
+    {
+      "epoch": 0.031792,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.760322580645161e-05,
+      "loss": 0.2323,
+      "step": 1987
+    },
+    {
+      "epoch": 0.031808,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.760161290322581e-05,
+      "loss": 0.1602,
+      "step": 1988
+    },
+    {
+      "epoch": 0.031824,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.76e-05,
+      "loss": 0.1981,
+      "step": 1989
+    },
+    {
+      "epoch": 0.03184,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.75983870967742e-05,
+      "loss": 0.199,
+      "step": 1990
+    },
+    {
+      "epoch": 0.031856,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.759677419354839e-05,
+      "loss": 0.154,
+      "step": 1991
+    },
+    {
+      "epoch": 0.031872,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.759516129032259e-05,
+      "loss": 0.1965,
+      "step": 1992
+    },
+    {
+      "epoch": 0.031888,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.759354838709677e-05,
+      "loss": 0.1941,
+      "step": 1993
+    },
+    {
+      "epoch": 0.031904,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.759193548387097e-05,
+      "loss": 0.2081,
+      "step": 1994
+    },
+    {
+      "epoch": 0.03192,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.759032258064517e-05,
+      "loss": 0.1667,
+      "step": 1995
+    },
+    {
+      "epoch": 0.031936,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.758870967741936e-05,
+      "loss": 0.1762,
+      "step": 1996
+    },
+    {
+      "epoch": 0.031952,
+      "grad_norm": 1.390625,
+      "learning_rate": 9.758709677419356e-05,
+      "loss": 0.1983,
+      "step": 1997
+    },
+    {
+      "epoch": 0.031968,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.758548387096774e-05,
+      "loss": 0.2065,
+      "step": 1998
+    },
+    {
+      "epoch": 0.031984,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.758387096774194e-05,
+      "loss": 0.1833,
+      "step": 1999
+    },
+    {
+      "epoch": 0.032,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.758225806451613e-05,
+      "loss": 0.205,
+      "step": 2000
+    },
+    {
+      "epoch": 0.032016,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.758064516129033e-05,
+      "loss": 0.2215,
+      "step": 2001
+    },
+    {
+      "epoch": 0.032032,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.757903225806451e-05,
+      "loss": 0.1842,
+      "step": 2002
+    },
+    {
+      "epoch": 0.032048,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.757741935483871e-05,
+      "loss": 0.1635,
+      "step": 2003
+    },
+    {
+      "epoch": 0.032064,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.75758064516129e-05,
+      "loss": 0.1536,
+      "step": 2004
+    },
+    {
+      "epoch": 0.03208,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.75741935483871e-05,
+      "loss": 0.1957,
+      "step": 2005
+    },
+    {
+      "epoch": 0.032096,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.75725806451613e-05,
+      "loss": 0.2031,
+      "step": 2006
+    },
+    {
+      "epoch": 0.032112,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.75709677419355e-05,
+      "loss": 0.2321,
+      "step": 2007
+    },
+    {
+      "epoch": 0.032128,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.756935483870969e-05,
+      "loss": 0.2223,
+      "step": 2008
+    },
+    {
+      "epoch": 0.032144,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.756774193548388e-05,
+      "loss": 0.1838,
+      "step": 2009
+    },
+    {
+      "epoch": 0.03216,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.756612903225807e-05,
+      "loss": 0.2025,
+      "step": 2010
+    },
+    {
+      "epoch": 0.032176,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.756451612903226e-05,
+      "loss": 0.1928,
+      "step": 2011
+    },
+    {
+      "epoch": 0.032192,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.756290322580646e-05,
+      "loss": 0.2387,
+      "step": 2012
+    },
+    {
+      "epoch": 0.032208,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.756129032258064e-05,
+      "loss": 0.2149,
+      "step": 2013
+    },
+    {
+      "epoch": 0.032224,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.755967741935484e-05,
+      "loss": 0.2356,
+      "step": 2014
+    },
+    {
+      "epoch": 0.03224,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.755806451612903e-05,
+      "loss": 0.2243,
+      "step": 2015
+    },
+    {
+      "epoch": 0.032256,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.755645161290323e-05,
+      "loss": 0.2539,
+      "step": 2016
+    },
+    {
+      "epoch": 0.032272,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.755483870967741e-05,
+      "loss": 0.196,
+      "step": 2017
+    },
+    {
+      "epoch": 0.032288,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.755322580645161e-05,
+      "loss": 0.2242,
+      "step": 2018
+    },
+    {
+      "epoch": 0.032304,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.755161290322581e-05,
+      "loss": 0.1993,
+      "step": 2019
+    },
+    {
+      "epoch": 0.03232,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.755000000000001e-05,
+      "loss": 0.1904,
+      "step": 2020
+    },
+    {
+      "epoch": 0.032336,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.75483870967742e-05,
+      "loss": 0.2103,
+      "step": 2021
+    },
+    {
+      "epoch": 0.032352,
+      "grad_norm": 0.875,
+      "learning_rate": 9.75467741935484e-05,
+      "loss": 0.1772,
+      "step": 2022
+    },
+    {
+      "epoch": 0.032368,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.754516129032258e-05,
+      "loss": 0.2246,
+      "step": 2023
+    },
+    {
+      "epoch": 0.032384,
+      "grad_norm": 0.875,
+      "learning_rate": 9.754354838709678e-05,
+      "loss": 0.216,
+      "step": 2024
+    },
+    {
+      "epoch": 0.0324,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.754193548387097e-05,
+      "loss": 0.1756,
+      "step": 2025
+    },
+    {
+      "epoch": 0.032416,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.754032258064517e-05,
+      "loss": 0.2004,
+      "step": 2026
+    },
+    {
+      "epoch": 0.032432,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.753870967741936e-05,
+      "loss": 0.2363,
+      "step": 2027
+    },
+    {
+      "epoch": 0.032448,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.753709677419354e-05,
+      "loss": 0.1596,
+      "step": 2028
+    },
+    {
+      "epoch": 0.032464,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.753548387096774e-05,
+      "loss": 0.2426,
+      "step": 2029
+    },
+    {
+      "epoch": 0.03248,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.753387096774194e-05,
+      "loss": 0.2078,
+      "step": 2030
+    },
+    {
+      "epoch": 0.032496,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.753225806451614e-05,
+      "loss": 0.2201,
+      "step": 2031
+    },
+    {
+      "epoch": 0.032512,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.753064516129033e-05,
+      "loss": 0.2215,
+      "step": 2032
+    },
+    {
+      "epoch": 0.032528,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.752903225806453e-05,
+      "loss": 0.1746,
+      "step": 2033
+    },
+    {
+      "epoch": 0.032544,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.752741935483871e-05,
+      "loss": 0.2281,
+      "step": 2034
+    },
+    {
+      "epoch": 0.03256,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.752580645161291e-05,
+      "loss": 0.1964,
+      "step": 2035
+    },
+    {
+      "epoch": 0.032576,
+      "grad_norm": 1.4375,
+      "learning_rate": 9.75241935483871e-05,
+      "loss": 0.1583,
+      "step": 2036
+    },
+    {
+      "epoch": 0.032592,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.75225806451613e-05,
+      "loss": 0.1989,
+      "step": 2037
+    },
+    {
+      "epoch": 0.032608,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.752096774193548e-05,
+      "loss": 0.1679,
+      "step": 2038
+    },
+    {
+      "epoch": 0.032624,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.751935483870968e-05,
+      "loss": 0.1838,
+      "step": 2039
+    },
+    {
+      "epoch": 0.03264,
+      "grad_norm": 1.125,
+      "learning_rate": 9.751774193548387e-05,
+      "loss": 0.2058,
+      "step": 2040
+    },
+    {
+      "epoch": 0.032656,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.751612903225807e-05,
+      "loss": 0.1723,
+      "step": 2041
+    },
+    {
+      "epoch": 0.032672,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.751451612903227e-05,
+      "loss": 0.1606,
+      "step": 2042
+    },
+    {
+      "epoch": 0.032688,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.751290322580646e-05,
+      "loss": 0.2094,
+      "step": 2043
+    },
+    {
+      "epoch": 0.032704,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.751129032258066e-05,
+      "loss": 0.1674,
+      "step": 2044
+    },
+    {
+      "epoch": 0.03272,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.750967741935484e-05,
+      "loss": 0.2389,
+      "step": 2045
+    },
+    {
+      "epoch": 0.032736,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.750806451612904e-05,
+      "loss": 0.2439,
+      "step": 2046
+    },
+    {
+      "epoch": 0.032752,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.750645161290323e-05,
+      "loss": 0.1983,
+      "step": 2047
+    },
+    {
+      "epoch": 0.032768,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.750483870967743e-05,
+      "loss": 0.1848,
+      "step": 2048
+    },
+    {
+      "epoch": 0.032784,
+      "grad_norm": 1.359375,
+      "learning_rate": 9.750322580645161e-05,
+      "loss": 0.1878,
+      "step": 2049
+    },
+    {
+      "epoch": 0.0328,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.750161290322581e-05,
+      "loss": 0.1556,
+      "step": 2050
+    },
+    {
+      "epoch": 0.032816,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.75e-05,
+      "loss": 0.1593,
+      "step": 2051
+    },
+    {
+      "epoch": 0.032832,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.74983870967742e-05,
+      "loss": 0.2277,
+      "step": 2052
+    },
+    {
+      "epoch": 0.032848,
+      "grad_norm": 1.796875,
+      "learning_rate": 9.749677419354838e-05,
+      "loss": 0.1995,
+      "step": 2053
+    },
+    {
+      "epoch": 0.032864,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.749516129032258e-05,
+      "loss": 0.2173,
+      "step": 2054
+    },
+    {
+      "epoch": 0.03288,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.749354838709678e-05,
+      "loss": 0.1855,
+      "step": 2055
+    },
+    {
+      "epoch": 0.032896,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.749193548387098e-05,
+      "loss": 0.156,
+      "step": 2056
+    },
+    {
+      "epoch": 0.032912,
+      "grad_norm": 1.359375,
+      "learning_rate": 9.749032258064517e-05,
+      "loss": 0.1987,
+      "step": 2057
+    },
+    {
+      "epoch": 0.032928,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.748870967741935e-05,
+      "loss": 0.1887,
+      "step": 2058
+    },
+    {
+      "epoch": 0.032944,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.748709677419355e-05,
+      "loss": 0.1588,
+      "step": 2059
+    },
+    {
+      "epoch": 0.03296,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.748548387096774e-05,
+      "loss": 0.2011,
+      "step": 2060
+    },
+    {
+      "epoch": 0.032976,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.748387096774194e-05,
+      "loss": 0.2281,
+      "step": 2061
+    },
+    {
+      "epoch": 0.032992,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.748225806451613e-05,
+      "loss": 0.1937,
+      "step": 2062
+    },
+    {
+      "epoch": 0.033008,
+      "grad_norm": 1.8203125,
+      "learning_rate": 9.748064516129033e-05,
+      "loss": 0.1721,
+      "step": 2063
+    },
+    {
+      "epoch": 0.033024,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.747903225806451e-05,
+      "loss": 0.2199,
+      "step": 2064
+    },
+    {
+      "epoch": 0.03304,
+      "grad_norm": 1.625,
+      "learning_rate": 9.747741935483871e-05,
+      "loss": 0.2021,
+      "step": 2065
+    },
+    {
+      "epoch": 0.033056,
+      "grad_norm": 1.4140625,
+      "learning_rate": 9.747580645161291e-05,
+      "loss": 0.176,
+      "step": 2066
+    },
+    {
+      "epoch": 0.033072,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.747419354838711e-05,
+      "loss": 0.2232,
+      "step": 2067
+    },
+    {
+      "epoch": 0.033088,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.74725806451613e-05,
+      "loss": 0.1883,
+      "step": 2068
+    },
+    {
+      "epoch": 0.033104,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.74709677419355e-05,
+      "loss": 0.1794,
+      "step": 2069
+    },
+    {
+      "epoch": 0.03312,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.746935483870968e-05,
+      "loss": 0.159,
+      "step": 2070
+    },
+    {
+      "epoch": 0.033136,
+      "grad_norm": 1.4140625,
+      "learning_rate": 9.746774193548388e-05,
+      "loss": 0.2019,
+      "step": 2071
+    },
+    {
+      "epoch": 0.033152,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.746612903225807e-05,
+      "loss": 0.2047,
+      "step": 2072
+    },
+    {
+      "epoch": 0.033168,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.746451612903225e-05,
+      "loss": 0.1579,
+      "step": 2073
+    },
+    {
+      "epoch": 0.033184,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.746290322580645e-05,
+      "loss": 0.1315,
+      "step": 2074
+    },
+    {
+      "epoch": 0.0332,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.746129032258064e-05,
+      "loss": 0.2163,
+      "step": 2075
+    },
+    {
+      "epoch": 0.033216,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.745967741935484e-05,
+      "loss": 0.145,
+      "step": 2076
+    },
+    {
+      "epoch": 0.033232,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.745806451612904e-05,
+      "loss": 0.1976,
+      "step": 2077
+    },
+    {
+      "epoch": 0.033248,
+      "grad_norm": 1.125,
+      "learning_rate": 9.745645161290323e-05,
+      "loss": 0.2074,
+      "step": 2078
+    },
+    {
+      "epoch": 0.033264,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.745483870967743e-05,
+      "loss": 0.224,
+      "step": 2079
+    },
+    {
+      "epoch": 0.03328,
+      "grad_norm": 1.3828125,
+      "learning_rate": 9.745322580645162e-05,
+      "loss": 0.235,
+      "step": 2080
+    },
+    {
+      "epoch": 0.033296,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.745161290322581e-05,
+      "loss": 0.227,
+      "step": 2081
+    },
+    {
+      "epoch": 0.033312,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.745000000000001e-05,
+      "loss": 0.1771,
+      "step": 2082
+    },
+    {
+      "epoch": 0.033328,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.74483870967742e-05,
+      "loss": 0.2028,
+      "step": 2083
+    },
+    {
+      "epoch": 0.033344,
+      "grad_norm": 1.296875,
+      "learning_rate": 9.74467741935484e-05,
+      "loss": 0.1773,
+      "step": 2084
+    },
+    {
+      "epoch": 0.03336,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.744516129032258e-05,
+      "loss": 0.1975,
+      "step": 2085
+    },
+    {
+      "epoch": 0.033376,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.744354838709678e-05,
+      "loss": 0.1905,
+      "step": 2086
+    },
+    {
+      "epoch": 0.033392,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.744193548387097e-05,
+      "loss": 0.1974,
+      "step": 2087
+    },
+    {
+      "epoch": 0.033408,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.744032258064517e-05,
+      "loss": 0.1651,
+      "step": 2088
+    },
+    {
+      "epoch": 0.033424,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.743870967741935e-05,
+      "loss": 0.2508,
+      "step": 2089
+    },
+    {
+      "epoch": 0.03344,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.743709677419355e-05,
+      "loss": 0.1808,
+      "step": 2090
+    },
+    {
+      "epoch": 0.033456,
+      "grad_norm": 1.8046875,
+      "learning_rate": 9.743548387096775e-05,
+      "loss": 0.2281,
+      "step": 2091
+    },
+    {
+      "epoch": 0.033472,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.743387096774194e-05,
+      "loss": 0.1717,
+      "step": 2092
+    },
+    {
+      "epoch": 0.033488,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.743225806451614e-05,
+      "loss": 0.2016,
+      "step": 2093
+    },
+    {
+      "epoch": 0.033504,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.743064516129032e-05,
+      "loss": 0.2019,
+      "step": 2094
+    },
+    {
+      "epoch": 0.03352,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.742903225806452e-05,
+      "loss": 0.1806,
+      "step": 2095
+    },
+    {
+      "epoch": 0.033536,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.742741935483871e-05,
+      "loss": 0.1484,
+      "step": 2096
+    },
+    {
+      "epoch": 0.033552,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.742580645161291e-05,
+      "loss": 0.1894,
+      "step": 2097
+    },
+    {
+      "epoch": 0.033568,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.74241935483871e-05,
+      "loss": 0.1902,
+      "step": 2098
+    },
+    {
+      "epoch": 0.033584,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.74225806451613e-05,
+      "loss": 0.1555,
+      "step": 2099
+    },
+    {
+      "epoch": 0.0336,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.742096774193548e-05,
+      "loss": 0.1976,
+      "step": 2100
+    },
+    {
+      "epoch": 0.033616,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.741935483870968e-05,
+      "loss": 0.2096,
+      "step": 2101
+    },
+    {
+      "epoch": 0.033632,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.741774193548388e-05,
+      "loss": 0.2353,
+      "step": 2102
+    },
+    {
+      "epoch": 0.033648,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.741612903225808e-05,
+      "loss": 0.2078,
+      "step": 2103
+    },
+    {
+      "epoch": 0.033664,
+      "grad_norm": 1.4609375,
+      "learning_rate": 9.741451612903227e-05,
+      "loss": 0.2134,
+      "step": 2104
+    },
+    {
+      "epoch": 0.03368,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.741290322580645e-05,
+      "loss": 0.2218,
+      "step": 2105
+    },
+    {
+      "epoch": 0.033696,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.741129032258065e-05,
+      "loss": 0.2071,
+      "step": 2106
+    },
+    {
+      "epoch": 0.033712,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.740967741935484e-05,
+      "loss": 0.1955,
+      "step": 2107
+    },
+    {
+      "epoch": 0.033728,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.740806451612904e-05,
+      "loss": 0.2102,
+      "step": 2108
+    },
+    {
+      "epoch": 0.033744,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.740645161290322e-05,
+      "loss": 0.1681,
+      "step": 2109
+    },
+    {
+      "epoch": 0.03376,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.740483870967742e-05,
+      "loss": 0.2117,
+      "step": 2110
+    },
+    {
+      "epoch": 0.033776,
+      "grad_norm": 0.875,
+      "learning_rate": 9.740322580645161e-05,
+      "loss": 0.2132,
+      "step": 2111
+    },
+    {
+      "epoch": 0.033792,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.740161290322581e-05,
+      "loss": 0.1871,
+      "step": 2112
+    },
+    {
+      "epoch": 0.033808,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.74e-05,
+      "loss": 0.1721,
+      "step": 2113
+    },
+    {
+      "epoch": 0.033824,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.73983870967742e-05,
+      "loss": 0.1701,
+      "step": 2114
+    },
+    {
+      "epoch": 0.03384,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.73967741935484e-05,
+      "loss": 0.2045,
+      "step": 2115
+    },
+    {
+      "epoch": 0.033856,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.73951612903226e-05,
+      "loss": 0.1778,
+      "step": 2116
+    },
+    {
+      "epoch": 0.033872,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.739354838709678e-05,
+      "loss": 0.1597,
+      "step": 2117
+    },
+    {
+      "epoch": 0.033888,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.739193548387098e-05,
+      "loss": 0.1756,
+      "step": 2118
+    },
+    {
+      "epoch": 0.033904,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.739032258064517e-05,
+      "loss": 0.2152,
+      "step": 2119
+    },
+    {
+      "epoch": 0.03392,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.738870967741935e-05,
+      "loss": 0.2031,
+      "step": 2120
+    },
+    {
+      "epoch": 0.033936,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.738709677419355e-05,
+      "loss": 0.181,
+      "step": 2121
+    },
+    {
+      "epoch": 0.033952,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.738548387096774e-05,
+      "loss": 0.183,
+      "step": 2122
+    },
+    {
+      "epoch": 0.033968,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.738387096774194e-05,
+      "loss": 0.1897,
+      "step": 2123
+    },
+    {
+      "epoch": 0.033984,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.738225806451612e-05,
+      "loss": 0.1644,
+      "step": 2124
+    },
+    {
+      "epoch": 0.034,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.738064516129032e-05,
+      "loss": 0.1856,
+      "step": 2125
+    },
+    {
+      "epoch": 0.034016,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.737903225806452e-05,
+      "loss": 0.2074,
+      "step": 2126
+    },
+    {
+      "epoch": 0.034032,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.737741935483872e-05,
+      "loss": 0.1749,
+      "step": 2127
+    },
+    {
+      "epoch": 0.034048,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.737580645161291e-05,
+      "loss": 0.1727,
+      "step": 2128
+    },
+    {
+      "epoch": 0.034064,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.737419354838711e-05,
+      "loss": 0.1391,
+      "step": 2129
+    },
+    {
+      "epoch": 0.03408,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.73725806451613e-05,
+      "loss": 0.1931,
+      "step": 2130
+    },
+    {
+      "epoch": 0.034096,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.73709677419355e-05,
+      "loss": 0.1804,
+      "step": 2131
+    },
+    {
+      "epoch": 0.034112,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.736935483870968e-05,
+      "loss": 0.1706,
+      "step": 2132
+    },
+    {
+      "epoch": 0.034128,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.736774193548388e-05,
+      "loss": 0.1769,
+      "step": 2133
+    },
+    {
+      "epoch": 0.034144,
+      "grad_norm": 1.4140625,
+      "learning_rate": 9.736612903225807e-05,
+      "loss": 0.2241,
+      "step": 2134
+    },
+    {
+      "epoch": 0.03416,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.736451612903227e-05,
+      "loss": 0.1934,
+      "step": 2135
+    },
+    {
+      "epoch": 0.034176,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.736290322580645e-05,
+      "loss": 0.2069,
+      "step": 2136
+    },
+    {
+      "epoch": 0.034192,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.736129032258065e-05,
+      "loss": 0.2068,
+      "step": 2137
+    },
+    {
+      "epoch": 0.034208,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.735967741935485e-05,
+      "loss": 0.2377,
+      "step": 2138
+    },
+    {
+      "epoch": 0.034224,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.735806451612904e-05,
+      "loss": 0.1862,
+      "step": 2139
+    },
+    {
+      "epoch": 0.03424,
+      "grad_norm": 1.4296875,
+      "learning_rate": 9.735645161290324e-05,
+      "loss": 0.2024,
+      "step": 2140
+    },
+    {
+      "epoch": 0.034256,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.735483870967742e-05,
+      "loss": 0.2032,
+      "step": 2141
+    },
+    {
+      "epoch": 0.034272,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.735322580645162e-05,
+      "loss": 0.1952,
+      "step": 2142
+    },
+    {
+      "epoch": 0.034288,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.735161290322581e-05,
+      "loss": 0.2133,
+      "step": 2143
+    },
+    {
+      "epoch": 0.034304,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.735000000000001e-05,
+      "loss": 0.1747,
+      "step": 2144
+    },
+    {
+      "epoch": 0.03432,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.73483870967742e-05,
+      "loss": 0.2327,
+      "step": 2145
+    },
+    {
+      "epoch": 0.034336,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.73467741935484e-05,
+      "loss": 0.2076,
+      "step": 2146
+    },
+    {
+      "epoch": 0.034352,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.734516129032258e-05,
+      "loss": 0.1872,
+      "step": 2147
+    },
+    {
+      "epoch": 0.034368,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.734354838709678e-05,
+      "loss": 0.1792,
+      "step": 2148
+    },
+    {
+      "epoch": 0.034384,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.734193548387097e-05,
+      "loss": 0.2065,
+      "step": 2149
+    },
+    {
+      "epoch": 0.0344,
+      "grad_norm": 1.390625,
+      "learning_rate": 9.734032258064517e-05,
+      "loss": 0.216,
+      "step": 2150
+    },
+    {
+      "epoch": 0.034416,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.733870967741936e-05,
+      "loss": 0.2445,
+      "step": 2151
+    },
+    {
+      "epoch": 0.034432,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.733709677419355e-05,
+      "loss": 0.1546,
+      "step": 2152
+    },
+    {
+      "epoch": 0.034448,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.733548387096775e-05,
+      "loss": 0.1913,
+      "step": 2153
+    },
+    {
+      "epoch": 0.034464,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.733387096774194e-05,
+      "loss": 0.2102,
+      "step": 2154
+    },
+    {
+      "epoch": 0.03448,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.733225806451614e-05,
+      "loss": 0.2222,
+      "step": 2155
+    },
+    {
+      "epoch": 0.034496,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.733064516129032e-05,
+      "loss": 0.1946,
+      "step": 2156
+    },
+    {
+      "epoch": 0.034512,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.732903225806452e-05,
+      "loss": 0.1637,
+      "step": 2157
+    },
+    {
+      "epoch": 0.034528,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.732741935483871e-05,
+      "loss": 0.1907,
+      "step": 2158
+    },
+    {
+      "epoch": 0.034544,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.732580645161291e-05,
+      "loss": 0.1863,
+      "step": 2159
+    },
+    {
+      "epoch": 0.03456,
+      "grad_norm": 1.0,
+      "learning_rate": 9.73241935483871e-05,
+      "loss": 0.215,
+      "step": 2160
+    },
+    {
+      "epoch": 0.034576,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.73225806451613e-05,
+      "loss": 0.203,
+      "step": 2161
+    },
+    {
+      "epoch": 0.034592,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.732096774193549e-05,
+      "loss": 0.1661,
+      "step": 2162
+    },
+    {
+      "epoch": 0.034608,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.731935483870969e-05,
+      "loss": 0.2275,
+      "step": 2163
+    },
+    {
+      "epoch": 0.034624,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.731774193548388e-05,
+      "loss": 0.1818,
+      "step": 2164
+    },
+    {
+      "epoch": 0.03464,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.731612903225808e-05,
+      "loss": 0.2012,
+      "step": 2165
+    },
+    {
+      "epoch": 0.034656,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.731451612903226e-05,
+      "loss": 0.1978,
+      "step": 2166
+    },
+    {
+      "epoch": 0.034672,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.731290322580645e-05,
+      "loss": 0.2705,
+      "step": 2167
+    },
+    {
+      "epoch": 0.034688,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.731129032258065e-05,
+      "loss": 0.2219,
+      "step": 2168
+    },
+    {
+      "epoch": 0.034704,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.730967741935484e-05,
+      "loss": 0.1719,
+      "step": 2169
+    },
+    {
+      "epoch": 0.03472,
+      "grad_norm": 1.3984375,
+      "learning_rate": 9.730806451612904e-05,
+      "loss": 0.1959,
+      "step": 2170
+    },
+    {
+      "epoch": 0.034736,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.730645161290322e-05,
+      "loss": 0.1889,
+      "step": 2171
+    },
+    {
+      "epoch": 0.034752,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.730483870967742e-05,
+      "loss": 0.1756,
+      "step": 2172
+    },
+    {
+      "epoch": 0.034768,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.730322580645162e-05,
+      "loss": 0.1685,
+      "step": 2173
+    },
+    {
+      "epoch": 0.034784,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.730161290322581e-05,
+      "loss": 0.1684,
+      "step": 2174
+    },
+    {
+      "epoch": 0.0348,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.730000000000001e-05,
+      "loss": 0.1682,
+      "step": 2175
+    },
+    {
+      "epoch": 0.034816,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.72983870967742e-05,
+      "loss": 0.1707,
+      "step": 2176
+    },
+    {
+      "epoch": 0.034832,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.729677419354839e-05,
+      "loss": 0.1866,
+      "step": 2177
+    },
+    {
+      "epoch": 0.034848,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.729516129032259e-05,
+      "loss": 0.1933,
+      "step": 2178
+    },
+    {
+      "epoch": 0.034864,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.729354838709678e-05,
+      "loss": 0.1845,
+      "step": 2179
+    },
+    {
+      "epoch": 0.03488,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.729193548387098e-05,
+      "loss": 0.2145,
+      "step": 2180
+    },
+    {
+      "epoch": 0.034896,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.729032258064516e-05,
+      "loss": 0.1553,
+      "step": 2181
+    },
+    {
+      "epoch": 0.034912,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.728870967741935e-05,
+      "loss": 0.2649,
+      "step": 2182
+    },
+    {
+      "epoch": 0.034928,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.728709677419355e-05,
+      "loss": 0.2054,
+      "step": 2183
+    },
+    {
+      "epoch": 0.034944,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.728548387096774e-05,
+      "loss": 0.2313,
+      "step": 2184
+    },
+    {
+      "epoch": 0.03496,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.728387096774194e-05,
+      "loss": 0.16,
+      "step": 2185
+    },
+    {
+      "epoch": 0.034976,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.728225806451614e-05,
+      "loss": 0.2665,
+      "step": 2186
+    },
+    {
+      "epoch": 0.034992,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.728064516129033e-05,
+      "loss": 0.214,
+      "step": 2187
+    },
+    {
+      "epoch": 0.035008,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.727903225806452e-05,
+      "loss": 0.1635,
+      "step": 2188
+    },
+    {
+      "epoch": 0.035024,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.727741935483872e-05,
+      "loss": 0.1924,
+      "step": 2189
+    },
+    {
+      "epoch": 0.03504,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.72758064516129e-05,
+      "loss": 0.1911,
+      "step": 2190
+    },
+    {
+      "epoch": 0.035056,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.72741935483871e-05,
+      "loss": 0.1834,
+      "step": 2191
+    },
+    {
+      "epoch": 0.035072,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.727258064516129e-05,
+      "loss": 0.2036,
+      "step": 2192
+    },
+    {
+      "epoch": 0.035088,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.727096774193549e-05,
+      "loss": 0.1809,
+      "step": 2193
+    },
+    {
+      "epoch": 0.035104,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.726935483870968e-05,
+      "loss": 0.2205,
+      "step": 2194
+    },
+    {
+      "epoch": 0.03512,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.726774193548388e-05,
+      "loss": 0.1807,
+      "step": 2195
+    },
+    {
+      "epoch": 0.035136,
+      "grad_norm": 1.515625,
+      "learning_rate": 9.726612903225806e-05,
+      "loss": 0.2638,
+      "step": 2196
+    },
+    {
+      "epoch": 0.035152,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.726451612903226e-05,
+      "loss": 0.1847,
+      "step": 2197
+    },
+    {
+      "epoch": 0.035168,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.726290322580646e-05,
+      "loss": 0.2419,
+      "step": 2198
+    },
+    {
+      "epoch": 0.035184,
+      "grad_norm": 0.875,
+      "learning_rate": 9.726129032258065e-05,
+      "loss": 0.1964,
+      "step": 2199
+    },
+    {
+      "epoch": 0.0352,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.725967741935485e-05,
+      "loss": 0.2166,
+      "step": 2200
+    },
+    {
+      "epoch": 0.035216,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.725806451612903e-05,
+      "loss": 0.1708,
+      "step": 2201
+    },
+    {
+      "epoch": 0.035232,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.725645161290323e-05,
+      "loss": 0.1732,
+      "step": 2202
+    },
+    {
+      "epoch": 0.035248,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.725483870967742e-05,
+      "loss": 0.2056,
+      "step": 2203
+    },
+    {
+      "epoch": 0.035264,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.725322580645162e-05,
+      "loss": 0.1935,
+      "step": 2204
+    },
+    {
+      "epoch": 0.03528,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.72516129032258e-05,
+      "loss": 0.1703,
+      "step": 2205
+    },
+    {
+      "epoch": 0.035296,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.725e-05,
+      "loss": 0.1744,
+      "step": 2206
+    },
+    {
+      "epoch": 0.035312,
+      "grad_norm": 1.4296875,
+      "learning_rate": 9.724838709677419e-05,
+      "loss": 0.2041,
+      "step": 2207
+    },
+    {
+      "epoch": 0.035328,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.724677419354839e-05,
+      "loss": 0.2213,
+      "step": 2208
+    },
+    {
+      "epoch": 0.035344,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.724516129032258e-05,
+      "loss": 0.1872,
+      "step": 2209
+    },
+    {
+      "epoch": 0.03536,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.724354838709678e-05,
+      "loss": 0.2099,
+      "step": 2210
+    },
+    {
+      "epoch": 0.035376,
+      "grad_norm": 1.3984375,
+      "learning_rate": 9.724193548387098e-05,
+      "loss": 0.2435,
+      "step": 2211
+    },
+    {
+      "epoch": 0.035392,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.724032258064518e-05,
+      "loss": 0.2296,
+      "step": 2212
+    },
+    {
+      "epoch": 0.035408,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.723870967741936e-05,
+      "loss": 0.1899,
+      "step": 2213
+    },
+    {
+      "epoch": 0.035424,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.723709677419355e-05,
+      "loss": 0.161,
+      "step": 2214
+    },
+    {
+      "epoch": 0.03544,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.723548387096775e-05,
+      "loss": 0.2,
+      "step": 2215
+    },
+    {
+      "epoch": 0.035456,
+      "grad_norm": 0.875,
+      "learning_rate": 9.723387096774193e-05,
+      "loss": 0.1617,
+      "step": 2216
+    },
+    {
+      "epoch": 0.035472,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.723225806451613e-05,
+      "loss": 0.2294,
+      "step": 2217
+    },
+    {
+      "epoch": 0.035488,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.723064516129032e-05,
+      "loss": 0.2131,
+      "step": 2218
+    },
+    {
+      "epoch": 0.035504,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.722903225806452e-05,
+      "loss": 0.1908,
+      "step": 2219
+    },
+    {
+      "epoch": 0.03552,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.72274193548387e-05,
+      "loss": 0.1713,
+      "step": 2220
+    },
+    {
+      "epoch": 0.035536,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.72258064516129e-05,
+      "loss": 0.2235,
+      "step": 2221
+    },
+    {
+      "epoch": 0.035552,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.72241935483871e-05,
+      "loss": 0.1862,
+      "step": 2222
+    },
+    {
+      "epoch": 0.035568,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.72225806451613e-05,
+      "loss": 0.1747,
+      "step": 2223
+    },
+    {
+      "epoch": 0.035584,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.722096774193549e-05,
+      "loss": 0.2023,
+      "step": 2224
+    },
+    {
+      "epoch": 0.0356,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.721935483870969e-05,
+      "loss": 0.1633,
+      "step": 2225
+    },
+    {
+      "epoch": 0.035616,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.721774193548388e-05,
+      "loss": 0.1629,
+      "step": 2226
+    },
+    {
+      "epoch": 0.035632,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.721612903225808e-05,
+      "loss": 0.2071,
+      "step": 2227
+    },
+    {
+      "epoch": 0.035648,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.721451612903226e-05,
+      "loss": 0.1783,
+      "step": 2228
+    },
+    {
+      "epoch": 0.035664,
+      "grad_norm": 0.875,
+      "learning_rate": 9.721290322580645e-05,
+      "loss": 0.2286,
+      "step": 2229
+    },
+    {
+      "epoch": 0.03568,
+      "grad_norm": 1.4453125,
+      "learning_rate": 9.721129032258065e-05,
+      "loss": 0.2056,
+      "step": 2230
+    },
+    {
+      "epoch": 0.035696,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.720967741935483e-05,
+      "loss": 0.1879,
+      "step": 2231
+    },
+    {
+      "epoch": 0.035712,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.720806451612903e-05,
+      "loss": 0.1975,
+      "step": 2232
+    },
+    {
+      "epoch": 0.035728,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.720645161290323e-05,
+      "loss": 0.1476,
+      "step": 2233
+    },
+    {
+      "epoch": 0.035744,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.720483870967743e-05,
+      "loss": 0.2446,
+      "step": 2234
+    },
+    {
+      "epoch": 0.03576,
+      "grad_norm": 1.4609375,
+      "learning_rate": 9.720322580645162e-05,
+      "loss": 0.2535,
+      "step": 2235
+    },
+    {
+      "epoch": 0.035776,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.720161290322582e-05,
+      "loss": 0.2031,
+      "step": 2236
+    },
+    {
+      "epoch": 0.035792,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.72e-05,
+      "loss": 0.1985,
+      "step": 2237
+    },
+    {
+      "epoch": 0.035808,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.71983870967742e-05,
+      "loss": 0.2132,
+      "step": 2238
+    },
+    {
+      "epoch": 0.035824,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.719677419354839e-05,
+      "loss": 0.2219,
+      "step": 2239
+    },
+    {
+      "epoch": 0.03584,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.719516129032259e-05,
+      "loss": 0.2198,
+      "step": 2240
+    },
+    {
+      "epoch": 0.035856,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.719354838709678e-05,
+      "loss": 0.1651,
+      "step": 2241
+    },
+    {
+      "epoch": 0.035872,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.719193548387098e-05,
+      "loss": 0.1995,
+      "step": 2242
+    },
+    {
+      "epoch": 0.035888,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.719032258064516e-05,
+      "loss": 0.1564,
+      "step": 2243
+    },
+    {
+      "epoch": 0.035904,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.718870967741935e-05,
+      "loss": 0.1715,
+      "step": 2244
+    },
+    {
+      "epoch": 0.03592,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.718709677419355e-05,
+      "loss": 0.1948,
+      "step": 2245
+    },
+    {
+      "epoch": 0.035936,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.718548387096775e-05,
+      "loss": 0.1917,
+      "step": 2246
+    },
+    {
+      "epoch": 0.035952,
+      "grad_norm": 1.5078125,
+      "learning_rate": 9.718387096774195e-05,
+      "loss": 0.2145,
+      "step": 2247
+    },
+    {
+      "epoch": 0.035968,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.718225806451613e-05,
+      "loss": 0.1937,
+      "step": 2248
+    },
+    {
+      "epoch": 0.035984,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.718064516129033e-05,
+      "loss": 0.2006,
+      "step": 2249
+    },
+    {
+      "epoch": 0.036,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.717903225806452e-05,
+      "loss": 0.218,
+      "step": 2250
+    },
+    {
+      "epoch": 0.036016,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.717741935483872e-05,
+      "loss": 0.173,
+      "step": 2251
+    },
+    {
+      "epoch": 0.036032,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.71758064516129e-05,
+      "loss": 0.2018,
+      "step": 2252
+    },
+    {
+      "epoch": 0.036048,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.71741935483871e-05,
+      "loss": 0.1463,
+      "step": 2253
+    },
+    {
+      "epoch": 0.036064,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.717258064516129e-05,
+      "loss": 0.1746,
+      "step": 2254
+    },
+    {
+      "epoch": 0.03608,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.717096774193549e-05,
+      "loss": 0.195,
+      "step": 2255
+    },
+    {
+      "epoch": 0.036096,
+      "grad_norm": 1.0,
+      "learning_rate": 9.716935483870968e-05,
+      "loss": 0.1875,
+      "step": 2256
+    },
+    {
+      "epoch": 0.036112,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.716774193548388e-05,
+      "loss": 0.1831,
+      "step": 2257
+    },
+    {
+      "epoch": 0.036128,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.716612903225807e-05,
+      "loss": 0.1994,
+      "step": 2258
+    },
+    {
+      "epoch": 0.036144,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.716451612903227e-05,
+      "loss": 0.1751,
+      "step": 2259
+    },
+    {
+      "epoch": 0.03616,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.716290322580646e-05,
+      "loss": 0.1838,
+      "step": 2260
+    },
+    {
+      "epoch": 0.036176,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.716129032258065e-05,
+      "loss": 0.1447,
+      "step": 2261
+    },
+    {
+      "epoch": 0.036192,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.715967741935485e-05,
+      "loss": 0.1981,
+      "step": 2262
+    },
+    {
+      "epoch": 0.036208,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.715806451612903e-05,
+      "loss": 0.2462,
+      "step": 2263
+    },
+    {
+      "epoch": 0.036224,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.715645161290323e-05,
+      "loss": 0.1475,
+      "step": 2264
+    },
+    {
+      "epoch": 0.03624,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.715483870967742e-05,
+      "loss": 0.2098,
+      "step": 2265
+    },
+    {
+      "epoch": 0.036256,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.715322580645162e-05,
+      "loss": 0.2352,
+      "step": 2266
+    },
+    {
+      "epoch": 0.036272,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.71516129032258e-05,
+      "loss": 0.205,
+      "step": 2267
+    },
+    {
+      "epoch": 0.036288,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.715e-05,
+      "loss": 0.2139,
+      "step": 2268
+    },
+    {
+      "epoch": 0.036304,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.714838709677419e-05,
+      "loss": 0.1875,
+      "step": 2269
+    },
+    {
+      "epoch": 0.03632,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.714677419354839e-05,
+      "loss": 0.1849,
+      "step": 2270
+    },
+    {
+      "epoch": 0.036336,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.714516129032259e-05,
+      "loss": 0.2253,
+      "step": 2271
+    },
+    {
+      "epoch": 0.036352,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.714354838709679e-05,
+      "loss": 0.1904,
+      "step": 2272
+    },
+    {
+      "epoch": 0.036368,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.714193548387097e-05,
+      "loss": 0.1877,
+      "step": 2273
+    },
+    {
+      "epoch": 0.036384,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.714032258064517e-05,
+      "loss": 0.1497,
+      "step": 2274
+    },
+    {
+      "epoch": 0.0364,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.713870967741936e-05,
+      "loss": 0.1804,
+      "step": 2275
+    },
+    {
+      "epoch": 0.036416,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.713709677419355e-05,
+      "loss": 0.2069,
+      "step": 2276
+    },
+    {
+      "epoch": 0.036432,
+      "grad_norm": 1.3984375,
+      "learning_rate": 9.713548387096775e-05,
+      "loss": 0.2147,
+      "step": 2277
+    },
+    {
+      "epoch": 0.036448,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.713387096774193e-05,
+      "loss": 0.2043,
+      "step": 2278
+    },
+    {
+      "epoch": 0.036464,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.713225806451613e-05,
+      "loss": 0.1765,
+      "step": 2279
+    },
+    {
+      "epoch": 0.03648,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.713064516129032e-05,
+      "loss": 0.1953,
+      "step": 2280
+    },
+    {
+      "epoch": 0.036496,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.712903225806452e-05,
+      "loss": 0.1504,
+      "step": 2281
+    },
+    {
+      "epoch": 0.036512,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.712741935483872e-05,
+      "loss": 0.1993,
+      "step": 2282
+    },
+    {
+      "epoch": 0.036528,
+      "grad_norm": 1.515625,
+      "learning_rate": 9.712580645161292e-05,
+      "loss": 0.2225,
+      "step": 2283
+    },
+    {
+      "epoch": 0.036544,
+      "grad_norm": 1.5,
+      "learning_rate": 9.71241935483871e-05,
+      "loss": 0.1992,
+      "step": 2284
+    },
+    {
+      "epoch": 0.03656,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.71225806451613e-05,
+      "loss": 0.1778,
+      "step": 2285
+    },
+    {
+      "epoch": 0.036576,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.712096774193549e-05,
+      "loss": 0.2085,
+      "step": 2286
+    },
+    {
+      "epoch": 0.036592,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.711935483870969e-05,
+      "loss": 0.1665,
+      "step": 2287
+    },
+    {
+      "epoch": 0.036608,
+      "grad_norm": 1.3125,
+      "learning_rate": 9.711774193548387e-05,
+      "loss": 0.2373,
+      "step": 2288
+    },
+    {
+      "epoch": 0.036624,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.711612903225807e-05,
+      "loss": 0.1942,
+      "step": 2289
+    },
+    {
+      "epoch": 0.03664,
+      "grad_norm": 1.3359375,
+      "learning_rate": 9.711451612903226e-05,
+      "loss": 0.1829,
+      "step": 2290
+    },
+    {
+      "epoch": 0.036656,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.711290322580645e-05,
+      "loss": 0.2259,
+      "step": 2291
+    },
+    {
+      "epoch": 0.036672,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.711129032258065e-05,
+      "loss": 0.1934,
+      "step": 2292
+    },
+    {
+      "epoch": 0.036688,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.710967741935484e-05,
+      "loss": 0.2188,
+      "step": 2293
+    },
+    {
+      "epoch": 0.036704,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.710806451612904e-05,
+      "loss": 0.191,
+      "step": 2294
+    },
+    {
+      "epoch": 0.03672,
+      "grad_norm": 1.484375,
+      "learning_rate": 9.710645161290323e-05,
+      "loss": 0.1769,
+      "step": 2295
+    },
+    {
+      "epoch": 0.036736,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.710483870967743e-05,
+      "loss": 0.1947,
+      "step": 2296
+    },
+    {
+      "epoch": 0.036752,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.710322580645162e-05,
+      "loss": 0.1791,
+      "step": 2297
+    },
+    {
+      "epoch": 0.036768,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.710161290322582e-05,
+      "loss": 0.2474,
+      "step": 2298
+    },
+    {
+      "epoch": 0.036784,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.71e-05,
+      "loss": 0.1998,
+      "step": 2299
+    },
+    {
+      "epoch": 0.0368,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.70983870967742e-05,
+      "loss": 0.202,
+      "step": 2300
+    },
+    {
+      "epoch": 0.036816,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.709677419354839e-05,
+      "loss": 0.1826,
+      "step": 2301
+    },
+    {
+      "epoch": 0.036832,
+      "grad_norm": 1.4921875,
+      "learning_rate": 9.709516129032259e-05,
+      "loss": 0.193,
+      "step": 2302
+    },
+    {
+      "epoch": 0.036848,
+      "grad_norm": 1.359375,
+      "learning_rate": 9.709354838709677e-05,
+      "loss": 0.1974,
+      "step": 2303
+    },
+    {
+      "epoch": 0.036864,
+      "grad_norm": 1.8203125,
+      "learning_rate": 9.709193548387097e-05,
+      "loss": 0.1927,
+      "step": 2304
+    },
+    {
+      "epoch": 0.03688,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.709032258064516e-05,
+      "loss": 0.1981,
+      "step": 2305
+    },
+    {
+      "epoch": 0.036896,
+      "grad_norm": 1.84375,
+      "learning_rate": 9.708870967741936e-05,
+      "loss": 0.2049,
+      "step": 2306
+    },
+    {
+      "epoch": 0.036912,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.708709677419356e-05,
+      "loss": 0.1887,
+      "step": 2307
+    },
+    {
+      "epoch": 0.036928,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.708548387096774e-05,
+      "loss": 0.189,
+      "step": 2308
+    },
+    {
+      "epoch": 0.036944,
+      "grad_norm": 1.5703125,
+      "learning_rate": 9.708387096774194e-05,
+      "loss": 0.1589,
+      "step": 2309
+    },
+    {
+      "epoch": 0.03696,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.708225806451613e-05,
+      "loss": 0.1668,
+      "step": 2310
+    },
+    {
+      "epoch": 0.036976,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.708064516129033e-05,
+      "loss": 0.2271,
+      "step": 2311
+    },
+    {
+      "epoch": 0.036992,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.707903225806452e-05,
+      "loss": 0.1996,
+      "step": 2312
+    },
+    {
+      "epoch": 0.037008,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.707741935483872e-05,
+      "loss": 0.2104,
+      "step": 2313
+    },
+    {
+      "epoch": 0.037024,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.70758064516129e-05,
+      "loss": 0.1646,
+      "step": 2314
+    },
+    {
+      "epoch": 0.03704,
+      "grad_norm": 1.125,
+      "learning_rate": 9.70741935483871e-05,
+      "loss": 0.1978,
+      "step": 2315
+    },
+    {
+      "epoch": 0.037056,
+      "grad_norm": 0.75,
+      "learning_rate": 9.707258064516129e-05,
+      "loss": 0.1408,
+      "step": 2316
+    },
+    {
+      "epoch": 0.037072,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.707096774193549e-05,
+      "loss": 0.1937,
+      "step": 2317
+    },
+    {
+      "epoch": 0.037088,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.706935483870969e-05,
+      "loss": 0.2319,
+      "step": 2318
+    },
+    {
+      "epoch": 0.037104,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.706774193548389e-05,
+      "loss": 0.2055,
+      "step": 2319
+    },
+    {
+      "epoch": 0.03712,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.706612903225807e-05,
+      "loss": 0.1895,
+      "step": 2320
+    },
+    {
+      "epoch": 0.037136,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.706451612903227e-05,
+      "loss": 0.1971,
+      "step": 2321
+    },
+    {
+      "epoch": 0.037152,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.706290322580646e-05,
+      "loss": 0.2147,
+      "step": 2322
+    },
+    {
+      "epoch": 0.037168,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.706129032258064e-05,
+      "loss": 0.1723,
+      "step": 2323
+    },
+    {
+      "epoch": 0.037184,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.705967741935484e-05,
+      "loss": 0.1979,
+      "step": 2324
+    },
+    {
+      "epoch": 0.0372,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.705806451612903e-05,
+      "loss": 0.1993,
+      "step": 2325
+    },
+    {
+      "epoch": 0.037216,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.705645161290323e-05,
+      "loss": 0.2292,
+      "step": 2326
+    },
+    {
+      "epoch": 0.037232,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.705483870967742e-05,
+      "loss": 0.1966,
+      "step": 2327
+    },
+    {
+      "epoch": 0.037248,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.705322580645162e-05,
+      "loss": 0.2055,
+      "step": 2328
+    },
+    {
+      "epoch": 0.037264,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.705161290322581e-05,
+      "loss": 0.1769,
+      "step": 2329
+    },
+    {
+      "epoch": 0.03728,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.705e-05,
+      "loss": 0.2029,
+      "step": 2330
+    },
+    {
+      "epoch": 0.037296,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.70483870967742e-05,
+      "loss": 0.2335,
+      "step": 2331
+    },
+    {
+      "epoch": 0.037312,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.70467741935484e-05,
+      "loss": 0.2112,
+      "step": 2332
+    },
+    {
+      "epoch": 0.037328,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.704516129032259e-05,
+      "loss": 0.1765,
+      "step": 2333
+    },
+    {
+      "epoch": 0.037344,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.704354838709679e-05,
+      "loss": 0.2104,
+      "step": 2334
+    },
+    {
+      "epoch": 0.03736,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.704193548387097e-05,
+      "loss": 0.2029,
+      "step": 2335
+    },
+    {
+      "epoch": 0.037376,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.704032258064517e-05,
+      "loss": 0.1834,
+      "step": 2336
+    },
+    {
+      "epoch": 0.037392,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.703870967741936e-05,
+      "loss": 0.206,
+      "step": 2337
+    },
+    {
+      "epoch": 0.037408,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.703709677419354e-05,
+      "loss": 0.1597,
+      "step": 2338
+    },
+    {
+      "epoch": 0.037424,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.703548387096774e-05,
+      "loss": 0.2291,
+      "step": 2339
+    },
+    {
+      "epoch": 0.03744,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.703387096774193e-05,
+      "loss": 0.2167,
+      "step": 2340
+    },
+    {
+      "epoch": 0.037456,
+      "grad_norm": 1.4296875,
+      "learning_rate": 9.703225806451613e-05,
+      "loss": 0.2098,
+      "step": 2341
+    },
+    {
+      "epoch": 0.037472,
+      "grad_norm": 1.3984375,
+      "learning_rate": 9.703064516129033e-05,
+      "loss": 0.2172,
+      "step": 2342
+    },
+    {
+      "epoch": 0.037488,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.702903225806453e-05,
+      "loss": 0.2302,
+      "step": 2343
+    },
+    {
+      "epoch": 0.037504,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.702741935483871e-05,
+      "loss": 0.2134,
+      "step": 2344
+    },
+    {
+      "epoch": 0.03752,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.702580645161291e-05,
+      "loss": 0.2139,
+      "step": 2345
+    },
+    {
+      "epoch": 0.037536,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.70241935483871e-05,
+      "loss": 0.1839,
+      "step": 2346
+    },
+    {
+      "epoch": 0.037552,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.70225806451613e-05,
+      "loss": 0.1653,
+      "step": 2347
+    },
+    {
+      "epoch": 0.037568,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.702096774193549e-05,
+      "loss": 0.2231,
+      "step": 2348
+    },
+    {
+      "epoch": 0.037584,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.701935483870969e-05,
+      "loss": 0.1836,
+      "step": 2349
+    },
+    {
+      "epoch": 0.0376,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.701774193548387e-05,
+      "loss": 0.1653,
+      "step": 2350
+    },
+    {
+      "epoch": 0.037616,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.701612903225807e-05,
+      "loss": 0.1387,
+      "step": 2351
+    },
+    {
+      "epoch": 0.037632,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.701451612903226e-05,
+      "loss": 0.228,
+      "step": 2352
+    },
+    {
+      "epoch": 0.037648,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.701290322580646e-05,
+      "loss": 0.2264,
+      "step": 2353
+    },
+    {
+      "epoch": 0.037664,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.701129032258066e-05,
+      "loss": 0.1898,
+      "step": 2354
+    },
+    {
+      "epoch": 0.03768,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.700967741935484e-05,
+      "loss": 0.1983,
+      "step": 2355
+    },
+    {
+      "epoch": 0.037696,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.700806451612904e-05,
+      "loss": 0.1769,
+      "step": 2356
+    },
+    {
+      "epoch": 0.037712,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.700645161290323e-05,
+      "loss": 0.1821,
+      "step": 2357
+    },
+    {
+      "epoch": 0.037728,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.700483870967743e-05,
+      "loss": 0.1481,
+      "step": 2358
+    },
+    {
+      "epoch": 0.037744,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.700322580645161e-05,
+      "loss": 0.184,
+      "step": 2359
+    },
+    {
+      "epoch": 0.03776,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.700161290322581e-05,
+      "loss": 0.1947,
+      "step": 2360
+    },
+    {
+      "epoch": 0.037776,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.7e-05,
+      "loss": 0.1437,
+      "step": 2361
+    },
+    {
+      "epoch": 0.037792,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.69983870967742e-05,
+      "loss": 0.177,
+      "step": 2362
+    },
+    {
+      "epoch": 0.037808,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.699677419354839e-05,
+      "loss": 0.1834,
+      "step": 2363
+    },
+    {
+      "epoch": 0.037824,
+      "grad_norm": 1.5078125,
+      "learning_rate": 9.699516129032258e-05,
+      "loss": 0.1609,
+      "step": 2364
+    },
+    {
+      "epoch": 0.03784,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.699354838709677e-05,
+      "loss": 0.1757,
+      "step": 2365
+    },
+    {
+      "epoch": 0.037856,
+      "grad_norm": 1.125,
+      "learning_rate": 9.699193548387097e-05,
+      "loss": 0.2176,
+      "step": 2366
+    },
+    {
+      "epoch": 0.037872,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.699032258064517e-05,
+      "loss": 0.1657,
+      "step": 2367
+    },
+    {
+      "epoch": 0.037888,
+      "grad_norm": 1.5078125,
+      "learning_rate": 9.698870967741937e-05,
+      "loss": 0.2199,
+      "step": 2368
+    },
+    {
+      "epoch": 0.037904,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.698709677419356e-05,
+      "loss": 0.186,
+      "step": 2369
+    },
+    {
+      "epoch": 0.03792,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.698548387096774e-05,
+      "loss": 0.1567,
+      "step": 2370
+    },
+    {
+      "epoch": 0.037936,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.698387096774194e-05,
+      "loss": 0.2144,
+      "step": 2371
+    },
+    {
+      "epoch": 0.037952,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.698225806451613e-05,
+      "loss": 0.1709,
+      "step": 2372
+    },
+    {
+      "epoch": 0.037968,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.698064516129033e-05,
+      "loss": 0.2146,
+      "step": 2373
+    },
+    {
+      "epoch": 0.037984,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.697903225806451e-05,
+      "loss": 0.1734,
+      "step": 2374
+    },
+    {
+      "epoch": 0.038,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.697741935483871e-05,
+      "loss": 0.1487,
+      "step": 2375
+    },
+    {
+      "epoch": 0.038016,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.69758064516129e-05,
+      "loss": 0.2259,
+      "step": 2376
+    },
+    {
+      "epoch": 0.038032,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.69741935483871e-05,
+      "loss": 0.2,
+      "step": 2377
+    },
+    {
+      "epoch": 0.038048,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.69725806451613e-05,
+      "loss": 0.2301,
+      "step": 2378
+    },
+    {
+      "epoch": 0.038064,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.69709677419355e-05,
+      "loss": 0.197,
+      "step": 2379
+    },
+    {
+      "epoch": 0.03808,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.696935483870968e-05,
+      "loss": 0.1651,
+      "step": 2380
+    },
+    {
+      "epoch": 0.038096,
+      "grad_norm": 1.3828125,
+      "learning_rate": 9.696774193548388e-05,
+      "loss": 0.2364,
+      "step": 2381
+    },
+    {
+      "epoch": 0.038112,
+      "grad_norm": 1.421875,
+      "learning_rate": 9.696612903225807e-05,
+      "loss": 0.217,
+      "step": 2382
+    },
+    {
+      "epoch": 0.038128,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.696451612903227e-05,
+      "loss": 0.1719,
+      "step": 2383
+    },
+    {
+      "epoch": 0.038144,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.696290322580646e-05,
+      "loss": 0.2079,
+      "step": 2384
+    },
+    {
+      "epoch": 0.03816,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.696129032258064e-05,
+      "loss": 0.1807,
+      "step": 2385
+    },
+    {
+      "epoch": 0.038176,
+      "grad_norm": 0.875,
+      "learning_rate": 9.695967741935484e-05,
+      "loss": 0.1963,
+      "step": 2386
+    },
+    {
+      "epoch": 0.038192,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.695806451612903e-05,
+      "loss": 0.187,
+      "step": 2387
+    },
+    {
+      "epoch": 0.038208,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.695645161290323e-05,
+      "loss": 0.1756,
+      "step": 2388
+    },
+    {
+      "epoch": 0.038224,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.695483870967743e-05,
+      "loss": 0.2218,
+      "step": 2389
+    },
+    {
+      "epoch": 0.03824,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.695322580645163e-05,
+      "loss": 0.1673,
+      "step": 2390
+    },
+    {
+      "epoch": 0.038256,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.695161290322581e-05,
+      "loss": 0.2091,
+      "step": 2391
+    },
+    {
+      "epoch": 0.038272,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.695000000000001e-05,
+      "loss": 0.2008,
+      "step": 2392
+    },
+    {
+      "epoch": 0.038288,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.69483870967742e-05,
+      "loss": 0.2135,
+      "step": 2393
+    },
+    {
+      "epoch": 0.038304,
+      "grad_norm": 1.3359375,
+      "learning_rate": 9.69467741935484e-05,
+      "loss": 0.2573,
+      "step": 2394
+    },
+    {
+      "epoch": 0.03832,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.694516129032258e-05,
+      "loss": 0.2056,
+      "step": 2395
+    },
+    {
+      "epoch": 0.038336,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.694354838709678e-05,
+      "loss": 0.1999,
+      "step": 2396
+    },
+    {
+      "epoch": 0.038352,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.694193548387097e-05,
+      "loss": 0.1666,
+      "step": 2397
+    },
+    {
+      "epoch": 0.038368,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.694032258064517e-05,
+      "loss": 0.1608,
+      "step": 2398
+    },
+    {
+      "epoch": 0.038384,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.693870967741936e-05,
+      "loss": 0.218,
+      "step": 2399
+    },
+    {
+      "epoch": 0.0384,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.693709677419354e-05,
+      "loss": 0.2025,
+      "step": 2400
+    },
+    {
+      "epoch": 0.038416,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.693548387096774e-05,
+      "loss": 0.17,
+      "step": 2401
+    },
+    {
+      "epoch": 0.038432,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.693387096774194e-05,
+      "loss": 0.1785,
+      "step": 2402
+    },
+    {
+      "epoch": 0.038448,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.693225806451614e-05,
+      "loss": 0.1605,
+      "step": 2403
+    },
+    {
+      "epoch": 0.038464,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.693064516129033e-05,
+      "loss": 0.215,
+      "step": 2404
+    },
+    {
+      "epoch": 0.03848,
+      "grad_norm": 1.0,
+      "learning_rate": 9.692903225806453e-05,
+      "loss": 0.2452,
+      "step": 2405
+    },
+    {
+      "epoch": 0.038496,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.692741935483871e-05,
+      "loss": 0.235,
+      "step": 2406
+    },
+    {
+      "epoch": 0.038512,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.692580645161291e-05,
+      "loss": 0.2202,
+      "step": 2407
+    },
+    {
+      "epoch": 0.038528,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.69241935483871e-05,
+      "loss": 0.1465,
+      "step": 2408
+    },
+    {
+      "epoch": 0.038544,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.69225806451613e-05,
+      "loss": 0.1843,
+      "step": 2409
+    },
+    {
+      "epoch": 0.03856,
+      "grad_norm": 1.4140625,
+      "learning_rate": 9.692096774193548e-05,
+      "loss": 0.2352,
+      "step": 2410
+    },
+    {
+      "epoch": 0.038576,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.691935483870968e-05,
+      "loss": 0.1679,
+      "step": 2411
+    },
+    {
+      "epoch": 0.038592,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.691774193548387e-05,
+      "loss": 0.2127,
+      "step": 2412
+    },
+    {
+      "epoch": 0.038608,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.691612903225807e-05,
+      "loss": 0.1917,
+      "step": 2413
+    },
+    {
+      "epoch": 0.038624,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.691451612903227e-05,
+      "loss": 0.1912,
+      "step": 2414
+    },
+    {
+      "epoch": 0.03864,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.691290322580647e-05,
+      "loss": 0.2211,
+      "step": 2415
+    },
+    {
+      "epoch": 0.038656,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.691129032258065e-05,
+      "loss": 0.2227,
+      "step": 2416
+    },
+    {
+      "epoch": 0.038672,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.690967741935484e-05,
+      "loss": 0.2125,
+      "step": 2417
+    },
+    {
+      "epoch": 0.038688,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.690806451612904e-05,
+      "loss": 0.1841,
+      "step": 2418
+    },
+    {
+      "epoch": 0.038704,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.690645161290323e-05,
+      "loss": 0.1758,
+      "step": 2419
+    },
+    {
+      "epoch": 0.03872,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.690483870967743e-05,
+      "loss": 0.2685,
+      "step": 2420
+    },
+    {
+      "epoch": 0.038736,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.690322580645161e-05,
+      "loss": 0.1963,
+      "step": 2421
+    },
+    {
+      "epoch": 0.038752,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.690161290322581e-05,
+      "loss": 0.1757,
+      "step": 2422
+    },
+    {
+      "epoch": 0.038768,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.69e-05,
+      "loss": 0.234,
+      "step": 2423
+    },
+    {
+      "epoch": 0.038784,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.68983870967742e-05,
+      "loss": 0.197,
+      "step": 2424
+    },
+    {
+      "epoch": 0.0388,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.68967741935484e-05,
+      "loss": 0.1818,
+      "step": 2425
+    },
+    {
+      "epoch": 0.038816,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.689516129032258e-05,
+      "loss": 0.1696,
+      "step": 2426
+    },
+    {
+      "epoch": 0.038832,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.689354838709678e-05,
+      "loss": 0.188,
+      "step": 2427
+    },
+    {
+      "epoch": 0.038848,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.689193548387098e-05,
+      "loss": 0.2015,
+      "step": 2428
+    },
+    {
+      "epoch": 0.038864,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.689032258064517e-05,
+      "loss": 0.2412,
+      "step": 2429
+    },
+    {
+      "epoch": 0.03888,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.688870967741937e-05,
+      "loss": 0.1975,
+      "step": 2430
+    },
+    {
+      "epoch": 0.038896,
+      "grad_norm": 1.34375,
+      "learning_rate": 9.688709677419355e-05,
+      "loss": 0.1558,
+      "step": 2431
+    },
+    {
+      "epoch": 0.038912,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.688548387096774e-05,
+      "loss": 0.1684,
+      "step": 2432
+    },
+    {
+      "epoch": 0.038928,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.688387096774194e-05,
+      "loss": 0.2573,
+      "step": 2433
+    },
+    {
+      "epoch": 0.038944,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.688225806451613e-05,
+      "loss": 0.1977,
+      "step": 2434
+    },
+    {
+      "epoch": 0.03896,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.688064516129032e-05,
+      "loss": 0.2011,
+      "step": 2435
+    },
+    {
+      "epoch": 0.038976,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.687903225806451e-05,
+      "loss": 0.2095,
+      "step": 2436
+    },
+    {
+      "epoch": 0.038992,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.687741935483871e-05,
+      "loss": 0.2005,
+      "step": 2437
+    },
+    {
+      "epoch": 0.039008,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.687580645161291e-05,
+      "loss": 0.1827,
+      "step": 2438
+    },
+    {
+      "epoch": 0.039024,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.687419354838711e-05,
+      "loss": 0.2087,
+      "step": 2439
+    },
+    {
+      "epoch": 0.03904,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.68725806451613e-05,
+      "loss": 0.1898,
+      "step": 2440
+    },
+    {
+      "epoch": 0.039056,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.68709677419355e-05,
+      "loss": 0.1988,
+      "step": 2441
+    },
+    {
+      "epoch": 0.039072,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.686935483870968e-05,
+      "loss": 0.2122,
+      "step": 2442
+    },
+    {
+      "epoch": 0.039088,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.686774193548388e-05,
+      "loss": 0.1828,
+      "step": 2443
+    },
+    {
+      "epoch": 0.039104,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.686612903225807e-05,
+      "loss": 0.2342,
+      "step": 2444
+    },
+    {
+      "epoch": 0.03912,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.686451612903227e-05,
+      "loss": 0.1767,
+      "step": 2445
+    },
+    {
+      "epoch": 0.039136,
+      "grad_norm": 1.3125,
+      "learning_rate": 9.686290322580645e-05,
+      "loss": 0.1717,
+      "step": 2446
+    },
+    {
+      "epoch": 0.039152,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.686129032258064e-05,
+      "loss": 0.2081,
+      "step": 2447
+    },
+    {
+      "epoch": 0.039168,
+      "grad_norm": 1.40625,
+      "learning_rate": 9.685967741935484e-05,
+      "loss": 0.155,
+      "step": 2448
+    },
+    {
+      "epoch": 0.039184,
+      "grad_norm": 1.625,
+      "learning_rate": 9.685806451612904e-05,
+      "loss": 0.186,
+      "step": 2449
+    },
+    {
+      "epoch": 0.0392,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.685645161290324e-05,
+      "loss": 0.1998,
+      "step": 2450
+    },
+    {
+      "epoch": 0.039216,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.685483870967742e-05,
+      "loss": 0.2268,
+      "step": 2451
+    },
+    {
+      "epoch": 0.039232,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.685322580645162e-05,
+      "loss": 0.1613,
+      "step": 2452
+    },
+    {
+      "epoch": 0.039248,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.685161290322581e-05,
+      "loss": 0.2379,
+      "step": 2453
+    },
+    {
+      "epoch": 0.039264,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.685000000000001e-05,
+      "loss": 0.2129,
+      "step": 2454
+    },
+    {
+      "epoch": 0.03928,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.68483870967742e-05,
+      "loss": 0.2157,
+      "step": 2455
+    },
+    {
+      "epoch": 0.039296,
+      "grad_norm": 1.4375,
+      "learning_rate": 9.68467741935484e-05,
+      "loss": 0.1844,
+      "step": 2456
+    },
+    {
+      "epoch": 0.039312,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.684516129032258e-05,
+      "loss": 0.1516,
+      "step": 2457
+    },
+    {
+      "epoch": 0.039328,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.684354838709678e-05,
+      "loss": 0.198,
+      "step": 2458
+    },
+    {
+      "epoch": 0.039344,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.684193548387097e-05,
+      "loss": 0.1819,
+      "step": 2459
+    },
+    {
+      "epoch": 0.03936,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.684032258064517e-05,
+      "loss": 0.1662,
+      "step": 2460
+    },
+    {
+      "epoch": 0.039376,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.683870967741935e-05,
+      "loss": 0.2031,
+      "step": 2461
+    },
+    {
+      "epoch": 0.039392,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.683709677419355e-05,
+      "loss": 0.1835,
+      "step": 2462
+    },
+    {
+      "epoch": 0.039408,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.683548387096775e-05,
+      "loss": 0.203,
+      "step": 2463
+    },
+    {
+      "epoch": 0.039424,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.683387096774194e-05,
+      "loss": 0.2181,
+      "step": 2464
+    },
+    {
+      "epoch": 0.03944,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.683225806451614e-05,
+      "loss": 0.2256,
+      "step": 2465
+    },
+    {
+      "epoch": 0.039456,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.683064516129032e-05,
+      "loss": 0.2023,
+      "step": 2466
+    },
+    {
+      "epoch": 0.039472,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.682903225806452e-05,
+      "loss": 0.1661,
+      "step": 2467
+    },
+    {
+      "epoch": 0.039488,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.682741935483871e-05,
+      "loss": 0.1995,
+      "step": 2468
+    },
+    {
+      "epoch": 0.039504,
+      "grad_norm": 1.5,
+      "learning_rate": 9.682580645161291e-05,
+      "loss": 0.215,
+      "step": 2469
+    },
+    {
+      "epoch": 0.03952,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.68241935483871e-05,
+      "loss": 0.1716,
+      "step": 2470
+    },
+    {
+      "epoch": 0.039536,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.68225806451613e-05,
+      "loss": 0.212,
+      "step": 2471
+    },
+    {
+      "epoch": 0.039552,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.682096774193548e-05,
+      "loss": 0.2679,
+      "step": 2472
+    },
+    {
+      "epoch": 0.039568,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.681935483870968e-05,
+      "loss": 0.1569,
+      "step": 2473
+    },
+    {
+      "epoch": 0.039584,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.681774193548388e-05,
+      "loss": 0.2324,
+      "step": 2474
+    },
+    {
+      "epoch": 0.0396,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.681612903225808e-05,
+      "loss": 0.2073,
+      "step": 2475
+    },
+    {
+      "epoch": 0.039616,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.681451612903227e-05,
+      "loss": 0.2351,
+      "step": 2476
+    },
+    {
+      "epoch": 0.039632,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.681290322580647e-05,
+      "loss": 0.2317,
+      "step": 2477
+    },
+    {
+      "epoch": 0.039648,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.681129032258065e-05,
+      "loss": 0.1712,
+      "step": 2478
+    },
+    {
+      "epoch": 0.039664,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.680967741935484e-05,
+      "loss": 0.1845,
+      "step": 2479
+    },
+    {
+      "epoch": 0.03968,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.680806451612904e-05,
+      "loss": 0.2123,
+      "step": 2480
+    },
+    {
+      "epoch": 0.039696,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.680645161290322e-05,
+      "loss": 0.2288,
+      "step": 2481
+    },
+    {
+      "epoch": 0.039712,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.680483870967742e-05,
+      "loss": 0.1424,
+      "step": 2482
+    },
+    {
+      "epoch": 0.039728,
+      "grad_norm": 1.4453125,
+      "learning_rate": 9.680322580645161e-05,
+      "loss": 0.2273,
+      "step": 2483
+    },
+    {
+      "epoch": 0.039744,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.680161290322581e-05,
+      "loss": 0.1995,
+      "step": 2484
+    },
+    {
+      "epoch": 0.03976,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.680000000000001e-05,
+      "loss": 0.1774,
+      "step": 2485
+    },
+    {
+      "epoch": 0.039776,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.679838709677421e-05,
+      "loss": 0.1973,
+      "step": 2486
+    },
+    {
+      "epoch": 0.039792,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.67967741935484e-05,
+      "loss": 0.2051,
+      "step": 2487
+    },
+    {
+      "epoch": 0.039808,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.67951612903226e-05,
+      "loss": 0.2134,
+      "step": 2488
+    },
+    {
+      "epoch": 0.039824,
+      "grad_norm": 1.4375,
+      "learning_rate": 9.679354838709678e-05,
+      "loss": 0.2031,
+      "step": 2489
+    },
+    {
+      "epoch": 0.03984,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.679193548387098e-05,
+      "loss": 0.1894,
+      "step": 2490
+    },
+    {
+      "epoch": 0.039856,
+      "grad_norm": 1.4375,
+      "learning_rate": 9.679032258064517e-05,
+      "loss": 0.2196,
+      "step": 2491
+    },
+    {
+      "epoch": 0.039872,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.678870967741937e-05,
+      "loss": 0.28,
+      "step": 2492
+    },
+    {
+      "epoch": 0.039888,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.678709677419355e-05,
+      "loss": 0.2215,
+      "step": 2493
+    },
+    {
+      "epoch": 0.039904,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.678548387096774e-05,
+      "loss": 0.2081,
+      "step": 2494
+    },
+    {
+      "epoch": 0.03992,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.678387096774194e-05,
+      "loss": 0.1812,
+      "step": 2495
+    },
+    {
+      "epoch": 0.039936,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.678225806451612e-05,
+      "loss": 0.229,
+      "step": 2496
+    },
+    {
+      "epoch": 0.039952,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.678064516129032e-05,
+      "loss": 0.1654,
+      "step": 2497
+    },
+    {
+      "epoch": 0.039968,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.677903225806452e-05,
+      "loss": 0.181,
+      "step": 2498
+    },
+    {
+      "epoch": 0.039984,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.677741935483872e-05,
+      "loss": 0.1875,
+      "step": 2499
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.677580645161291e-05,
+      "loss": 0.1899,
+      "step": 2500
+    },
+    {
+      "epoch": 0.040016,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.677419354838711e-05,
+      "loss": 0.1578,
+      "step": 2501
+    },
+    {
+      "epoch": 0.040032,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.67725806451613e-05,
+      "loss": 0.1929,
+      "step": 2502
+    },
+    {
+      "epoch": 0.040048,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.67709677419355e-05,
+      "loss": 0.2074,
+      "step": 2503
+    },
+    {
+      "epoch": 0.040064,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.676935483870968e-05,
+      "loss": 0.1618,
+      "step": 2504
+    },
+    {
+      "epoch": 0.04008,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.676774193548388e-05,
+      "loss": 0.1861,
+      "step": 2505
+    },
+    {
+      "epoch": 0.040096,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.676612903225807e-05,
+      "loss": 0.2147,
+      "step": 2506
+    },
+    {
+      "epoch": 0.040112,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.676451612903226e-05,
+      "loss": 0.1629,
+      "step": 2507
+    },
+    {
+      "epoch": 0.040128,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.676290322580645e-05,
+      "loss": 0.1409,
+      "step": 2508
+    },
+    {
+      "epoch": 0.040144,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.676129032258065e-05,
+      "loss": 0.2151,
+      "step": 2509
+    },
+    {
+      "epoch": 0.04016,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.675967741935485e-05,
+      "loss": 0.2228,
+      "step": 2510
+    },
+    {
+      "epoch": 0.040176,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.675806451612904e-05,
+      "loss": 0.1944,
+      "step": 2511
+    },
+    {
+      "epoch": 0.040192,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.675645161290324e-05,
+      "loss": 0.165,
+      "step": 2512
+    },
+    {
+      "epoch": 0.040208,
+      "grad_norm": 0.75,
+      "learning_rate": 9.675483870967742e-05,
+      "loss": 0.1785,
+      "step": 2513
+    },
+    {
+      "epoch": 0.040224,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.675322580645162e-05,
+      "loss": 0.1917,
+      "step": 2514
+    },
+    {
+      "epoch": 0.04024,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.675161290322581e-05,
+      "loss": 0.1825,
+      "step": 2515
+    },
+    {
+      "epoch": 0.040256,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.675000000000001e-05,
+      "loss": 0.1677,
+      "step": 2516
+    },
+    {
+      "epoch": 0.040272,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.674838709677419e-05,
+      "loss": 0.1525,
+      "step": 2517
+    },
+    {
+      "epoch": 0.040288,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.674677419354839e-05,
+      "loss": 0.1964,
+      "step": 2518
+    },
+    {
+      "epoch": 0.040304,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.674516129032258e-05,
+      "loss": 0.1678,
+      "step": 2519
+    },
+    {
+      "epoch": 0.04032,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.674354838709678e-05,
+      "loss": 0.2112,
+      "step": 2520
+    },
+    {
+      "epoch": 0.040336,
+      "grad_norm": 1.3828125,
+      "learning_rate": 9.674193548387096e-05,
+      "loss": 0.2037,
+      "step": 2521
+    },
+    {
+      "epoch": 0.040352,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.674032258064516e-05,
+      "loss": 0.2063,
+      "step": 2522
+    },
+    {
+      "epoch": 0.040368,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.673870967741936e-05,
+      "loss": 0.2261,
+      "step": 2523
+    },
+    {
+      "epoch": 0.040384,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.673709677419356e-05,
+      "loss": 0.2025,
+      "step": 2524
+    },
+    {
+      "epoch": 0.0404,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.673548387096775e-05,
+      "loss": 0.1977,
+      "step": 2525
+    },
+    {
+      "epoch": 0.040416,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.673387096774194e-05,
+      "loss": 0.1784,
+      "step": 2526
+    },
+    {
+      "epoch": 0.040432,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.673225806451614e-05,
+      "loss": 0.2044,
+      "step": 2527
+    },
+    {
+      "epoch": 0.040448,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.673064516129032e-05,
+      "loss": 0.1902,
+      "step": 2528
+    },
+    {
+      "epoch": 0.040464,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.672903225806452e-05,
+      "loss": 0.1971,
+      "step": 2529
+    },
+    {
+      "epoch": 0.04048,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.672741935483871e-05,
+      "loss": 0.1463,
+      "step": 2530
+    },
+    {
+      "epoch": 0.040496,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.67258064516129e-05,
+      "loss": 0.1858,
+      "step": 2531
+    },
+    {
+      "epoch": 0.040512,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.672419354838709e-05,
+      "loss": 0.2006,
+      "step": 2532
+    },
+    {
+      "epoch": 0.040528,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.672258064516129e-05,
+      "loss": 0.165,
+      "step": 2533
+    },
+    {
+      "epoch": 0.040544,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.672096774193549e-05,
+      "loss": 0.2242,
+      "step": 2534
+    },
+    {
+      "epoch": 0.04056,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.671935483870969e-05,
+      "loss": 0.1974,
+      "step": 2535
+    },
+    {
+      "epoch": 0.040576,
+      "grad_norm": 1.359375,
+      "learning_rate": 9.671774193548388e-05,
+      "loss": 0.1845,
+      "step": 2536
+    },
+    {
+      "epoch": 0.040592,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.671612903225808e-05,
+      "loss": 0.1856,
+      "step": 2537
+    },
+    {
+      "epoch": 0.040608,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.671451612903226e-05,
+      "loss": 0.1955,
+      "step": 2538
+    },
+    {
+      "epoch": 0.040624,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.671290322580646e-05,
+      "loss": 0.2186,
+      "step": 2539
+    },
+    {
+      "epoch": 0.04064,
+      "grad_norm": 0.875,
+      "learning_rate": 9.671129032258065e-05,
+      "loss": 0.1901,
+      "step": 2540
+    },
+    {
+      "epoch": 0.040656,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.670967741935484e-05,
+      "loss": 0.2408,
+      "step": 2541
+    },
+    {
+      "epoch": 0.040672,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.670806451612903e-05,
+      "loss": 0.1659,
+      "step": 2542
+    },
+    {
+      "epoch": 0.040688,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.670645161290322e-05,
+      "loss": 0.2157,
+      "step": 2543
+    },
+    {
+      "epoch": 0.040704,
+      "grad_norm": 1.3828125,
+      "learning_rate": 9.670483870967742e-05,
+      "loss": 0.1616,
+      "step": 2544
+    },
+    {
+      "epoch": 0.04072,
+      "grad_norm": 1.125,
+      "learning_rate": 9.670322580645162e-05,
+      "loss": 0.2126,
+      "step": 2545
+    },
+    {
+      "epoch": 0.040736,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.670161290322582e-05,
+      "loss": 0.2096,
+      "step": 2546
+    },
+    {
+      "epoch": 0.040752,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.67e-05,
+      "loss": 0.1993,
+      "step": 2547
+    },
+    {
+      "epoch": 0.040768,
+      "grad_norm": 1.3359375,
+      "learning_rate": 9.66983870967742e-05,
+      "loss": 0.2304,
+      "step": 2548
+    },
+    {
+      "epoch": 0.040784,
+      "grad_norm": 1.4375,
+      "learning_rate": 9.669677419354839e-05,
+      "loss": 0.2305,
+      "step": 2549
+    },
+    {
+      "epoch": 0.0408,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.669516129032259e-05,
+      "loss": 0.1921,
+      "step": 2550
+    },
+    {
+      "epoch": 0.040816,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.669354838709678e-05,
+      "loss": 0.1629,
+      "step": 2551
+    },
+    {
+      "epoch": 0.040832,
+      "grad_norm": 1.0,
+      "learning_rate": 9.669193548387098e-05,
+      "loss": 0.2144,
+      "step": 2552
+    },
+    {
+      "epoch": 0.040848,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.669032258064516e-05,
+      "loss": 0.2093,
+      "step": 2553
+    },
+    {
+      "epoch": 0.040864,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.668870967741936e-05,
+      "loss": 0.1717,
+      "step": 2554
+    },
+    {
+      "epoch": 0.04088,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.668709677419355e-05,
+      "loss": 0.1732,
+      "step": 2555
+    },
+    {
+      "epoch": 0.040896,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.668548387096773e-05,
+      "loss": 0.194,
+      "step": 2556
+    },
+    {
+      "epoch": 0.040912,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.668387096774193e-05,
+      "loss": 0.2152,
+      "step": 2557
+    },
+    {
+      "epoch": 0.040928,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.668225806451613e-05,
+      "loss": 0.2232,
+      "step": 2558
+    },
+    {
+      "epoch": 0.040944,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.668064516129033e-05,
+      "loss": 0.1699,
+      "step": 2559
+    },
+    {
+      "epoch": 0.04096,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.667903225806452e-05,
+      "loss": 0.1881,
+      "step": 2560
+    },
+    {
+      "epoch": 0.040976,
+      "grad_norm": 1.4375,
+      "learning_rate": 9.667741935483872e-05,
+      "loss": 0.1617,
+      "step": 2561
+    },
+    {
+      "epoch": 0.040992,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.66758064516129e-05,
+      "loss": 0.1768,
+      "step": 2562
+    },
+    {
+      "epoch": 0.041008,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.66741935483871e-05,
+      "loss": 0.1716,
+      "step": 2563
+    },
+    {
+      "epoch": 0.041024,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.667258064516129e-05,
+      "loss": 0.2005,
+      "step": 2564
+    },
+    {
+      "epoch": 0.04104,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.667096774193549e-05,
+      "loss": 0.1916,
+      "step": 2565
+    },
+    {
+      "epoch": 0.041056,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.666935483870968e-05,
+      "loss": 0.2101,
+      "step": 2566
+    },
+    {
+      "epoch": 0.041072,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.666774193548388e-05,
+      "loss": 0.1836,
+      "step": 2567
+    },
+    {
+      "epoch": 0.041088,
+      "grad_norm": 1.96875,
+      "learning_rate": 9.666612903225806e-05,
+      "loss": 0.2468,
+      "step": 2568
+    },
+    {
+      "epoch": 0.041104,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.666451612903226e-05,
+      "loss": 0.184,
+      "step": 2569
+    },
+    {
+      "epoch": 0.04112,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.666290322580646e-05,
+      "loss": 0.183,
+      "step": 2570
+    },
+    {
+      "epoch": 0.041136,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.666129032258065e-05,
+      "loss": 0.2399,
+      "step": 2571
+    },
+    {
+      "epoch": 0.041152,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.665967741935485e-05,
+      "loss": 0.2443,
+      "step": 2572
+    },
+    {
+      "epoch": 0.041168,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.665806451612903e-05,
+      "loss": 0.2237,
+      "step": 2573
+    },
+    {
+      "epoch": 0.041184,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.665645161290323e-05,
+      "loss": 0.1761,
+      "step": 2574
+    },
+    {
+      "epoch": 0.0412,
+      "grad_norm": 0.875,
+      "learning_rate": 9.665483870967742e-05,
+      "loss": 0.1925,
+      "step": 2575
+    },
+    {
+      "epoch": 0.041216,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.665322580645162e-05,
+      "loss": 0.1691,
+      "step": 2576
+    },
+    {
+      "epoch": 0.041232,
+      "grad_norm": 1.515625,
+      "learning_rate": 9.66516129032258e-05,
+      "loss": 0.2166,
+      "step": 2577
+    },
+    {
+      "epoch": 0.041248,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.665e-05,
+      "loss": 0.1461,
+      "step": 2578
+    },
+    {
+      "epoch": 0.041264,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.664838709677419e-05,
+      "loss": 0.1779,
+      "step": 2579
+    },
+    {
+      "epoch": 0.04128,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.664677419354839e-05,
+      "loss": 0.1521,
+      "step": 2580
+    },
+    {
+      "epoch": 0.041296,
+      "grad_norm": 1.375,
+      "learning_rate": 9.664516129032259e-05,
+      "loss": 0.2108,
+      "step": 2581
+    },
+    {
+      "epoch": 0.041312,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.664354838709678e-05,
+      "loss": 0.158,
+      "step": 2582
+    },
+    {
+      "epoch": 0.041328,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.664193548387098e-05,
+      "loss": 0.183,
+      "step": 2583
+    },
+    {
+      "epoch": 0.041344,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.664032258064518e-05,
+      "loss": 0.1996,
+      "step": 2584
+    },
+    {
+      "epoch": 0.04136,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.663870967741936e-05,
+      "loss": 0.1751,
+      "step": 2585
+    },
+    {
+      "epoch": 0.041376,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.663709677419356e-05,
+      "loss": 0.189,
+      "step": 2586
+    },
+    {
+      "epoch": 0.041392,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.663548387096775e-05,
+      "loss": 0.1703,
+      "step": 2587
+    },
+    {
+      "epoch": 0.041408,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.663387096774193e-05,
+      "loss": 0.1909,
+      "step": 2588
+    },
+    {
+      "epoch": 0.041424,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.663225806451613e-05,
+      "loss": 0.1486,
+      "step": 2589
+    },
+    {
+      "epoch": 0.04144,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.663064516129032e-05,
+      "loss": 0.2006,
+      "step": 2590
+    },
+    {
+      "epoch": 0.041456,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.662903225806452e-05,
+      "loss": 0.2143,
+      "step": 2591
+    },
+    {
+      "epoch": 0.041472,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.66274193548387e-05,
+      "loss": 0.1855,
+      "step": 2592
+    },
+    {
+      "epoch": 0.041488,
+      "grad_norm": 1.25,
+      "learning_rate": 9.66258064516129e-05,
+      "loss": 0.2038,
+      "step": 2593
+    },
+    {
+      "epoch": 0.041504,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.66241935483871e-05,
+      "loss": 0.1828,
+      "step": 2594
+    },
+    {
+      "epoch": 0.04152,
+      "grad_norm": 1.46875,
+      "learning_rate": 9.66225806451613e-05,
+      "loss": 0.2161,
+      "step": 2595
+    },
+    {
+      "epoch": 0.041536,
+      "grad_norm": 1.3515625,
+      "learning_rate": 9.662096774193549e-05,
+      "loss": 0.193,
+      "step": 2596
+    },
+    {
+      "epoch": 0.041552,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.661935483870969e-05,
+      "loss": 0.1794,
+      "step": 2597
+    },
+    {
+      "epoch": 0.041568,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.661774193548388e-05,
+      "loss": 0.2124,
+      "step": 2598
+    },
+    {
+      "epoch": 0.041584,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.661612903225807e-05,
+      "loss": 0.1932,
+      "step": 2599
+    },
+    {
+      "epoch": 0.0416,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.661451612903226e-05,
+      "loss": 0.1725,
+      "step": 2600
+    },
+    {
+      "epoch": 0.041616,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.661290322580646e-05,
+      "loss": 0.2139,
+      "step": 2601
+    },
+    {
+      "epoch": 0.041632,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.661129032258065e-05,
+      "loss": 0.1692,
+      "step": 2602
+    },
+    {
+      "epoch": 0.041648,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.660967741935483e-05,
+      "loss": 0.1786,
+      "step": 2603
+    },
+    {
+      "epoch": 0.041664,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.660806451612903e-05,
+      "loss": 0.1948,
+      "step": 2604
+    },
+    {
+      "epoch": 0.04168,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.660645161290323e-05,
+      "loss": 0.201,
+      "step": 2605
+    },
+    {
+      "epoch": 0.041696,
+      "grad_norm": 1.296875,
+      "learning_rate": 9.660483870967743e-05,
+      "loss": 0.2223,
+      "step": 2606
+    },
+    {
+      "epoch": 0.041712,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.660322580645162e-05,
+      "loss": 0.1908,
+      "step": 2607
+    },
+    {
+      "epoch": 0.041728,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.660161290322582e-05,
+      "loss": 0.1909,
+      "step": 2608
+    },
+    {
+      "epoch": 0.041744,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.66e-05,
+      "loss": 0.2055,
+      "step": 2609
+    },
+    {
+      "epoch": 0.04176,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.65983870967742e-05,
+      "loss": 0.1739,
+      "step": 2610
+    },
+    {
+      "epoch": 0.041776,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.659677419354839e-05,
+      "loss": 0.2319,
+      "step": 2611
+    },
+    {
+      "epoch": 0.041792,
+      "grad_norm": 1.390625,
+      "learning_rate": 9.659516129032259e-05,
+      "loss": 0.214,
+      "step": 2612
+    },
+    {
+      "epoch": 0.041808,
+      "grad_norm": 1.3125,
+      "learning_rate": 9.659354838709677e-05,
+      "loss": 0.1877,
+      "step": 2613
+    },
+    {
+      "epoch": 0.041824,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.659193548387097e-05,
+      "loss": 0.2042,
+      "step": 2614
+    },
+    {
+      "epoch": 0.04184,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.659032258064516e-05,
+      "loss": 0.1962,
+      "step": 2615
+    },
+    {
+      "epoch": 0.041856,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.658870967741936e-05,
+      "loss": 0.1853,
+      "step": 2616
+    },
+    {
+      "epoch": 0.041872,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.658709677419355e-05,
+      "loss": 0.1834,
+      "step": 2617
+    },
+    {
+      "epoch": 0.041888,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.658548387096775e-05,
+      "loss": 0.1888,
+      "step": 2618
+    },
+    {
+      "epoch": 0.041904,
+      "grad_norm": 1.3515625,
+      "learning_rate": 9.658387096774195e-05,
+      "loss": 0.1897,
+      "step": 2619
+    },
+    {
+      "epoch": 0.04192,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.658225806451613e-05,
+      "loss": 0.1455,
+      "step": 2620
+    },
+    {
+      "epoch": 0.041936,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.658064516129033e-05,
+      "loss": 0.1846,
+      "step": 2621
+    },
+    {
+      "epoch": 0.041952,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.657903225806452e-05,
+      "loss": 0.1772,
+      "step": 2622
+    },
+    {
+      "epoch": 0.041968,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.657741935483872e-05,
+      "loss": 0.156,
+      "step": 2623
+    },
+    {
+      "epoch": 0.041984,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.65758064516129e-05,
+      "loss": 0.1889,
+      "step": 2624
+    },
+    {
+      "epoch": 0.042,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.65741935483871e-05,
+      "loss": 0.2029,
+      "step": 2625
+    },
+    {
+      "epoch": 0.042016,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.657258064516129e-05,
+      "loss": 0.2247,
+      "step": 2626
+    },
+    {
+      "epoch": 0.042032,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.657096774193549e-05,
+      "loss": 0.2005,
+      "step": 2627
+    },
+    {
+      "epoch": 0.042048,
+      "grad_norm": 1.3359375,
+      "learning_rate": 9.656935483870967e-05,
+      "loss": 0.2069,
+      "step": 2628
+    },
+    {
+      "epoch": 0.042064,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.656774193548387e-05,
+      "loss": 0.1854,
+      "step": 2629
+    },
+    {
+      "epoch": 0.04208,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.656612903225807e-05,
+      "loss": 0.2114,
+      "step": 2630
+    },
+    {
+      "epoch": 0.042096,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.656451612903227e-05,
+      "loss": 0.1759,
+      "step": 2631
+    },
+    {
+      "epoch": 0.042112,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.656290322580646e-05,
+      "loss": 0.1962,
+      "step": 2632
+    },
+    {
+      "epoch": 0.042128,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.656129032258066e-05,
+      "loss": 0.17,
+      "step": 2633
+    },
+    {
+      "epoch": 0.042144,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.655967741935485e-05,
+      "loss": 0.1656,
+      "step": 2634
+    },
+    {
+      "epoch": 0.04216,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.655806451612903e-05,
+      "loss": 0.1797,
+      "step": 2635
+    },
+    {
+      "epoch": 0.042176,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.655645161290323e-05,
+      "loss": 0.1752,
+      "step": 2636
+    },
+    {
+      "epoch": 0.042192,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.655483870967742e-05,
+      "loss": 0.16,
+      "step": 2637
+    },
+    {
+      "epoch": 0.042208,
+      "grad_norm": 1.5,
+      "learning_rate": 9.655322580645162e-05,
+      "loss": 0.2051,
+      "step": 2638
+    },
+    {
+      "epoch": 0.042224,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.65516129032258e-05,
+      "loss": 0.2221,
+      "step": 2639
+    },
+    {
+      "epoch": 0.04224,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.655e-05,
+      "loss": 0.2057,
+      "step": 2640
+    },
+    {
+      "epoch": 0.042256,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.65483870967742e-05,
+      "loss": 0.1662,
+      "step": 2641
+    },
+    {
+      "epoch": 0.042272,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.65467741935484e-05,
+      "loss": 0.145,
+      "step": 2642
+    },
+    {
+      "epoch": 0.042288,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.654516129032259e-05,
+      "loss": 0.2175,
+      "step": 2643
+    },
+    {
+      "epoch": 0.042304,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.654354838709679e-05,
+      "loss": 0.1715,
+      "step": 2644
+    },
+    {
+      "epoch": 0.04232,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.654193548387097e-05,
+      "loss": 0.2042,
+      "step": 2645
+    },
+    {
+      "epoch": 0.042336,
+      "grad_norm": 1.4296875,
+      "learning_rate": 9.654032258064517e-05,
+      "loss": 0.2001,
+      "step": 2646
+    },
+    {
+      "epoch": 0.042352,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.653870967741936e-05,
+      "loss": 0.1766,
+      "step": 2647
+    },
+    {
+      "epoch": 0.042368,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.653709677419356e-05,
+      "loss": 0.1984,
+      "step": 2648
+    },
+    {
+      "epoch": 0.042384,
+      "grad_norm": 0.75,
+      "learning_rate": 9.653548387096774e-05,
+      "loss": 0.1615,
+      "step": 2649
+    },
+    {
+      "epoch": 0.0424,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.653387096774193e-05,
+      "loss": 0.164,
+      "step": 2650
+    },
+    {
+      "epoch": 0.042416,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.653225806451613e-05,
+      "loss": 0.1605,
+      "step": 2651
+    },
+    {
+      "epoch": 0.042432,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.653064516129032e-05,
+      "loss": 0.2173,
+      "step": 2652
+    },
+    {
+      "epoch": 0.042448,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.652903225806452e-05,
+      "loss": 0.1642,
+      "step": 2653
+    },
+    {
+      "epoch": 0.042464,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.652741935483872e-05,
+      "loss": 0.1817,
+      "step": 2654
+    },
+    {
+      "epoch": 0.04248,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.652580645161292e-05,
+      "loss": 0.1893,
+      "step": 2655
+    },
+    {
+      "epoch": 0.042496,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.65241935483871e-05,
+      "loss": 0.1548,
+      "step": 2656
+    },
+    {
+      "epoch": 0.042512,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.65225806451613e-05,
+      "loss": 0.1694,
+      "step": 2657
+    },
+    {
+      "epoch": 0.042528,
+      "grad_norm": 1.0,
+      "learning_rate": 9.652096774193549e-05,
+      "loss": 0.191,
+      "step": 2658
+    },
+    {
+      "epoch": 0.042544,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.651935483870969e-05,
+      "loss": 0.1426,
+      "step": 2659
+    },
+    {
+      "epoch": 0.04256,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.651774193548387e-05,
+      "loss": 0.1746,
+      "step": 2660
+    },
+    {
+      "epoch": 0.042576,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.651612903225807e-05,
+      "loss": 0.1904,
+      "step": 2661
+    },
+    {
+      "epoch": 0.042592,
+      "grad_norm": 1.25,
+      "learning_rate": 9.651451612903226e-05,
+      "loss": 0.1662,
+      "step": 2662
+    },
+    {
+      "epoch": 0.042608,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.651290322580646e-05,
+      "loss": 0.1458,
+      "step": 2663
+    },
+    {
+      "epoch": 0.042624,
+      "grad_norm": 1.484375,
+      "learning_rate": 9.651129032258064e-05,
+      "loss": 0.204,
+      "step": 2664
+    },
+    {
+      "epoch": 0.04264,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.650967741935484e-05,
+      "loss": 0.1933,
+      "step": 2665
+    },
+    {
+      "epoch": 0.042656,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.650806451612904e-05,
+      "loss": 0.1992,
+      "step": 2666
+    },
+    {
+      "epoch": 0.042672,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.650645161290323e-05,
+      "loss": 0.166,
+      "step": 2667
+    },
+    {
+      "epoch": 0.042688,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.650483870967743e-05,
+      "loss": 0.2205,
+      "step": 2668
+    },
+    {
+      "epoch": 0.042704,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.650322580645162e-05,
+      "loss": 0.1497,
+      "step": 2669
+    },
+    {
+      "epoch": 0.04272,
+      "grad_norm": 1.375,
+      "learning_rate": 9.650161290322581e-05,
+      "loss": 0.2422,
+      "step": 2670
+    },
+    {
+      "epoch": 0.042736,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.65e-05,
+      "loss": 0.2274,
+      "step": 2671
+    },
+    {
+      "epoch": 0.042752,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.64983870967742e-05,
+      "loss": 0.2236,
+      "step": 2672
+    },
+    {
+      "epoch": 0.042768,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.649677419354839e-05,
+      "loss": 0.1939,
+      "step": 2673
+    },
+    {
+      "epoch": 0.042784,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.649516129032259e-05,
+      "loss": 0.1745,
+      "step": 2674
+    },
+    {
+      "epoch": 0.0428,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.649354838709677e-05,
+      "loss": 0.192,
+      "step": 2675
+    },
+    {
+      "epoch": 0.042816,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.649193548387097e-05,
+      "loss": 0.1918,
+      "step": 2676
+    },
+    {
+      "epoch": 0.042832,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.649032258064517e-05,
+      "loss": 0.2028,
+      "step": 2677
+    },
+    {
+      "epoch": 0.042848,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.648870967741936e-05,
+      "loss": 0.1882,
+      "step": 2678
+    },
+    {
+      "epoch": 0.042864,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.648709677419356e-05,
+      "loss": 0.1999,
+      "step": 2679
+    },
+    {
+      "epoch": 0.04288,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.648548387096774e-05,
+      "loss": 0.1486,
+      "step": 2680
+    },
+    {
+      "epoch": 0.042896,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.648387096774194e-05,
+      "loss": 0.1797,
+      "step": 2681
+    },
+    {
+      "epoch": 0.042912,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.648225806451613e-05,
+      "loss": 0.1863,
+      "step": 2682
+    },
+    {
+      "epoch": 0.042928,
+      "grad_norm": 0.75,
+      "learning_rate": 9.648064516129033e-05,
+      "loss": 0.1729,
+      "step": 2683
+    },
+    {
+      "epoch": 0.042944,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.647903225806451e-05,
+      "loss": 0.1929,
+      "step": 2684
+    },
+    {
+      "epoch": 0.04296,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.647741935483871e-05,
+      "loss": 0.2228,
+      "step": 2685
+    },
+    {
+      "epoch": 0.042976,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.64758064516129e-05,
+      "loss": 0.2175,
+      "step": 2686
+    },
+    {
+      "epoch": 0.042992,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.64741935483871e-05,
+      "loss": 0.1434,
+      "step": 2687
+    },
+    {
+      "epoch": 0.043008,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.647258064516129e-05,
+      "loss": 0.1329,
+      "step": 2688
+    },
+    {
+      "epoch": 0.043024,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.647096774193549e-05,
+      "loss": 0.1721,
+      "step": 2689
+    },
+    {
+      "epoch": 0.04304,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.646935483870969e-05,
+      "loss": 0.1972,
+      "step": 2690
+    },
+    {
+      "epoch": 0.043056,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.646774193548389e-05,
+      "loss": 0.245,
+      "step": 2691
+    },
+    {
+      "epoch": 0.043072,
+      "grad_norm": 1.328125,
+      "learning_rate": 9.646612903225807e-05,
+      "loss": 0.1788,
+      "step": 2692
+    },
+    {
+      "epoch": 0.043088,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.646451612903227e-05,
+      "loss": 0.19,
+      "step": 2693
+    },
+    {
+      "epoch": 0.043104,
+      "grad_norm": 1.34375,
+      "learning_rate": 9.646290322580646e-05,
+      "loss": 0.2039,
+      "step": 2694
+    },
+    {
+      "epoch": 0.04312,
+      "grad_norm": 1.375,
+      "learning_rate": 9.646129032258066e-05,
+      "loss": 0.1812,
+      "step": 2695
+    },
+    {
+      "epoch": 0.043136,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.645967741935484e-05,
+      "loss": 0.1534,
+      "step": 2696
+    },
+    {
+      "epoch": 0.043152,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.645806451612903e-05,
+      "loss": 0.1834,
+      "step": 2697
+    },
+    {
+      "epoch": 0.043168,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.645645161290323e-05,
+      "loss": 0.175,
+      "step": 2698
+    },
+    {
+      "epoch": 0.043184,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.645483870967741e-05,
+      "loss": 0.1824,
+      "step": 2699
+    },
+    {
+      "epoch": 0.0432,
+      "grad_norm": 1.40625,
+      "learning_rate": 9.645322580645161e-05,
+      "loss": 0.2144,
+      "step": 2700
+    },
+    {
+      "epoch": 0.043216,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.645161290322581e-05,
+      "loss": 0.1851,
+      "step": 2701
+    },
+    {
+      "epoch": 0.043232,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.645000000000001e-05,
+      "loss": 0.1815,
+      "step": 2702
+    },
+    {
+      "epoch": 0.043248,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.64483870967742e-05,
+      "loss": 0.1957,
+      "step": 2703
+    },
+    {
+      "epoch": 0.043264,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.64467741935484e-05,
+      "loss": 0.2025,
+      "step": 2704
+    },
+    {
+      "epoch": 0.04328,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.644516129032259e-05,
+      "loss": 0.1851,
+      "step": 2705
+    },
+    {
+      "epoch": 0.043296,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.644354838709678e-05,
+      "loss": 0.2297,
+      "step": 2706
+    },
+    {
+      "epoch": 0.043312,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.644193548387097e-05,
+      "loss": 0.2034,
+      "step": 2707
+    },
+    {
+      "epoch": 0.043328,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.644032258064517e-05,
+      "loss": 0.1505,
+      "step": 2708
+    },
+    {
+      "epoch": 0.043344,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.643870967741936e-05,
+      "loss": 0.2084,
+      "step": 2709
+    },
+    {
+      "epoch": 0.04336,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.643709677419356e-05,
+      "loss": 0.1766,
+      "step": 2710
+    },
+    {
+      "epoch": 0.043376,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.643548387096774e-05,
+      "loss": 0.2174,
+      "step": 2711
+    },
+    {
+      "epoch": 0.043392,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.643387096774193e-05,
+      "loss": 0.2005,
+      "step": 2712
+    },
+    {
+      "epoch": 0.043408,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.643225806451613e-05,
+      "loss": 0.2009,
+      "step": 2713
+    },
+    {
+      "epoch": 0.043424,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.643064516129033e-05,
+      "loss": 0.1811,
+      "step": 2714
+    },
+    {
+      "epoch": 0.04344,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.642903225806453e-05,
+      "loss": 0.2321,
+      "step": 2715
+    },
+    {
+      "epoch": 0.043456,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.642741935483871e-05,
+      "loss": 0.1562,
+      "step": 2716
+    },
+    {
+      "epoch": 0.043472,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.642580645161291e-05,
+      "loss": 0.1736,
+      "step": 2717
+    },
+    {
+      "epoch": 0.043488,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.64241935483871e-05,
+      "loss": 0.1945,
+      "step": 2718
+    },
+    {
+      "epoch": 0.043504,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.64225806451613e-05,
+      "loss": 0.159,
+      "step": 2719
+    },
+    {
+      "epoch": 0.04352,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.642096774193548e-05,
+      "loss": 0.166,
+      "step": 2720
+    },
+    {
+      "epoch": 0.043536,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.641935483870968e-05,
+      "loss": 0.1512,
+      "step": 2721
+    },
+    {
+      "epoch": 0.043552,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.641774193548387e-05,
+      "loss": 0.2057,
+      "step": 2722
+    },
+    {
+      "epoch": 0.043568,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.641612903225807e-05,
+      "loss": 0.1649,
+      "step": 2723
+    },
+    {
+      "epoch": 0.043584,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.641451612903226e-05,
+      "loss": 0.1536,
+      "step": 2724
+    },
+    {
+      "epoch": 0.0436,
+      "grad_norm": 1.3359375,
+      "learning_rate": 9.641290322580646e-05,
+      "loss": 0.1732,
+      "step": 2725
+    },
+    {
+      "epoch": 0.043616,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.641129032258066e-05,
+      "loss": 0.2138,
+      "step": 2726
+    },
+    {
+      "epoch": 0.043632,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.640967741935484e-05,
+      "loss": 0.1607,
+      "step": 2727
+    },
+    {
+      "epoch": 0.043648,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.640806451612904e-05,
+      "loss": 0.1575,
+      "step": 2728
+    },
+    {
+      "epoch": 0.043664,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.640645161290323e-05,
+      "loss": 0.2036,
+      "step": 2729
+    },
+    {
+      "epoch": 0.04368,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.640483870967743e-05,
+      "loss": 0.1901,
+      "step": 2730
+    },
+    {
+      "epoch": 0.043696,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.640322580645161e-05,
+      "loss": 0.2083,
+      "step": 2731
+    },
+    {
+      "epoch": 0.043712,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.640161290322581e-05,
+      "loss": 0.1826,
+      "step": 2732
+    },
+    {
+      "epoch": 0.043728,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.64e-05,
+      "loss": 0.1781,
+      "step": 2733
+    },
+    {
+      "epoch": 0.043744,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.63983870967742e-05,
+      "loss": 0.1781,
+      "step": 2734
+    },
+    {
+      "epoch": 0.04376,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.639677419354838e-05,
+      "loss": 0.2002,
+      "step": 2735
+    },
+    {
+      "epoch": 0.043776,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.639516129032258e-05,
+      "loss": 0.1988,
+      "step": 2736
+    },
+    {
+      "epoch": 0.043792,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.639354838709678e-05,
+      "loss": 0.1748,
+      "step": 2737
+    },
+    {
+      "epoch": 0.043808,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.639193548387098e-05,
+      "loss": 0.1687,
+      "step": 2738
+    },
+    {
+      "epoch": 0.043824,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.639032258064517e-05,
+      "loss": 0.1845,
+      "step": 2739
+    },
+    {
+      "epoch": 0.04384,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.638870967741937e-05,
+      "loss": 0.1883,
+      "step": 2740
+    },
+    {
+      "epoch": 0.043856,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.638709677419355e-05,
+      "loss": 0.1876,
+      "step": 2741
+    },
+    {
+      "epoch": 0.043872,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.638548387096774e-05,
+      "loss": 0.1942,
+      "step": 2742
+    },
+    {
+      "epoch": 0.043888,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.638387096774194e-05,
+      "loss": 0.1823,
+      "step": 2743
+    },
+    {
+      "epoch": 0.043904,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.638225806451613e-05,
+      "loss": 0.2198,
+      "step": 2744
+    },
+    {
+      "epoch": 0.04392,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.638064516129033e-05,
+      "loss": 0.2471,
+      "step": 2745
+    },
+    {
+      "epoch": 0.043936,
+      "grad_norm": 1.0,
+      "learning_rate": 9.637903225806451e-05,
+      "loss": 0.1586,
+      "step": 2746
+    },
+    {
+      "epoch": 0.043952,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.637741935483871e-05,
+      "loss": 0.2143,
+      "step": 2747
+    },
+    {
+      "epoch": 0.043968,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.63758064516129e-05,
+      "loss": 0.1681,
+      "step": 2748
+    },
+    {
+      "epoch": 0.043984,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.63741935483871e-05,
+      "loss": 0.1942,
+      "step": 2749
+    },
+    {
+      "epoch": 0.044,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.63725806451613e-05,
+      "loss": 0.158,
+      "step": 2750
+    },
+    {
+      "epoch": 0.044016,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.63709677419355e-05,
+      "loss": 0.1551,
+      "step": 2751
+    },
+    {
+      "epoch": 0.044032,
+      "grad_norm": 1.3515625,
+      "learning_rate": 9.636935483870968e-05,
+      "loss": 0.19,
+      "step": 2752
+    },
+    {
+      "epoch": 0.044048,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.636774193548388e-05,
+      "loss": 0.2148,
+      "step": 2753
+    },
+    {
+      "epoch": 0.044064,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.636612903225807e-05,
+      "loss": 0.1869,
+      "step": 2754
+    },
+    {
+      "epoch": 0.04408,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.636451612903227e-05,
+      "loss": 0.2126,
+      "step": 2755
+    },
+    {
+      "epoch": 0.044096,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.636290322580645e-05,
+      "loss": 0.1924,
+      "step": 2756
+    },
+    {
+      "epoch": 0.044112,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.636129032258065e-05,
+      "loss": 0.2033,
+      "step": 2757
+    },
+    {
+      "epoch": 0.044128,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.635967741935484e-05,
+      "loss": 0.2055,
+      "step": 2758
+    },
+    {
+      "epoch": 0.044144,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.635806451612903e-05,
+      "loss": 0.225,
+      "step": 2759
+    },
+    {
+      "epoch": 0.04416,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.635645161290323e-05,
+      "loss": 0.1832,
+      "step": 2760
+    },
+    {
+      "epoch": 0.044176,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.635483870967743e-05,
+      "loss": 0.1857,
+      "step": 2761
+    },
+    {
+      "epoch": 0.044192,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.635322580645163e-05,
+      "loss": 0.2198,
+      "step": 2762
+    },
+    {
+      "epoch": 0.044208,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.635161290322581e-05,
+      "loss": 0.1865,
+      "step": 2763
+    },
+    {
+      "epoch": 0.044224,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.635000000000001e-05,
+      "loss": 0.2134,
+      "step": 2764
+    },
+    {
+      "epoch": 0.04424,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.63483870967742e-05,
+      "loss": 0.1679,
+      "step": 2765
+    },
+    {
+      "epoch": 0.044256,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.63467741935484e-05,
+      "loss": 0.2013,
+      "step": 2766
+    },
+    {
+      "epoch": 0.044272,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.634516129032258e-05,
+      "loss": 0.1631,
+      "step": 2767
+    },
+    {
+      "epoch": 0.044288,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.634354838709678e-05,
+      "loss": 0.1637,
+      "step": 2768
+    },
+    {
+      "epoch": 0.044304,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.634193548387097e-05,
+      "loss": 0.1861,
+      "step": 2769
+    },
+    {
+      "epoch": 0.04432,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.634032258064517e-05,
+      "loss": 0.2075,
+      "step": 2770
+    },
+    {
+      "epoch": 0.044336,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.633870967741935e-05,
+      "loss": 0.1533,
+      "step": 2771
+    },
+    {
+      "epoch": 0.044352,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.633709677419355e-05,
+      "loss": 0.1808,
+      "step": 2772
+    },
+    {
+      "epoch": 0.044368,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.633548387096774e-05,
+      "loss": 0.2111,
+      "step": 2773
+    },
+    {
+      "epoch": 0.044384,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.633387096774194e-05,
+      "loss": 0.1627,
+      "step": 2774
+    },
+    {
+      "epoch": 0.0444,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.633225806451614e-05,
+      "loss": 0.1838,
+      "step": 2775
+    },
+    {
+      "epoch": 0.044416,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.633064516129033e-05,
+      "loss": 0.1638,
+      "step": 2776
+    },
+    {
+      "epoch": 0.044432,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.632903225806452e-05,
+      "loss": 0.1725,
+      "step": 2777
+    },
+    {
+      "epoch": 0.044448,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.632741935483871e-05,
+      "loss": 0.1936,
+      "step": 2778
+    },
+    {
+      "epoch": 0.044464,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.632580645161291e-05,
+      "loss": 0.222,
+      "step": 2779
+    },
+    {
+      "epoch": 0.04448,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.63241935483871e-05,
+      "loss": 0.1685,
+      "step": 2780
+    },
+    {
+      "epoch": 0.044496,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.63225806451613e-05,
+      "loss": 0.1548,
+      "step": 2781
+    },
+    {
+      "epoch": 0.044512,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.632096774193548e-05,
+      "loss": 0.192,
+      "step": 2782
+    },
+    {
+      "epoch": 0.044528,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.631935483870968e-05,
+      "loss": 0.1925,
+      "step": 2783
+    },
+    {
+      "epoch": 0.044544,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.631774193548387e-05,
+      "loss": 0.2078,
+      "step": 2784
+    },
+    {
+      "epoch": 0.04456,
+      "grad_norm": 1.8203125,
+      "learning_rate": 9.631612903225807e-05,
+      "loss": 0.2385,
+      "step": 2785
+    },
+    {
+      "epoch": 0.044576,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.631451612903227e-05,
+      "loss": 0.2228,
+      "step": 2786
+    },
+    {
+      "epoch": 0.044592,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.631290322580647e-05,
+      "loss": 0.1876,
+      "step": 2787
+    },
+    {
+      "epoch": 0.044608,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.631129032258065e-05,
+      "loss": 0.1462,
+      "step": 2788
+    },
+    {
+      "epoch": 0.044624,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.630967741935484e-05,
+      "loss": 0.174,
+      "step": 2789
+    },
+    {
+      "epoch": 0.04464,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.630806451612904e-05,
+      "loss": 0.1552,
+      "step": 2790
+    },
+    {
+      "epoch": 0.044656,
+      "grad_norm": 1.3828125,
+      "learning_rate": 9.630645161290322e-05,
+      "loss": 0.1688,
+      "step": 2791
+    },
+    {
+      "epoch": 0.044672,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.630483870967742e-05,
+      "loss": 0.1643,
+      "step": 2792
+    },
+    {
+      "epoch": 0.044688,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.630322580645161e-05,
+      "loss": 0.1731,
+      "step": 2793
+    },
+    {
+      "epoch": 0.044704,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.630161290322581e-05,
+      "loss": 0.1852,
+      "step": 2794
+    },
+    {
+      "epoch": 0.04472,
+      "grad_norm": 1.421875,
+      "learning_rate": 9.63e-05,
+      "loss": 0.2295,
+      "step": 2795
+    },
+    {
+      "epoch": 0.044736,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.62983870967742e-05,
+      "loss": 0.2318,
+      "step": 2796
+    },
+    {
+      "epoch": 0.044752,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.62967741935484e-05,
+      "loss": 0.1488,
+      "step": 2797
+    },
+    {
+      "epoch": 0.044768,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.62951612903226e-05,
+      "loss": 0.2252,
+      "step": 2798
+    },
+    {
+      "epoch": 0.044784,
+      "grad_norm": 1.5390625,
+      "learning_rate": 9.629354838709678e-05,
+      "loss": 0.1784,
+      "step": 2799
+    },
+    {
+      "epoch": 0.0448,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.629193548387098e-05,
+      "loss": 0.1979,
+      "step": 2800
+    },
+    {
+      "epoch": 0.044816,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.629032258064517e-05,
+      "loss": 0.1893,
+      "step": 2801
+    },
+    {
+      "epoch": 0.044832,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.628870967741937e-05,
+      "loss": 0.161,
+      "step": 2802
+    },
+    {
+      "epoch": 0.044848,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.628709677419355e-05,
+      "loss": 0.209,
+      "step": 2803
+    },
+    {
+      "epoch": 0.044864,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.628548387096775e-05,
+      "loss": 0.1468,
+      "step": 2804
+    },
+    {
+      "epoch": 0.04488,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.628387096774194e-05,
+      "loss": 0.1863,
+      "step": 2805
+    },
+    {
+      "epoch": 0.044896,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.628225806451612e-05,
+      "loss": 0.1511,
+      "step": 2806
+    },
+    {
+      "epoch": 0.044912,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.628064516129032e-05,
+      "loss": 0.2397,
+      "step": 2807
+    },
+    {
+      "epoch": 0.044928,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.627903225806451e-05,
+      "loss": 0.2429,
+      "step": 2808
+    },
+    {
+      "epoch": 0.044944,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.627741935483871e-05,
+      "loss": 0.1874,
+      "step": 2809
+    },
+    {
+      "epoch": 0.04496,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.627580645161291e-05,
+      "loss": 0.1989,
+      "step": 2810
+    },
+    {
+      "epoch": 0.044976,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.627419354838711e-05,
+      "loss": 0.1882,
+      "step": 2811
+    },
+    {
+      "epoch": 0.044992,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.62725806451613e-05,
+      "loss": 0.2134,
+      "step": 2812
+    },
+    {
+      "epoch": 0.045008,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.62709677419355e-05,
+      "loss": 0.2065,
+      "step": 2813
+    },
+    {
+      "epoch": 0.045024,
+      "grad_norm": 0.875,
+      "learning_rate": 9.626935483870968e-05,
+      "loss": 0.2148,
+      "step": 2814
+    },
+    {
+      "epoch": 0.04504,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.626774193548388e-05,
+      "loss": 0.2036,
+      "step": 2815
+    },
+    {
+      "epoch": 0.045056,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.626612903225807e-05,
+      "loss": 0.2185,
+      "step": 2816
+    },
+    {
+      "epoch": 0.045072,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.626451612903227e-05,
+      "loss": 0.1848,
+      "step": 2817
+    },
+    {
+      "epoch": 0.045088,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.626290322580645e-05,
+      "loss": 0.178,
+      "step": 2818
+    },
+    {
+      "epoch": 0.045104,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.626129032258065e-05,
+      "loss": 0.1653,
+      "step": 2819
+    },
+    {
+      "epoch": 0.04512,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.625967741935484e-05,
+      "loss": 0.1957,
+      "step": 2820
+    },
+    {
+      "epoch": 0.045136,
+      "grad_norm": 1.25,
+      "learning_rate": 9.625806451612904e-05,
+      "loss": 0.2214,
+      "step": 2821
+    },
+    {
+      "epoch": 0.045152,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.625645161290324e-05,
+      "loss": 0.1628,
+      "step": 2822
+    },
+    {
+      "epoch": 0.045168,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.625483870967742e-05,
+      "loss": 0.1949,
+      "step": 2823
+    },
+    {
+      "epoch": 0.045184,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.625322580645162e-05,
+      "loss": 0.2171,
+      "step": 2824
+    },
+    {
+      "epoch": 0.0452,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.625161290322581e-05,
+      "loss": 0.1751,
+      "step": 2825
+    },
+    {
+      "epoch": 0.045216,
+      "grad_norm": 1.4453125,
+      "learning_rate": 9.625000000000001e-05,
+      "loss": 0.2106,
+      "step": 2826
+    },
+    {
+      "epoch": 0.045232,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.62483870967742e-05,
+      "loss": 0.1554,
+      "step": 2827
+    },
+    {
+      "epoch": 0.045248,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.62467741935484e-05,
+      "loss": 0.1926,
+      "step": 2828
+    },
+    {
+      "epoch": 0.045264,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.624516129032258e-05,
+      "loss": 0.1585,
+      "step": 2829
+    },
+    {
+      "epoch": 0.04528,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.624354838709678e-05,
+      "loss": 0.1643,
+      "step": 2830
+    },
+    {
+      "epoch": 0.045296,
+      "grad_norm": 1.4140625,
+      "learning_rate": 9.624193548387097e-05,
+      "loss": 0.2375,
+      "step": 2831
+    },
+    {
+      "epoch": 0.045312,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.624032258064517e-05,
+      "loss": 0.2085,
+      "step": 2832
+    },
+    {
+      "epoch": 0.045328,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.623870967741937e-05,
+      "loss": 0.2147,
+      "step": 2833
+    },
+    {
+      "epoch": 0.045344,
+      "grad_norm": 1.3125,
+      "learning_rate": 9.623709677419355e-05,
+      "loss": 0.1974,
+      "step": 2834
+    },
+    {
+      "epoch": 0.04536,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.623548387096775e-05,
+      "loss": 0.172,
+      "step": 2835
+    },
+    {
+      "epoch": 0.045376,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.623387096774194e-05,
+      "loss": 0.1548,
+      "step": 2836
+    },
+    {
+      "epoch": 0.045392,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.623225806451614e-05,
+      "loss": 0.1594,
+      "step": 2837
+    },
+    {
+      "epoch": 0.045408,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.623064516129032e-05,
+      "loss": 0.173,
+      "step": 2838
+    },
+    {
+      "epoch": 0.045424,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.622903225806452e-05,
+      "loss": 0.2315,
+      "step": 2839
+    },
+    {
+      "epoch": 0.04544,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.622741935483871e-05,
+      "loss": 0.241,
+      "step": 2840
+    },
+    {
+      "epoch": 0.045456,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.622580645161291e-05,
+      "loss": 0.2244,
+      "step": 2841
+    },
+    {
+      "epoch": 0.045472,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.62241935483871e-05,
+      "loss": 0.186,
+      "step": 2842
+    },
+    {
+      "epoch": 0.045488,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.62225806451613e-05,
+      "loss": 0.1683,
+      "step": 2843
+    },
+    {
+      "epoch": 0.045504,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.622096774193548e-05,
+      "loss": 0.1847,
+      "step": 2844
+    },
+    {
+      "epoch": 0.04552,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.621935483870968e-05,
+      "loss": 0.1943,
+      "step": 2845
+    },
+    {
+      "epoch": 0.045536,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.621774193548388e-05,
+      "loss": 0.2383,
+      "step": 2846
+    },
+    {
+      "epoch": 0.045552,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.621612903225808e-05,
+      "loss": 0.2089,
+      "step": 2847
+    },
+    {
+      "epoch": 0.045568,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.621451612903226e-05,
+      "loss": 0.1572,
+      "step": 2848
+    },
+    {
+      "epoch": 0.045584,
+      "grad_norm": 1.3125,
+      "learning_rate": 9.621290322580646e-05,
+      "loss": 0.1861,
+      "step": 2849
+    },
+    {
+      "epoch": 0.0456,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.621129032258065e-05,
+      "loss": 0.1573,
+      "step": 2850
+    },
+    {
+      "epoch": 0.045616,
+      "grad_norm": 1.4375,
+      "learning_rate": 9.620967741935484e-05,
+      "loss": 0.1663,
+      "step": 2851
+    },
+    {
+      "epoch": 0.045632,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.620806451612904e-05,
+      "loss": 0.202,
+      "step": 2852
+    },
+    {
+      "epoch": 0.045648,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.620645161290322e-05,
+      "loss": 0.2117,
+      "step": 2853
+    },
+    {
+      "epoch": 0.045664,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.620483870967742e-05,
+      "loss": 0.1847,
+      "step": 2854
+    },
+    {
+      "epoch": 0.04568,
+      "grad_norm": 1.390625,
+      "learning_rate": 9.620322580645161e-05,
+      "loss": 0.1904,
+      "step": 2855
+    },
+    {
+      "epoch": 0.045696,
+      "grad_norm": 1.421875,
+      "learning_rate": 9.620161290322581e-05,
+      "loss": 0.2164,
+      "step": 2856
+    },
+    {
+      "epoch": 0.045712,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.620000000000001e-05,
+      "loss": 0.1859,
+      "step": 2857
+    },
+    {
+      "epoch": 0.045728,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.619838709677421e-05,
+      "loss": 0.1924,
+      "step": 2858
+    },
+    {
+      "epoch": 0.045744,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.619677419354839e-05,
+      "loss": 0.1795,
+      "step": 2859
+    },
+    {
+      "epoch": 0.04576,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.619516129032259e-05,
+      "loss": 0.1628,
+      "step": 2860
+    },
+    {
+      "epoch": 0.045776,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.619354838709678e-05,
+      "loss": 0.1546,
+      "step": 2861
+    },
+    {
+      "epoch": 0.045792,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.619193548387098e-05,
+      "loss": 0.1848,
+      "step": 2862
+    },
+    {
+      "epoch": 0.045808,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.619032258064516e-05,
+      "loss": 0.1502,
+      "step": 2863
+    },
+    {
+      "epoch": 0.045824,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.618870967741936e-05,
+      "loss": 0.2126,
+      "step": 2864
+    },
+    {
+      "epoch": 0.04584,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.618709677419355e-05,
+      "loss": 0.1834,
+      "step": 2865
+    },
+    {
+      "epoch": 0.045856,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.618548387096775e-05,
+      "loss": 0.1902,
+      "step": 2866
+    },
+    {
+      "epoch": 0.045872,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.618387096774194e-05,
+      "loss": 0.1656,
+      "step": 2867
+    },
+    {
+      "epoch": 0.045888,
+      "grad_norm": 1.296875,
+      "learning_rate": 9.618225806451612e-05,
+      "loss": 0.174,
+      "step": 2868
+    },
+    {
+      "epoch": 0.045904,
+      "grad_norm": 1.46875,
+      "learning_rate": 9.618064516129032e-05,
+      "loss": 0.2344,
+      "step": 2869
+    },
+    {
+      "epoch": 0.04592,
+      "grad_norm": 1.5234375,
+      "learning_rate": 9.617903225806452e-05,
+      "loss": 0.2155,
+      "step": 2870
+    },
+    {
+      "epoch": 0.045936,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.617741935483872e-05,
+      "loss": 0.1799,
+      "step": 2871
+    },
+    {
+      "epoch": 0.045952,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.617580645161291e-05,
+      "loss": 0.1945,
+      "step": 2872
+    },
+    {
+      "epoch": 0.045968,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.61741935483871e-05,
+      "loss": 0.2181,
+      "step": 2873
+    },
+    {
+      "epoch": 0.045984,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.617258064516129e-05,
+      "loss": 0.2,
+      "step": 2874
+    },
+    {
+      "epoch": 0.046,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.617096774193549e-05,
+      "loss": 0.2063,
+      "step": 2875
+    },
+    {
+      "epoch": 0.046016,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.616935483870968e-05,
+      "loss": 0.1826,
+      "step": 2876
+    },
+    {
+      "epoch": 0.046032,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.616774193548388e-05,
+      "loss": 0.1379,
+      "step": 2877
+    },
+    {
+      "epoch": 0.046048,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.616612903225806e-05,
+      "loss": 0.1878,
+      "step": 2878
+    },
+    {
+      "epoch": 0.046064,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.616451612903226e-05,
+      "loss": 0.2311,
+      "step": 2879
+    },
+    {
+      "epoch": 0.04608,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.616290322580645e-05,
+      "loss": 0.1785,
+      "step": 2880
+    },
+    {
+      "epoch": 0.046096,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.616129032258065e-05,
+      "loss": 0.1683,
+      "step": 2881
+    },
+    {
+      "epoch": 0.046112,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.615967741935485e-05,
+      "loss": 0.1851,
+      "step": 2882
+    },
+    {
+      "epoch": 0.046128,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.615806451612904e-05,
+      "loss": 0.2014,
+      "step": 2883
+    },
+    {
+      "epoch": 0.046144,
+      "grad_norm": 1.0,
+      "learning_rate": 9.615645161290323e-05,
+      "loss": 0.2074,
+      "step": 2884
+    },
+    {
+      "epoch": 0.04616,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.615483870967742e-05,
+      "loss": 0.2117,
+      "step": 2885
+    },
+    {
+      "epoch": 0.046176,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.615322580645162e-05,
+      "loss": 0.1856,
+      "step": 2886
+    },
+    {
+      "epoch": 0.046192,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.61516129032258e-05,
+      "loss": 0.2167,
+      "step": 2887
+    },
+    {
+      "epoch": 0.046208,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.615e-05,
+      "loss": 0.2537,
+      "step": 2888
+    },
+    {
+      "epoch": 0.046224,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.614838709677419e-05,
+      "loss": 0.223,
+      "step": 2889
+    },
+    {
+      "epoch": 0.04624,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.614677419354839e-05,
+      "loss": 0.1762,
+      "step": 2890
+    },
+    {
+      "epoch": 0.046256,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.614516129032258e-05,
+      "loss": 0.1483,
+      "step": 2891
+    },
+    {
+      "epoch": 0.046272,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.614354838709678e-05,
+      "loss": 0.2237,
+      "step": 2892
+    },
+    {
+      "epoch": 0.046288,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.614193548387098e-05,
+      "loss": 0.1822,
+      "step": 2893
+    },
+    {
+      "epoch": 0.046304,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.614032258064518e-05,
+      "loss": 0.1703,
+      "step": 2894
+    },
+    {
+      "epoch": 0.04632,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.613870967741936e-05,
+      "loss": 0.1997,
+      "step": 2895
+    },
+    {
+      "epoch": 0.046336,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.613709677419356e-05,
+      "loss": 0.1636,
+      "step": 2896
+    },
+    {
+      "epoch": 0.046352,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.613548387096775e-05,
+      "loss": 0.2189,
+      "step": 2897
+    },
+    {
+      "epoch": 0.046368,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.613387096774193e-05,
+      "loss": 0.1641,
+      "step": 2898
+    },
+    {
+      "epoch": 0.046384,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.613225806451613e-05,
+      "loss": 0.1835,
+      "step": 2899
+    },
+    {
+      "epoch": 0.0464,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.613064516129032e-05,
+      "loss": 0.1699,
+      "step": 2900
+    },
+    {
+      "epoch": 0.046416,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.612903225806452e-05,
+      "loss": 0.1724,
+      "step": 2901
+    },
+    {
+      "epoch": 0.046432,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.61274193548387e-05,
+      "loss": 0.1885,
+      "step": 2902
+    },
+    {
+      "epoch": 0.046448,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.61258064516129e-05,
+      "loss": 0.165,
+      "step": 2903
+    },
+    {
+      "epoch": 0.046464,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.612419354838709e-05,
+      "loss": 0.165,
+      "step": 2904
+    },
+    {
+      "epoch": 0.04648,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.612258064516129e-05,
+      "loss": 0.168,
+      "step": 2905
+    },
+    {
+      "epoch": 0.046496,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.612096774193549e-05,
+      "loss": 0.2191,
+      "step": 2906
+    },
+    {
+      "epoch": 0.046512,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.611935483870969e-05,
+      "loss": 0.2549,
+      "step": 2907
+    },
+    {
+      "epoch": 0.046528,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.611774193548388e-05,
+      "loss": 0.2054,
+      "step": 2908
+    },
+    {
+      "epoch": 0.046544,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.611612903225808e-05,
+      "loss": 0.1818,
+      "step": 2909
+    },
+    {
+      "epoch": 0.04656,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.611451612903226e-05,
+      "loss": 0.1662,
+      "step": 2910
+    },
+    {
+      "epoch": 0.046576,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.611290322580646e-05,
+      "loss": 0.1646,
+      "step": 2911
+    },
+    {
+      "epoch": 0.046592,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.611129032258065e-05,
+      "loss": 0.1449,
+      "step": 2912
+    },
+    {
+      "epoch": 0.046608,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.610967741935485e-05,
+      "loss": 0.2027,
+      "step": 2913
+    },
+    {
+      "epoch": 0.046624,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.610806451612903e-05,
+      "loss": 0.1541,
+      "step": 2914
+    },
+    {
+      "epoch": 0.04664,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.610645161290322e-05,
+      "loss": 0.2178,
+      "step": 2915
+    },
+    {
+      "epoch": 0.046656,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.610483870967742e-05,
+      "loss": 0.1567,
+      "step": 2916
+    },
+    {
+      "epoch": 0.046672,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.610322580645162e-05,
+      "loss": 0.215,
+      "step": 2917
+    },
+    {
+      "epoch": 0.046688,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.610161290322582e-05,
+      "loss": 0.1889,
+      "step": 2918
+    },
+    {
+      "epoch": 0.046704,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.61e-05,
+      "loss": 0.2138,
+      "step": 2919
+    },
+    {
+      "epoch": 0.04672,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.60983870967742e-05,
+      "loss": 0.1688,
+      "step": 2920
+    },
+    {
+      "epoch": 0.046736,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.609677419354839e-05,
+      "loss": 0.1904,
+      "step": 2921
+    },
+    {
+      "epoch": 0.046752,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.609516129032259e-05,
+      "loss": 0.1886,
+      "step": 2922
+    },
+    {
+      "epoch": 0.046768,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.609354838709678e-05,
+      "loss": 0.2174,
+      "step": 2923
+    },
+    {
+      "epoch": 0.046784,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.609193548387098e-05,
+      "loss": 0.2325,
+      "step": 2924
+    },
+    {
+      "epoch": 0.0468,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.609032258064516e-05,
+      "loss": 0.1685,
+      "step": 2925
+    },
+    {
+      "epoch": 0.046816,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.608870967741936e-05,
+      "loss": 0.1733,
+      "step": 2926
+    },
+    {
+      "epoch": 0.046832,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.608709677419355e-05,
+      "loss": 0.1817,
+      "step": 2927
+    },
+    {
+      "epoch": 0.046848,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.608548387096775e-05,
+      "loss": 0.2003,
+      "step": 2928
+    },
+    {
+      "epoch": 0.046864,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.608387096774195e-05,
+      "loss": 0.1585,
+      "step": 2929
+    },
+    {
+      "epoch": 0.04688,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.608225806451613e-05,
+      "loss": 0.2096,
+      "step": 2930
+    },
+    {
+      "epoch": 0.046896,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.608064516129033e-05,
+      "loss": 0.1856,
+      "step": 2931
+    },
+    {
+      "epoch": 0.046912,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.607903225806452e-05,
+      "loss": 0.2367,
+      "step": 2932
+    },
+    {
+      "epoch": 0.046928,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.607741935483872e-05,
+      "loss": 0.1846,
+      "step": 2933
+    },
+    {
+      "epoch": 0.046944,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.60758064516129e-05,
+      "loss": 0.1528,
+      "step": 2934
+    },
+    {
+      "epoch": 0.04696,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.60741935483871e-05,
+      "loss": 0.1926,
+      "step": 2935
+    },
+    {
+      "epoch": 0.046976,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.607258064516129e-05,
+      "loss": 0.2191,
+      "step": 2936
+    },
+    {
+      "epoch": 0.046992,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.607096774193549e-05,
+      "loss": 0.1742,
+      "step": 2937
+    },
+    {
+      "epoch": 0.047008,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.606935483870968e-05,
+      "loss": 0.16,
+      "step": 2938
+    },
+    {
+      "epoch": 0.047024,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.606774193548388e-05,
+      "loss": 0.168,
+      "step": 2939
+    },
+    {
+      "epoch": 0.04704,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.606612903225806e-05,
+      "loss": 0.1577,
+      "step": 2940
+    },
+    {
+      "epoch": 0.047056,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.606451612903226e-05,
+      "loss": 0.1633,
+      "step": 2941
+    },
+    {
+      "epoch": 0.047072,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.606290322580646e-05,
+      "loss": 0.1705,
+      "step": 2942
+    },
+    {
+      "epoch": 0.047088,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.606129032258066e-05,
+      "loss": 0.1577,
+      "step": 2943
+    },
+    {
+      "epoch": 0.047104,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.605967741935485e-05,
+      "loss": 0.1908,
+      "step": 2944
+    },
+    {
+      "epoch": 0.04712,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.605806451612903e-05,
+      "loss": 0.2032,
+      "step": 2945
+    },
+    {
+      "epoch": 0.047136,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.605645161290323e-05,
+      "loss": 0.1544,
+      "step": 2946
+    },
+    {
+      "epoch": 0.047152,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.605483870967742e-05,
+      "loss": 0.2147,
+      "step": 2947
+    },
+    {
+      "epoch": 0.047168,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.605322580645162e-05,
+      "loss": 0.2061,
+      "step": 2948
+    },
+    {
+      "epoch": 0.047184,
+      "grad_norm": 1.328125,
+      "learning_rate": 9.60516129032258e-05,
+      "loss": 0.1918,
+      "step": 2949
+    },
+    {
+      "epoch": 0.0472,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.605e-05,
+      "loss": 0.1916,
+      "step": 2950
+    },
+    {
+      "epoch": 0.047216,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.604838709677419e-05,
+      "loss": 0.1966,
+      "step": 2951
+    },
+    {
+      "epoch": 0.047232,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.604677419354839e-05,
+      "loss": 0.2084,
+      "step": 2952
+    },
+    {
+      "epoch": 0.047248,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.604516129032259e-05,
+      "loss": 0.1896,
+      "step": 2953
+    },
+    {
+      "epoch": 0.047264,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.604354838709679e-05,
+      "loss": 0.202,
+      "step": 2954
+    },
+    {
+      "epoch": 0.04728,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.604193548387097e-05,
+      "loss": 0.1849,
+      "step": 2955
+    },
+    {
+      "epoch": 0.047296,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.604032258064517e-05,
+      "loss": 0.1871,
+      "step": 2956
+    },
+    {
+      "epoch": 0.047312,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.603870967741936e-05,
+      "loss": 0.1719,
+      "step": 2957
+    },
+    {
+      "epoch": 0.047328,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.603709677419356e-05,
+      "loss": 0.1538,
+      "step": 2958
+    },
+    {
+      "epoch": 0.047344,
+      "grad_norm": 1.453125,
+      "learning_rate": 9.603548387096775e-05,
+      "loss": 0.1836,
+      "step": 2959
+    },
+    {
+      "epoch": 0.04736,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.603387096774193e-05,
+      "loss": 0.1862,
+      "step": 2960
+    },
+    {
+      "epoch": 0.047376,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.603225806451613e-05,
+      "loss": 0.1643,
+      "step": 2961
+    },
+    {
+      "epoch": 0.047392,
+      "grad_norm": 0.75,
+      "learning_rate": 9.603064516129032e-05,
+      "loss": 0.1578,
+      "step": 2962
+    },
+    {
+      "epoch": 0.047408,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.602903225806452e-05,
+      "loss": 0.2027,
+      "step": 2963
+    },
+    {
+      "epoch": 0.047424,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.60274193548387e-05,
+      "loss": 0.1989,
+      "step": 2964
+    },
+    {
+      "epoch": 0.04744,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.60258064516129e-05,
+      "loss": 0.2531,
+      "step": 2965
+    },
+    {
+      "epoch": 0.047456,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.60241935483871e-05,
+      "loss": 0.1934,
+      "step": 2966
+    },
+    {
+      "epoch": 0.047472,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.60225806451613e-05,
+      "loss": 0.1819,
+      "step": 2967
+    },
+    {
+      "epoch": 0.047488,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.602096774193549e-05,
+      "loss": 0.1913,
+      "step": 2968
+    },
+    {
+      "epoch": 0.047504,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.601935483870969e-05,
+      "loss": 0.1891,
+      "step": 2969
+    },
+    {
+      "epoch": 0.04752,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.601774193548387e-05,
+      "loss": 0.1826,
+      "step": 2970
+    },
+    {
+      "epoch": 0.047536,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.601612903225807e-05,
+      "loss": 0.1706,
+      "step": 2971
+    },
+    {
+      "epoch": 0.047552,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.601451612903226e-05,
+      "loss": 0.1566,
+      "step": 2972
+    },
+    {
+      "epoch": 0.047568,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.601290322580646e-05,
+      "loss": 0.1888,
+      "step": 2973
+    },
+    {
+      "epoch": 0.047584,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.601129032258065e-05,
+      "loss": 0.1938,
+      "step": 2974
+    },
+    {
+      "epoch": 0.0476,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.600967741935485e-05,
+      "loss": 0.1818,
+      "step": 2975
+    },
+    {
+      "epoch": 0.047616,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.600806451612903e-05,
+      "loss": 0.2394,
+      "step": 2976
+    },
+    {
+      "epoch": 0.047632,
+      "grad_norm": 1.296875,
+      "learning_rate": 9.600645161290323e-05,
+      "loss": 0.2246,
+      "step": 2977
+    },
+    {
+      "epoch": 0.047648,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.600483870967743e-05,
+      "loss": 0.1906,
+      "step": 2978
+    },
+    {
+      "epoch": 0.047664,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.600322580645162e-05,
+      "loss": 0.1725,
+      "step": 2979
+    },
+    {
+      "epoch": 0.04768,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.600161290322582e-05,
+      "loss": 0.139,
+      "step": 2980
+    },
+    {
+      "epoch": 0.047696,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.6e-05,
+      "loss": 0.1736,
+      "step": 2981
+    },
+    {
+      "epoch": 0.047712,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.59983870967742e-05,
+      "loss": 0.1988,
+      "step": 2982
+    },
+    {
+      "epoch": 0.047728,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.599677419354839e-05,
+      "loss": 0.2156,
+      "step": 2983
+    },
+    {
+      "epoch": 0.047744,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.599516129032259e-05,
+      "loss": 0.1951,
+      "step": 2984
+    },
+    {
+      "epoch": 0.04776,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.599354838709677e-05,
+      "loss": 0.1662,
+      "step": 2985
+    },
+    {
+      "epoch": 0.047776,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.599193548387097e-05,
+      "loss": 0.1749,
+      "step": 2986
+    },
+    {
+      "epoch": 0.047792,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.599032258064516e-05,
+      "loss": 0.1619,
+      "step": 2987
+    },
+    {
+      "epoch": 0.047808,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.598870967741936e-05,
+      "loss": 0.1667,
+      "step": 2988
+    },
+    {
+      "epoch": 0.047824,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.598709677419356e-05,
+      "loss": 0.188,
+      "step": 2989
+    },
+    {
+      "epoch": 0.04784,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.598548387096776e-05,
+      "loss": 0.1364,
+      "step": 2990
+    },
+    {
+      "epoch": 0.047856,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.598387096774194e-05,
+      "loss": 0.2516,
+      "step": 2991
+    },
+    {
+      "epoch": 0.047872,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.598225806451613e-05,
+      "loss": 0.1682,
+      "step": 2992
+    },
+    {
+      "epoch": 0.047888,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.598064516129033e-05,
+      "loss": 0.259,
+      "step": 2993
+    },
+    {
+      "epoch": 0.047904,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.597903225806452e-05,
+      "loss": 0.1802,
+      "step": 2994
+    },
+    {
+      "epoch": 0.04792,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.597741935483872e-05,
+      "loss": 0.1776,
+      "step": 2995
+    },
+    {
+      "epoch": 0.047936,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.59758064516129e-05,
+      "loss": 0.2207,
+      "step": 2996
+    },
+    {
+      "epoch": 0.047952,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.59741935483871e-05,
+      "loss": 0.1744,
+      "step": 2997
+    },
+    {
+      "epoch": 0.047968,
+      "grad_norm": 1.390625,
+      "learning_rate": 9.597258064516129e-05,
+      "loss": 0.1954,
+      "step": 2998
+    },
+    {
+      "epoch": 0.047984,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.597096774193549e-05,
+      "loss": 0.1954,
+      "step": 2999
+    },
+    {
+      "epoch": 0.048,
+      "grad_norm": 1.4296875,
+      "learning_rate": 9.596935483870967e-05,
+      "loss": 0.168,
+      "step": 3000
+    },
+    {
+      "epoch": 0.048016,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.596774193548387e-05,
+      "loss": 0.1775,
+      "step": 3001
+    },
+    {
+      "epoch": 0.048032,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.596612903225807e-05,
+      "loss": 0.1668,
+      "step": 3002
+    },
+    {
+      "epoch": 0.048048,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.596451612903227e-05,
+      "loss": 0.162,
+      "step": 3003
+    },
+    {
+      "epoch": 0.048064,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.596290322580646e-05,
+      "loss": 0.2113,
+      "step": 3004
+    },
+    {
+      "epoch": 0.04808,
+      "grad_norm": 1.5234375,
+      "learning_rate": 9.596129032258066e-05,
+      "loss": 0.2058,
+      "step": 3005
+    },
+    {
+      "epoch": 0.048096,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.595967741935484e-05,
+      "loss": 0.19,
+      "step": 3006
+    },
+    {
+      "epoch": 0.048112,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.595806451612903e-05,
+      "loss": 0.1858,
+      "step": 3007
+    },
+    {
+      "epoch": 0.048128,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.595645161290323e-05,
+      "loss": 0.1768,
+      "step": 3008
+    },
+    {
+      "epoch": 0.048144,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.595483870967742e-05,
+      "loss": 0.1826,
+      "step": 3009
+    },
+    {
+      "epoch": 0.04816,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.595322580645162e-05,
+      "loss": 0.1781,
+      "step": 3010
+    },
+    {
+      "epoch": 0.048176,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.59516129032258e-05,
+      "loss": 0.1582,
+      "step": 3011
+    },
+    {
+      "epoch": 0.048192,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.595e-05,
+      "loss": 0.1996,
+      "step": 3012
+    },
+    {
+      "epoch": 0.048208,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.59483870967742e-05,
+      "loss": 0.2472,
+      "step": 3013
+    },
+    {
+      "epoch": 0.048224,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.59467741935484e-05,
+      "loss": 0.1868,
+      "step": 3014
+    },
+    {
+      "epoch": 0.04824,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.594516129032259e-05,
+      "loss": 0.1947,
+      "step": 3015
+    },
+    {
+      "epoch": 0.048256,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.594354838709679e-05,
+      "loss": 0.1854,
+      "step": 3016
+    },
+    {
+      "epoch": 0.048272,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.594193548387097e-05,
+      "loss": 0.1851,
+      "step": 3017
+    },
+    {
+      "epoch": 0.048288,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.594032258064517e-05,
+      "loss": 0.1843,
+      "step": 3018
+    },
+    {
+      "epoch": 0.048304,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.593870967741936e-05,
+      "loss": 0.1976,
+      "step": 3019
+    },
+    {
+      "epoch": 0.04832,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.593709677419356e-05,
+      "loss": 0.1576,
+      "step": 3020
+    },
+    {
+      "epoch": 0.048336,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.593548387096774e-05,
+      "loss": 0.2033,
+      "step": 3021
+    },
+    {
+      "epoch": 0.048352,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.593387096774194e-05,
+      "loss": 0.1804,
+      "step": 3022
+    },
+    {
+      "epoch": 0.048368,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.593225806451613e-05,
+      "loss": 0.1542,
+      "step": 3023
+    },
+    {
+      "epoch": 0.048384,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.593064516129033e-05,
+      "loss": 0.2051,
+      "step": 3024
+    },
+    {
+      "epoch": 0.0484,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.592903225806452e-05,
+      "loss": 0.16,
+      "step": 3025
+    },
+    {
+      "epoch": 0.048416,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.592741935483871e-05,
+      "loss": 0.1945,
+      "step": 3026
+    },
+    {
+      "epoch": 0.048432,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.592580645161291e-05,
+      "loss": 0.1912,
+      "step": 3027
+    },
+    {
+      "epoch": 0.048448,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.59241935483871e-05,
+      "loss": 0.1806,
+      "step": 3028
+    },
+    {
+      "epoch": 0.048464,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.59225806451613e-05,
+      "loss": 0.1649,
+      "step": 3029
+    },
+    {
+      "epoch": 0.04848,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.592096774193549e-05,
+      "loss": 0.1709,
+      "step": 3030
+    },
+    {
+      "epoch": 0.048496,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.591935483870969e-05,
+      "loss": 0.1815,
+      "step": 3031
+    },
+    {
+      "epoch": 0.048512,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.591774193548387e-05,
+      "loss": 0.1883,
+      "step": 3032
+    },
+    {
+      "epoch": 0.048528,
+      "grad_norm": 1.4296875,
+      "learning_rate": 9.591612903225807e-05,
+      "loss": 0.2123,
+      "step": 3033
+    },
+    {
+      "epoch": 0.048544,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.591451612903226e-05,
+      "loss": 0.2101,
+      "step": 3034
+    },
+    {
+      "epoch": 0.04856,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.591290322580646e-05,
+      "loss": 0.1885,
+      "step": 3035
+    },
+    {
+      "epoch": 0.048576,
+      "grad_norm": 0.48828125,
+      "learning_rate": 9.591129032258064e-05,
+      "loss": 0.1654,
+      "step": 3036
+    },
+    {
+      "epoch": 0.048592,
+      "grad_norm": 0.75,
+      "learning_rate": 9.590967741935484e-05,
+      "loss": 0.1721,
+      "step": 3037
+    },
+    {
+      "epoch": 0.048608,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.590806451612904e-05,
+      "loss": 0.2043,
+      "step": 3038
+    },
+    {
+      "epoch": 0.048624,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.590645161290323e-05,
+      "loss": 0.1706,
+      "step": 3039
+    },
+    {
+      "epoch": 0.04864,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.590483870967743e-05,
+      "loss": 0.1557,
+      "step": 3040
+    },
+    {
+      "epoch": 0.048656,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.590322580645161e-05,
+      "loss": 0.2304,
+      "step": 3041
+    },
+    {
+      "epoch": 0.048672,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.590161290322581e-05,
+      "loss": 0.1904,
+      "step": 3042
+    },
+    {
+      "epoch": 0.048688,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.59e-05,
+      "loss": 0.1923,
+      "step": 3043
+    },
+    {
+      "epoch": 0.048704,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.58983870967742e-05,
+      "loss": 0.1702,
+      "step": 3044
+    },
+    {
+      "epoch": 0.04872,
+      "grad_norm": 1.125,
+      "learning_rate": 9.589677419354839e-05,
+      "loss": 0.1685,
+      "step": 3045
+    },
+    {
+      "epoch": 0.048736,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.589516129032259e-05,
+      "loss": 0.1779,
+      "step": 3046
+    },
+    {
+      "epoch": 0.048752,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.589354838709677e-05,
+      "loss": 0.1957,
+      "step": 3047
+    },
+    {
+      "epoch": 0.048768,
+      "grad_norm": 1.0,
+      "learning_rate": 9.589193548387097e-05,
+      "loss": 0.2087,
+      "step": 3048
+    },
+    {
+      "epoch": 0.048784,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.589032258064517e-05,
+      "loss": 0.1652,
+      "step": 3049
+    },
+    {
+      "epoch": 0.0488,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.588870967741937e-05,
+      "loss": 0.1962,
+      "step": 3050
+    },
+    {
+      "epoch": 0.048816,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.588709677419356e-05,
+      "loss": 0.1621,
+      "step": 3051
+    },
+    {
+      "epoch": 0.048832,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.588548387096776e-05,
+      "loss": 0.1711,
+      "step": 3052
+    },
+    {
+      "epoch": 0.048848,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.588387096774194e-05,
+      "loss": 0.1934,
+      "step": 3053
+    },
+    {
+      "epoch": 0.048864,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.588225806451613e-05,
+      "loss": 0.182,
+      "step": 3054
+    },
+    {
+      "epoch": 0.04888,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.588064516129033e-05,
+      "loss": 0.1858,
+      "step": 3055
+    },
+    {
+      "epoch": 0.048896,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.587903225806451e-05,
+      "loss": 0.2309,
+      "step": 3056
+    },
+    {
+      "epoch": 0.048912,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.587741935483871e-05,
+      "loss": 0.2139,
+      "step": 3057
+    },
+    {
+      "epoch": 0.048928,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.58758064516129e-05,
+      "loss": 0.1632,
+      "step": 3058
+    },
+    {
+      "epoch": 0.048944,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.58741935483871e-05,
+      "loss": 0.1909,
+      "step": 3059
+    },
+    {
+      "epoch": 0.04896,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.587258064516129e-05,
+      "loss": 0.2563,
+      "step": 3060
+    },
+    {
+      "epoch": 0.048976,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.587096774193548e-05,
+      "loss": 0.1752,
+      "step": 3061
+    },
+    {
+      "epoch": 0.048992,
+      "grad_norm": 1.328125,
+      "learning_rate": 9.586935483870968e-05,
+      "loss": 0.1978,
+      "step": 3062
+    },
+    {
+      "epoch": 0.049008,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.586774193548388e-05,
+      "loss": 0.166,
+      "step": 3063
+    },
+    {
+      "epoch": 0.049024,
+      "grad_norm": 1.0,
+      "learning_rate": 9.586612903225807e-05,
+      "loss": 0.1969,
+      "step": 3064
+    },
+    {
+      "epoch": 0.04904,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.586451612903227e-05,
+      "loss": 0.1863,
+      "step": 3065
+    },
+    {
+      "epoch": 0.049056,
+      "grad_norm": 1.375,
+      "learning_rate": 9.586290322580646e-05,
+      "loss": 0.2265,
+      "step": 3066
+    },
+    {
+      "epoch": 0.049072,
+      "grad_norm": 1.484375,
+      "learning_rate": 9.586129032258066e-05,
+      "loss": 0.204,
+      "step": 3067
+    },
+    {
+      "epoch": 0.049088,
+      "grad_norm": 1.359375,
+      "learning_rate": 9.585967741935484e-05,
+      "loss": 0.1859,
+      "step": 3068
+    },
+    {
+      "epoch": 0.049104,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.585806451612903e-05,
+      "loss": 0.2041,
+      "step": 3069
+    },
+    {
+      "epoch": 0.04912,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.585645161290323e-05,
+      "loss": 0.2137,
+      "step": 3070
+    },
+    {
+      "epoch": 0.049136,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.585483870967741e-05,
+      "loss": 0.2523,
+      "step": 3071
+    },
+    {
+      "epoch": 0.049152,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.585322580645161e-05,
+      "loss": 0.1758,
+      "step": 3072
+    },
+    {
+      "epoch": 0.049168,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.585161290322581e-05,
+      "loss": 0.1602,
+      "step": 3073
+    },
+    {
+      "epoch": 0.049184,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.585000000000001e-05,
+      "loss": 0.1965,
+      "step": 3074
+    },
+    {
+      "epoch": 0.0492,
+      "grad_norm": 1.875,
+      "learning_rate": 9.58483870967742e-05,
+      "loss": 0.2018,
+      "step": 3075
+    },
+    {
+      "epoch": 0.049216,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.58467741935484e-05,
+      "loss": 0.2117,
+      "step": 3076
+    },
+    {
+      "epoch": 0.049232,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.584516129032258e-05,
+      "loss": 0.2018,
+      "step": 3077
+    },
+    {
+      "epoch": 0.049248,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.584354838709678e-05,
+      "loss": 0.2308,
+      "step": 3078
+    },
+    {
+      "epoch": 0.049264,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.584193548387097e-05,
+      "loss": 0.2037,
+      "step": 3079
+    },
+    {
+      "epoch": 0.04928,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.584032258064517e-05,
+      "loss": 0.1906,
+      "step": 3080
+    },
+    {
+      "epoch": 0.049296,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.583870967741936e-05,
+      "loss": 0.1842,
+      "step": 3081
+    },
+    {
+      "epoch": 0.049312,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.583709677419356e-05,
+      "loss": 0.1531,
+      "step": 3082
+    },
+    {
+      "epoch": 0.049328,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.583548387096774e-05,
+      "loss": 0.1954,
+      "step": 3083
+    },
+    {
+      "epoch": 0.049344,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.583387096774194e-05,
+      "loss": 0.1839,
+      "step": 3084
+    },
+    {
+      "epoch": 0.04936,
+      "grad_norm": 1.3515625,
+      "learning_rate": 9.583225806451614e-05,
+      "loss": 0.1654,
+      "step": 3085
+    },
+    {
+      "epoch": 0.049376,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.583064516129033e-05,
+      "loss": 0.1922,
+      "step": 3086
+    },
+    {
+      "epoch": 0.049392,
+      "grad_norm": 1.25,
+      "learning_rate": 9.582903225806453e-05,
+      "loss": 0.185,
+      "step": 3087
+    },
+    {
+      "epoch": 0.049408,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.582741935483871e-05,
+      "loss": 0.1597,
+      "step": 3088
+    },
+    {
+      "epoch": 0.049424,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.582580645161291e-05,
+      "loss": 0.1541,
+      "step": 3089
+    },
+    {
+      "epoch": 0.04944,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.58241935483871e-05,
+      "loss": 0.1895,
+      "step": 3090
+    },
+    {
+      "epoch": 0.049456,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.58225806451613e-05,
+      "loss": 0.1864,
+      "step": 3091
+    },
+    {
+      "epoch": 0.049472,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.582096774193548e-05,
+      "loss": 0.1615,
+      "step": 3092
+    },
+    {
+      "epoch": 0.049488,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.581935483870968e-05,
+      "loss": 0.1736,
+      "step": 3093
+    },
+    {
+      "epoch": 0.049504,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.581774193548387e-05,
+      "loss": 0.2397,
+      "step": 3094
+    },
+    {
+      "epoch": 0.04952,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.581612903225807e-05,
+      "loss": 0.2397,
+      "step": 3095
+    },
+    {
+      "epoch": 0.049536,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.581451612903226e-05,
+      "loss": 0.1465,
+      "step": 3096
+    },
+    {
+      "epoch": 0.049552,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.581290322580645e-05,
+      "loss": 0.1966,
+      "step": 3097
+    },
+    {
+      "epoch": 0.049568,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.581129032258065e-05,
+      "loss": 0.2117,
+      "step": 3098
+    },
+    {
+      "epoch": 0.049584,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.580967741935485e-05,
+      "loss": 0.1535,
+      "step": 3099
+    },
+    {
+      "epoch": 0.0496,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.580806451612904e-05,
+      "loss": 0.1781,
+      "step": 3100
+    },
+    {
+      "epoch": 0.049616,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.580645161290323e-05,
+      "loss": 0.2007,
+      "step": 3101
+    },
+    {
+      "epoch": 0.049632,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.580483870967743e-05,
+      "loss": 0.167,
+      "step": 3102
+    },
+    {
+      "epoch": 0.049648,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.580322580645161e-05,
+      "loss": 0.1721,
+      "step": 3103
+    },
+    {
+      "epoch": 0.049664,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.580161290322581e-05,
+      "loss": 0.1489,
+      "step": 3104
+    },
+    {
+      "epoch": 0.04968,
+      "grad_norm": 0.875,
+      "learning_rate": 9.58e-05,
+      "loss": 0.1996,
+      "step": 3105
+    },
+    {
+      "epoch": 0.049696,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.57983870967742e-05,
+      "loss": 0.2335,
+      "step": 3106
+    },
+    {
+      "epoch": 0.049712,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.579677419354838e-05,
+      "loss": 0.2009,
+      "step": 3107
+    },
+    {
+      "epoch": 0.049728,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.579516129032258e-05,
+      "loss": 0.2116,
+      "step": 3108
+    },
+    {
+      "epoch": 0.049744,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.579354838709678e-05,
+      "loss": 0.2131,
+      "step": 3109
+    },
+    {
+      "epoch": 0.04976,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.579193548387098e-05,
+      "loss": 0.1409,
+      "step": 3110
+    },
+    {
+      "epoch": 0.049776,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.579032258064517e-05,
+      "loss": 0.1777,
+      "step": 3111
+    },
+    {
+      "epoch": 0.049792,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.578870967741937e-05,
+      "loss": 0.1866,
+      "step": 3112
+    },
+    {
+      "epoch": 0.049808,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.578709677419355e-05,
+      "loss": 0.18,
+      "step": 3113
+    },
+    {
+      "epoch": 0.049824,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.578548387096775e-05,
+      "loss": 0.1591,
+      "step": 3114
+    },
+    {
+      "epoch": 0.04984,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.578387096774194e-05,
+      "loss": 0.1725,
+      "step": 3115
+    },
+    {
+      "epoch": 0.049856,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.578225806451613e-05,
+      "loss": 0.1941,
+      "step": 3116
+    },
+    {
+      "epoch": 0.049872,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.578064516129033e-05,
+      "loss": 0.1648,
+      "step": 3117
+    },
+    {
+      "epoch": 0.049888,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.577903225806451e-05,
+      "loss": 0.1676,
+      "step": 3118
+    },
+    {
+      "epoch": 0.049904,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.577741935483871e-05,
+      "loss": 0.2091,
+      "step": 3119
+    },
+    {
+      "epoch": 0.04992,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.57758064516129e-05,
+      "loss": 0.1375,
+      "step": 3120
+    },
+    {
+      "epoch": 0.049936,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.57741935483871e-05,
+      "loss": 0.2127,
+      "step": 3121
+    },
+    {
+      "epoch": 0.049952,
+      "grad_norm": 1.5078125,
+      "learning_rate": 9.57725806451613e-05,
+      "loss": 0.1922,
+      "step": 3122
+    },
+    {
+      "epoch": 0.049968,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.57709677419355e-05,
+      "loss": 0.1724,
+      "step": 3123
+    },
+    {
+      "epoch": 0.049984,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.576935483870968e-05,
+      "loss": 0.2059,
+      "step": 3124
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.576774193548388e-05,
+      "loss": 0.1365,
+      "step": 3125
+    },
+    {
+      "epoch": 0.050016,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.576612903225807e-05,
+      "loss": 0.2059,
+      "step": 3126
+    },
+    {
+      "epoch": 0.050032,
+      "grad_norm": 1.25,
+      "learning_rate": 9.576451612903227e-05,
+      "loss": 0.2226,
+      "step": 3127
+    },
+    {
+      "epoch": 0.050048,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.576290322580645e-05,
+      "loss": 0.1529,
+      "step": 3128
+    },
+    {
+      "epoch": 0.050064,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.576129032258065e-05,
+      "loss": 0.1837,
+      "step": 3129
+    },
+    {
+      "epoch": 0.05008,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.575967741935484e-05,
+      "loss": 0.2053,
+      "step": 3130
+    },
+    {
+      "epoch": 0.050096,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.575806451612904e-05,
+      "loss": 0.1853,
+      "step": 3131
+    },
+    {
+      "epoch": 0.050112,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.575645161290322e-05,
+      "loss": 0.1825,
+      "step": 3132
+    },
+    {
+      "epoch": 0.050128,
+      "grad_norm": 1.375,
+      "learning_rate": 9.575483870967742e-05,
+      "loss": 0.2459,
+      "step": 3133
+    },
+    {
+      "epoch": 0.050144,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.575322580645162e-05,
+      "loss": 0.185,
+      "step": 3134
+    },
+    {
+      "epoch": 0.05016,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.575161290322581e-05,
+      "loss": 0.2057,
+      "step": 3135
+    },
+    {
+      "epoch": 0.050176,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.575000000000001e-05,
+      "loss": 0.1806,
+      "step": 3136
+    },
+    {
+      "epoch": 0.050192,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.57483870967742e-05,
+      "loss": 0.1767,
+      "step": 3137
+    },
+    {
+      "epoch": 0.050208,
+      "grad_norm": 1.453125,
+      "learning_rate": 9.57467741935484e-05,
+      "loss": 0.2059,
+      "step": 3138
+    },
+    {
+      "epoch": 0.050224,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.574516129032258e-05,
+      "loss": 0.1901,
+      "step": 3139
+    },
+    {
+      "epoch": 0.05024,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.574354838709678e-05,
+      "loss": 0.2364,
+      "step": 3140
+    },
+    {
+      "epoch": 0.050256,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.574193548387097e-05,
+      "loss": 0.1558,
+      "step": 3141
+    },
+    {
+      "epoch": 0.050272,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.574032258064517e-05,
+      "loss": 0.1833,
+      "step": 3142
+    },
+    {
+      "epoch": 0.050288,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.573870967741935e-05,
+      "loss": 0.1947,
+      "step": 3143
+    },
+    {
+      "epoch": 0.050304,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.573709677419355e-05,
+      "loss": 0.2065,
+      "step": 3144
+    },
+    {
+      "epoch": 0.05032,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.573548387096775e-05,
+      "loss": 0.2016,
+      "step": 3145
+    },
+    {
+      "epoch": 0.050336,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.573387096774195e-05,
+      "loss": 0.1781,
+      "step": 3146
+    },
+    {
+      "epoch": 0.050352,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.573225806451614e-05,
+      "loss": 0.1813,
+      "step": 3147
+    },
+    {
+      "epoch": 0.050368,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.573064516129032e-05,
+      "loss": 0.1796,
+      "step": 3148
+    },
+    {
+      "epoch": 0.050384,
+      "grad_norm": 1.34375,
+      "learning_rate": 9.572903225806452e-05,
+      "loss": 0.1993,
+      "step": 3149
+    },
+    {
+      "epoch": 0.0504,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.572741935483871e-05,
+      "loss": 0.1981,
+      "step": 3150
+    },
+    {
+      "epoch": 0.050416,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.572580645161291e-05,
+      "loss": 0.1943,
+      "step": 3151
+    },
+    {
+      "epoch": 0.050432,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.57241935483871e-05,
+      "loss": 0.1874,
+      "step": 3152
+    },
+    {
+      "epoch": 0.050448,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.57225806451613e-05,
+      "loss": 0.1844,
+      "step": 3153
+    },
+    {
+      "epoch": 0.050464,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.572096774193548e-05,
+      "loss": 0.1591,
+      "step": 3154
+    },
+    {
+      "epoch": 0.05048,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.571935483870968e-05,
+      "loss": 0.1993,
+      "step": 3155
+    },
+    {
+      "epoch": 0.050496,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.571774193548387e-05,
+      "loss": 0.214,
+      "step": 3156
+    },
+    {
+      "epoch": 0.050512,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.571612903225807e-05,
+      "loss": 0.1745,
+      "step": 3157
+    },
+    {
+      "epoch": 0.050528,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.571451612903227e-05,
+      "loss": 0.1774,
+      "step": 3158
+    },
+    {
+      "epoch": 0.050544,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.571290322580647e-05,
+      "loss": 0.166,
+      "step": 3159
+    },
+    {
+      "epoch": 0.05056,
+      "grad_norm": 0.47265625,
+      "learning_rate": 9.571129032258065e-05,
+      "loss": 0.1299,
+      "step": 3160
+    },
+    {
+      "epoch": 0.050576,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.570967741935485e-05,
+      "loss": 0.1986,
+      "step": 3161
+    },
+    {
+      "epoch": 0.050592,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.570806451612904e-05,
+      "loss": 0.2054,
+      "step": 3162
+    },
+    {
+      "epoch": 0.050608,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.570645161290322e-05,
+      "loss": 0.1699,
+      "step": 3163
+    },
+    {
+      "epoch": 0.050624,
+      "grad_norm": 1.125,
+      "learning_rate": 9.570483870967742e-05,
+      "loss": 0.1303,
+      "step": 3164
+    },
+    {
+      "epoch": 0.05064,
+      "grad_norm": 1.0,
+      "learning_rate": 9.570322580645161e-05,
+      "loss": 0.1608,
+      "step": 3165
+    },
+    {
+      "epoch": 0.050656,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.570161290322581e-05,
+      "loss": 0.1988,
+      "step": 3166
+    },
+    {
+      "epoch": 0.050672,
+      "grad_norm": 1.296875,
+      "learning_rate": 9.57e-05,
+      "loss": 0.2332,
+      "step": 3167
+    },
+    {
+      "epoch": 0.050688,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.56983870967742e-05,
+      "loss": 0.1666,
+      "step": 3168
+    },
+    {
+      "epoch": 0.050704,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.56967741935484e-05,
+      "loss": 0.1708,
+      "step": 3169
+    },
+    {
+      "epoch": 0.05072,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.56951612903226e-05,
+      "loss": 0.1724,
+      "step": 3170
+    },
+    {
+      "epoch": 0.050736,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.569354838709678e-05,
+      "loss": 0.2062,
+      "step": 3171
+    },
+    {
+      "epoch": 0.050752,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.569193548387098e-05,
+      "loss": 0.1728,
+      "step": 3172
+    },
+    {
+      "epoch": 0.050768,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.569032258064517e-05,
+      "loss": 0.1915,
+      "step": 3173
+    },
+    {
+      "epoch": 0.050784,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.568870967741937e-05,
+      "loss": 0.204,
+      "step": 3174
+    },
+    {
+      "epoch": 0.0508,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.568709677419355e-05,
+      "loss": 0.2027,
+      "step": 3175
+    },
+    {
+      "epoch": 0.050816,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.568548387096775e-05,
+      "loss": 0.19,
+      "step": 3176
+    },
+    {
+      "epoch": 0.050832,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.568387096774194e-05,
+      "loss": 0.1952,
+      "step": 3177
+    },
+    {
+      "epoch": 0.050848,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.568225806451612e-05,
+      "loss": 0.2092,
+      "step": 3178
+    },
+    {
+      "epoch": 0.050864,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.568064516129032e-05,
+      "loss": 0.148,
+      "step": 3179
+    },
+    {
+      "epoch": 0.05088,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.567903225806452e-05,
+      "loss": 0.1855,
+      "step": 3180
+    },
+    {
+      "epoch": 0.050896,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.567741935483872e-05,
+      "loss": 0.2112,
+      "step": 3181
+    },
+    {
+      "epoch": 0.050912,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.567580645161291e-05,
+      "loss": 0.1876,
+      "step": 3182
+    },
+    {
+      "epoch": 0.050928,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.567419354838711e-05,
+      "loss": 0.1724,
+      "step": 3183
+    },
+    {
+      "epoch": 0.050944,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.56725806451613e-05,
+      "loss": 0.2163,
+      "step": 3184
+    },
+    {
+      "epoch": 0.05096,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.56709677419355e-05,
+      "loss": 0.2648,
+      "step": 3185
+    },
+    {
+      "epoch": 0.050976,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.566935483870968e-05,
+      "loss": 0.1815,
+      "step": 3186
+    },
+    {
+      "epoch": 0.050992,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.566774193548388e-05,
+      "loss": 0.1823,
+      "step": 3187
+    },
+    {
+      "epoch": 0.051008,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.566612903225807e-05,
+      "loss": 0.2279,
+      "step": 3188
+    },
+    {
+      "epoch": 0.051024,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.566451612903227e-05,
+      "loss": 0.189,
+      "step": 3189
+    },
+    {
+      "epoch": 0.05104,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.566290322580645e-05,
+      "loss": 0.1836,
+      "step": 3190
+    },
+    {
+      "epoch": 0.051056,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.566129032258065e-05,
+      "loss": 0.1555,
+      "step": 3191
+    },
+    {
+      "epoch": 0.051072,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.565967741935484e-05,
+      "loss": 0.1721,
+      "step": 3192
+    },
+    {
+      "epoch": 0.051088,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.565806451612904e-05,
+      "loss": 0.1657,
+      "step": 3193
+    },
+    {
+      "epoch": 0.051104,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.565645161290324e-05,
+      "loss": 0.2038,
+      "step": 3194
+    },
+    {
+      "epoch": 0.05112,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.565483870967742e-05,
+      "loss": 0.2042,
+      "step": 3195
+    },
+    {
+      "epoch": 0.051136,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.565322580645162e-05,
+      "loss": 0.1992,
+      "step": 3196
+    },
+    {
+      "epoch": 0.051152,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.565161290322581e-05,
+      "loss": 0.1968,
+      "step": 3197
+    },
+    {
+      "epoch": 0.051168,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.565000000000001e-05,
+      "loss": 0.1444,
+      "step": 3198
+    },
+    {
+      "epoch": 0.051184,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.56483870967742e-05,
+      "loss": 0.1673,
+      "step": 3199
+    },
+    {
+      "epoch": 0.0512,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.564677419354839e-05,
+      "loss": 0.1754,
+      "step": 3200
+    },
+    {
+      "epoch": 0.051216,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.564516129032258e-05,
+      "loss": 0.1934,
+      "step": 3201
+    },
+    {
+      "epoch": 0.051232,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.564354838709678e-05,
+      "loss": 0.1847,
+      "step": 3202
+    },
+    {
+      "epoch": 0.051248,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.564193548387096e-05,
+      "loss": 0.1815,
+      "step": 3203
+    },
+    {
+      "epoch": 0.051264,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.564032258064516e-05,
+      "loss": 0.1617,
+      "step": 3204
+    },
+    {
+      "epoch": 0.05128,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.563870967741936e-05,
+      "loss": 0.1965,
+      "step": 3205
+    },
+    {
+      "epoch": 0.051296,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.563709677419356e-05,
+      "loss": 0.1599,
+      "step": 3206
+    },
+    {
+      "epoch": 0.051312,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.563548387096775e-05,
+      "loss": 0.1636,
+      "step": 3207
+    },
+    {
+      "epoch": 0.051328,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.563387096774195e-05,
+      "loss": 0.1272,
+      "step": 3208
+    },
+    {
+      "epoch": 0.051344,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.563225806451614e-05,
+      "loss": 0.1824,
+      "step": 3209
+    },
+    {
+      "epoch": 0.05136,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.563064516129032e-05,
+      "loss": 0.1909,
+      "step": 3210
+    },
+    {
+      "epoch": 0.051376,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.562903225806452e-05,
+      "loss": 0.1957,
+      "step": 3211
+    },
+    {
+      "epoch": 0.051392,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.562741935483871e-05,
+      "loss": 0.2156,
+      "step": 3212
+    },
+    {
+      "epoch": 0.051408,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.562580645161291e-05,
+      "loss": 0.217,
+      "step": 3213
+    },
+    {
+      "epoch": 0.051424,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.562419354838709e-05,
+      "loss": 0.1902,
+      "step": 3214
+    },
+    {
+      "epoch": 0.05144,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.562258064516129e-05,
+      "loss": 0.1772,
+      "step": 3215
+    },
+    {
+      "epoch": 0.051456,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.562096774193548e-05,
+      "loss": 0.1717,
+      "step": 3216
+    },
+    {
+      "epoch": 0.051472,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.561935483870968e-05,
+      "loss": 0.1764,
+      "step": 3217
+    },
+    {
+      "epoch": 0.051488,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.561774193548388e-05,
+      "loss": 0.1673,
+      "step": 3218
+    },
+    {
+      "epoch": 0.051504,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.561612903225808e-05,
+      "loss": 0.2016,
+      "step": 3219
+    },
+    {
+      "epoch": 0.05152,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.561451612903226e-05,
+      "loss": 0.1858,
+      "step": 3220
+    },
+    {
+      "epoch": 0.051536,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.561290322580646e-05,
+      "loss": 0.2379,
+      "step": 3221
+    },
+    {
+      "epoch": 0.051552,
+      "grad_norm": 1.3828125,
+      "learning_rate": 9.561129032258065e-05,
+      "loss": 0.196,
+      "step": 3222
+    },
+    {
+      "epoch": 0.051568,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.560967741935485e-05,
+      "loss": 0.2006,
+      "step": 3223
+    },
+    {
+      "epoch": 0.051584,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.560806451612904e-05,
+      "loss": 0.1789,
+      "step": 3224
+    },
+    {
+      "epoch": 0.0516,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.560645161290322e-05,
+      "loss": 0.2084,
+      "step": 3225
+    },
+    {
+      "epoch": 0.051616,
+      "grad_norm": 1.34375,
+      "learning_rate": 9.560483870967742e-05,
+      "loss": 0.187,
+      "step": 3226
+    },
+    {
+      "epoch": 0.051632,
+      "grad_norm": 1.4140625,
+      "learning_rate": 9.560322580645161e-05,
+      "loss": 0.1607,
+      "step": 3227
+    },
+    {
+      "epoch": 0.051648,
+      "grad_norm": 0.75,
+      "learning_rate": 9.56016129032258e-05,
+      "loss": 0.1862,
+      "step": 3228
+    },
+    {
+      "epoch": 0.051664,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.56e-05,
+      "loss": 0.1758,
+      "step": 3229
+    },
+    {
+      "epoch": 0.05168,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.55983870967742e-05,
+      "loss": 0.1987,
+      "step": 3230
+    },
+    {
+      "epoch": 0.051696,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.559677419354839e-05,
+      "loss": 0.1818,
+      "step": 3231
+    },
+    {
+      "epoch": 0.051712,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.559516129032259e-05,
+      "loss": 0.1631,
+      "step": 3232
+    },
+    {
+      "epoch": 0.051728,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.559354838709678e-05,
+      "loss": 0.144,
+      "step": 3233
+    },
+    {
+      "epoch": 0.051744,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.559193548387098e-05,
+      "loss": 0.1528,
+      "step": 3234
+    },
+    {
+      "epoch": 0.05176,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.559032258064516e-05,
+      "loss": 0.2231,
+      "step": 3235
+    },
+    {
+      "epoch": 0.051776,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.558870967741936e-05,
+      "loss": 0.224,
+      "step": 3236
+    },
+    {
+      "epoch": 0.051792,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.558709677419355e-05,
+      "loss": 0.1863,
+      "step": 3237
+    },
+    {
+      "epoch": 0.051808,
+      "grad_norm": 1.5078125,
+      "learning_rate": 9.558548387096775e-05,
+      "loss": 0.2397,
+      "step": 3238
+    },
+    {
+      "epoch": 0.051824,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.558387096774193e-05,
+      "loss": 0.2149,
+      "step": 3239
+    },
+    {
+      "epoch": 0.05184,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.558225806451613e-05,
+      "loss": 0.1649,
+      "step": 3240
+    },
+    {
+      "epoch": 0.051856,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.558064516129033e-05,
+      "loss": 0.1723,
+      "step": 3241
+    },
+    {
+      "epoch": 0.051872,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.557903225806452e-05,
+      "loss": 0.1744,
+      "step": 3242
+    },
+    {
+      "epoch": 0.051888,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.557741935483872e-05,
+      "loss": 0.1973,
+      "step": 3243
+    },
+    {
+      "epoch": 0.051904,
+      "grad_norm": 0.75,
+      "learning_rate": 9.55758064516129e-05,
+      "loss": 0.1641,
+      "step": 3244
+    },
+    {
+      "epoch": 0.05192,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.55741935483871e-05,
+      "loss": 0.2192,
+      "step": 3245
+    },
+    {
+      "epoch": 0.051936,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.557258064516129e-05,
+      "loss": 0.2412,
+      "step": 3246
+    },
+    {
+      "epoch": 0.051952,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.557096774193549e-05,
+      "loss": 0.1723,
+      "step": 3247
+    },
+    {
+      "epoch": 0.051968,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.556935483870968e-05,
+      "loss": 0.1762,
+      "step": 3248
+    },
+    {
+      "epoch": 0.051984,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.556774193548388e-05,
+      "loss": 0.1483,
+      "step": 3249
+    },
+    {
+      "epoch": 0.052,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.556612903225806e-05,
+      "loss": 0.2041,
+      "step": 3250
+    },
+    {
+      "epoch": 0.052016,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.556451612903226e-05,
+      "loss": 0.192,
+      "step": 3251
+    },
+    {
+      "epoch": 0.052032,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.556290322580645e-05,
+      "loss": 0.2096,
+      "step": 3252
+    },
+    {
+      "epoch": 0.052048,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.556129032258065e-05,
+      "loss": 0.2428,
+      "step": 3253
+    },
+    {
+      "epoch": 0.052064,
+      "grad_norm": 1.421875,
+      "learning_rate": 9.555967741935485e-05,
+      "loss": 0.2213,
+      "step": 3254
+    },
+    {
+      "epoch": 0.05208,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.555806451612905e-05,
+      "loss": 0.1853,
+      "step": 3255
+    },
+    {
+      "epoch": 0.052096,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.555645161290323e-05,
+      "loss": 0.1872,
+      "step": 3256
+    },
+    {
+      "epoch": 0.052112,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.555483870967742e-05,
+      "loss": 0.175,
+      "step": 3257
+    },
+    {
+      "epoch": 0.052128,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.555322580645162e-05,
+      "loss": 0.1911,
+      "step": 3258
+    },
+    {
+      "epoch": 0.052144,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.55516129032258e-05,
+      "loss": 0.181,
+      "step": 3259
+    },
+    {
+      "epoch": 0.05216,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.555e-05,
+      "loss": 0.1939,
+      "step": 3260
+    },
+    {
+      "epoch": 0.052176,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.554838709677419e-05,
+      "loss": 0.1478,
+      "step": 3261
+    },
+    {
+      "epoch": 0.052192,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.554677419354839e-05,
+      "loss": 0.1539,
+      "step": 3262
+    },
+    {
+      "epoch": 0.052208,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.554516129032258e-05,
+      "loss": 0.1829,
+      "step": 3263
+    },
+    {
+      "epoch": 0.052224,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.554354838709678e-05,
+      "loss": 0.1729,
+      "step": 3264
+    },
+    {
+      "epoch": 0.05224,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.554193548387098e-05,
+      "loss": 0.1864,
+      "step": 3265
+    },
+    {
+      "epoch": 0.052256,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.554032258064518e-05,
+      "loss": 0.1615,
+      "step": 3266
+    },
+    {
+      "epoch": 0.052272,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.553870967741936e-05,
+      "loss": 0.1974,
+      "step": 3267
+    },
+    {
+      "epoch": 0.052288,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.553709677419356e-05,
+      "loss": 0.1659,
+      "step": 3268
+    },
+    {
+      "epoch": 0.052304,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.553548387096775e-05,
+      "loss": 0.1723,
+      "step": 3269
+    },
+    {
+      "epoch": 0.05232,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.553387096774195e-05,
+      "loss": 0.1838,
+      "step": 3270
+    },
+    {
+      "epoch": 0.052336,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.553225806451613e-05,
+      "loss": 0.2007,
+      "step": 3271
+    },
+    {
+      "epoch": 0.052352,
+      "grad_norm": 1.25,
+      "learning_rate": 9.553064516129032e-05,
+      "loss": 0.1897,
+      "step": 3272
+    },
+    {
+      "epoch": 0.052368,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.552903225806452e-05,
+      "loss": 0.229,
+      "step": 3273
+    },
+    {
+      "epoch": 0.052384,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.55274193548387e-05,
+      "loss": 0.1618,
+      "step": 3274
+    },
+    {
+      "epoch": 0.0524,
+      "grad_norm": 1.46875,
+      "learning_rate": 9.55258064516129e-05,
+      "loss": 0.2105,
+      "step": 3275
+    },
+    {
+      "epoch": 0.052416,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.55241935483871e-05,
+      "loss": 0.1521,
+      "step": 3276
+    },
+    {
+      "epoch": 0.052432,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.552258064516129e-05,
+      "loss": 0.1882,
+      "step": 3277
+    },
+    {
+      "epoch": 0.052448,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.552096774193549e-05,
+      "loss": 0.1673,
+      "step": 3278
+    },
+    {
+      "epoch": 0.052464,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.551935483870969e-05,
+      "loss": 0.1808,
+      "step": 3279
+    },
+    {
+      "epoch": 0.05248,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.551774193548388e-05,
+      "loss": 0.229,
+      "step": 3280
+    },
+    {
+      "epoch": 0.052496,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.551612903225808e-05,
+      "loss": 0.1723,
+      "step": 3281
+    },
+    {
+      "epoch": 0.052512,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.551451612903226e-05,
+      "loss": 0.1965,
+      "step": 3282
+    },
+    {
+      "epoch": 0.052528,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.551290322580646e-05,
+      "loss": 0.1893,
+      "step": 3283
+    },
+    {
+      "epoch": 0.052544,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.551129032258065e-05,
+      "loss": 0.187,
+      "step": 3284
+    },
+    {
+      "epoch": 0.05256,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.550967741935485e-05,
+      "loss": 0.1808,
+      "step": 3285
+    },
+    {
+      "epoch": 0.052576,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.550806451612903e-05,
+      "loss": 0.1384,
+      "step": 3286
+    },
+    {
+      "epoch": 0.052592,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.550645161290322e-05,
+      "loss": 0.1796,
+      "step": 3287
+    },
+    {
+      "epoch": 0.052608,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.550483870967742e-05,
+      "loss": 0.1982,
+      "step": 3288
+    },
+    {
+      "epoch": 0.052624,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.550322580645162e-05,
+      "loss": 0.156,
+      "step": 3289
+    },
+    {
+      "epoch": 0.05264,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.550161290322582e-05,
+      "loss": 0.2036,
+      "step": 3290
+    },
+    {
+      "epoch": 0.052656,
+      "grad_norm": 0.625,
+      "learning_rate": 9.55e-05,
+      "loss": 0.1625,
+      "step": 3291
+    },
+    {
+      "epoch": 0.052672,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.54983870967742e-05,
+      "loss": 0.1983,
+      "step": 3292
+    },
+    {
+      "epoch": 0.052688,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.549677419354839e-05,
+      "loss": 0.1999,
+      "step": 3293
+    },
+    {
+      "epoch": 0.052704,
+      "grad_norm": 1.421875,
+      "learning_rate": 9.549516129032259e-05,
+      "loss": 0.1629,
+      "step": 3294
+    },
+    {
+      "epoch": 0.05272,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.549354838709678e-05,
+      "loss": 0.2266,
+      "step": 3295
+    },
+    {
+      "epoch": 0.052736,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.549193548387097e-05,
+      "loss": 0.2082,
+      "step": 3296
+    },
+    {
+      "epoch": 0.052752,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.549032258064516e-05,
+      "loss": 0.1607,
+      "step": 3297
+    },
+    {
+      "epoch": 0.052768,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.548870967741936e-05,
+      "loss": 0.1648,
+      "step": 3298
+    },
+    {
+      "epoch": 0.052784,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.548709677419355e-05,
+      "loss": 0.1875,
+      "step": 3299
+    },
+    {
+      "epoch": 0.0528,
+      "grad_norm": 1.421875,
+      "learning_rate": 9.548548387096775e-05,
+      "loss": 0.1971,
+      "step": 3300
+    },
+    {
+      "epoch": 0.052816,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.548387096774195e-05,
+      "loss": 0.1733,
+      "step": 3301
+    },
+    {
+      "epoch": 0.052832,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.548225806451615e-05,
+      "loss": 0.1614,
+      "step": 3302
+    },
+    {
+      "epoch": 0.052848,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.548064516129033e-05,
+      "loss": 0.2337,
+      "step": 3303
+    },
+    {
+      "epoch": 0.052864,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.547903225806452e-05,
+      "loss": 0.2293,
+      "step": 3304
+    },
+    {
+      "epoch": 0.05288,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.547741935483872e-05,
+      "loss": 0.1831,
+      "step": 3305
+    },
+    {
+      "epoch": 0.052896,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.54758064516129e-05,
+      "loss": 0.1849,
+      "step": 3306
+    },
+    {
+      "epoch": 0.052912,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.54741935483871e-05,
+      "loss": 0.1518,
+      "step": 3307
+    },
+    {
+      "epoch": 0.052928,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.547258064516129e-05,
+      "loss": 0.1585,
+      "step": 3308
+    },
+    {
+      "epoch": 0.052944,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.547096774193549e-05,
+      "loss": 0.1748,
+      "step": 3309
+    },
+    {
+      "epoch": 0.05296,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.546935483870967e-05,
+      "loss": 0.1578,
+      "step": 3310
+    },
+    {
+      "epoch": 0.052976,
+      "grad_norm": 0.875,
+      "learning_rate": 9.546774193548387e-05,
+      "loss": 0.1942,
+      "step": 3311
+    },
+    {
+      "epoch": 0.052992,
+      "grad_norm": 1.421875,
+      "learning_rate": 9.546612903225806e-05,
+      "loss": 0.1947,
+      "step": 3312
+    },
+    {
+      "epoch": 0.053008,
+      "grad_norm": 1.0,
+      "learning_rate": 9.546451612903226e-05,
+      "loss": 0.2015,
+      "step": 3313
+    },
+    {
+      "epoch": 0.053024,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.546290322580646e-05,
+      "loss": 0.1833,
+      "step": 3314
+    },
+    {
+      "epoch": 0.05304,
+      "grad_norm": 1.125,
+      "learning_rate": 9.546129032258066e-05,
+      "loss": 0.1739,
+      "step": 3315
+    },
+    {
+      "epoch": 0.053056,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.545967741935485e-05,
+      "loss": 0.1834,
+      "step": 3316
+    },
+    {
+      "epoch": 0.053072,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.545806451612905e-05,
+      "loss": 0.19,
+      "step": 3317
+    },
+    {
+      "epoch": 0.053088,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.545645161290323e-05,
+      "loss": 0.147,
+      "step": 3318
+    },
+    {
+      "epoch": 0.053104,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.545483870967742e-05,
+      "loss": 0.2068,
+      "step": 3319
+    },
+    {
+      "epoch": 0.05312,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.545322580645162e-05,
+      "loss": 0.1797,
+      "step": 3320
+    },
+    {
+      "epoch": 0.053136,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.54516129032258e-05,
+      "loss": 0.2193,
+      "step": 3321
+    },
+    {
+      "epoch": 0.053152,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.545e-05,
+      "loss": 0.203,
+      "step": 3322
+    },
+    {
+      "epoch": 0.053168,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.544838709677419e-05,
+      "loss": 0.2072,
+      "step": 3323
+    },
+    {
+      "epoch": 0.053184,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.544677419354839e-05,
+      "loss": 0.1631,
+      "step": 3324
+    },
+    {
+      "epoch": 0.0532,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.544516129032259e-05,
+      "loss": 0.1608,
+      "step": 3325
+    },
+    {
+      "epoch": 0.053216,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.544354838709679e-05,
+      "loss": 0.2263,
+      "step": 3326
+    },
+    {
+      "epoch": 0.053232,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.544193548387097e-05,
+      "loss": 0.2338,
+      "step": 3327
+    },
+    {
+      "epoch": 0.053248,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.544032258064517e-05,
+      "loss": 0.2069,
+      "step": 3328
+    },
+    {
+      "epoch": 0.053264,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.543870967741936e-05,
+      "loss": 0.1566,
+      "step": 3329
+    },
+    {
+      "epoch": 0.05328,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.543709677419356e-05,
+      "loss": 0.2174,
+      "step": 3330
+    },
+    {
+      "epoch": 0.053296,
+      "grad_norm": 1.4765625,
+      "learning_rate": 9.543548387096775e-05,
+      "loss": 0.2009,
+      "step": 3331
+    },
+    {
+      "epoch": 0.053312,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.543387096774194e-05,
+      "loss": 0.1766,
+      "step": 3332
+    },
+    {
+      "epoch": 0.053328,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.543225806451613e-05,
+      "loss": 0.1841,
+      "step": 3333
+    },
+    {
+      "epoch": 0.053344,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.543064516129032e-05,
+      "loss": 0.2023,
+      "step": 3334
+    },
+    {
+      "epoch": 0.05336,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.542903225806452e-05,
+      "loss": 0.2073,
+      "step": 3335
+    },
+    {
+      "epoch": 0.053376,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.542741935483872e-05,
+      "loss": 0.1539,
+      "step": 3336
+    },
+    {
+      "epoch": 0.053392,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.542580645161292e-05,
+      "loss": 0.1501,
+      "step": 3337
+    },
+    {
+      "epoch": 0.053408,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.54241935483871e-05,
+      "loss": 0.2013,
+      "step": 3338
+    },
+    {
+      "epoch": 0.053424,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.54225806451613e-05,
+      "loss": 0.154,
+      "step": 3339
+    },
+    {
+      "epoch": 0.05344,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.542096774193549e-05,
+      "loss": 0.195,
+      "step": 3340
+    },
+    {
+      "epoch": 0.053456,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.541935483870969e-05,
+      "loss": 0.2172,
+      "step": 3341
+    },
+    {
+      "epoch": 0.053472,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.541774193548387e-05,
+      "loss": 0.1873,
+      "step": 3342
+    },
+    {
+      "epoch": 0.053488,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.541612903225807e-05,
+      "loss": 0.1661,
+      "step": 3343
+    },
+    {
+      "epoch": 0.053504,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.541451612903226e-05,
+      "loss": 0.1483,
+      "step": 3344
+    },
+    {
+      "epoch": 0.05352,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.541290322580646e-05,
+      "loss": 0.2019,
+      "step": 3345
+    },
+    {
+      "epoch": 0.053536,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.541129032258064e-05,
+      "loss": 0.2273,
+      "step": 3346
+    },
+    {
+      "epoch": 0.053552,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.540967741935484e-05,
+      "loss": 0.229,
+      "step": 3347
+    },
+    {
+      "epoch": 0.053568,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.540806451612903e-05,
+      "loss": 0.1534,
+      "step": 3348
+    },
+    {
+      "epoch": 0.053584,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.540645161290323e-05,
+      "loss": 0.1739,
+      "step": 3349
+    },
+    {
+      "epoch": 0.0536,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.540483870967743e-05,
+      "loss": 0.1822,
+      "step": 3350
+    },
+    {
+      "epoch": 0.053616,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.540322580645162e-05,
+      "loss": 0.1781,
+      "step": 3351
+    },
+    {
+      "epoch": 0.053632,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.540161290322582e-05,
+      "loss": 0.1533,
+      "step": 3352
+    },
+    {
+      "epoch": 0.053648,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.54e-05,
+      "loss": 0.1869,
+      "step": 3353
+    },
+    {
+      "epoch": 0.053664,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.53983870967742e-05,
+      "loss": 0.2015,
+      "step": 3354
+    },
+    {
+      "epoch": 0.05368,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.539677419354839e-05,
+      "loss": 0.2141,
+      "step": 3355
+    },
+    {
+      "epoch": 0.053696,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.539516129032259e-05,
+      "loss": 0.1892,
+      "step": 3356
+    },
+    {
+      "epoch": 0.053712,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.539354838709677e-05,
+      "loss": 0.1871,
+      "step": 3357
+    },
+    {
+      "epoch": 0.053728,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.539193548387097e-05,
+      "loss": 0.1879,
+      "step": 3358
+    },
+    {
+      "epoch": 0.053744,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.539032258064516e-05,
+      "loss": 0.1743,
+      "step": 3359
+    },
+    {
+      "epoch": 0.05376,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.538870967741936e-05,
+      "loss": 0.2037,
+      "step": 3360
+    },
+    {
+      "epoch": 0.053776,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.538709677419356e-05,
+      "loss": 0.1747,
+      "step": 3361
+    },
+    {
+      "epoch": 0.053792,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.538548387096776e-05,
+      "loss": 0.1533,
+      "step": 3362
+    },
+    {
+      "epoch": 0.053808,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.538387096774194e-05,
+      "loss": 0.2209,
+      "step": 3363
+    },
+    {
+      "epoch": 0.053824,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.538225806451614e-05,
+      "loss": 0.1671,
+      "step": 3364
+    },
+    {
+      "epoch": 0.05384,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.538064516129033e-05,
+      "loss": 0.1735,
+      "step": 3365
+    },
+    {
+      "epoch": 0.053856,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.537903225806452e-05,
+      "loss": 0.1736,
+      "step": 3366
+    },
+    {
+      "epoch": 0.053872,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.537741935483871e-05,
+      "loss": 0.1574,
+      "step": 3367
+    },
+    {
+      "epoch": 0.053888,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.53758064516129e-05,
+      "loss": 0.1757,
+      "step": 3368
+    },
+    {
+      "epoch": 0.053904,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.53741935483871e-05,
+      "loss": 0.1381,
+      "step": 3369
+    },
+    {
+      "epoch": 0.05392,
+      "grad_norm": 1.25,
+      "learning_rate": 9.537258064516129e-05,
+      "loss": 0.1652,
+      "step": 3370
+    },
+    {
+      "epoch": 0.053936,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.537096774193549e-05,
+      "loss": 0.1463,
+      "step": 3371
+    },
+    {
+      "epoch": 0.053952,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.536935483870967e-05,
+      "loss": 0.1621,
+      "step": 3372
+    },
+    {
+      "epoch": 0.053968,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.536774193548387e-05,
+      "loss": 0.17,
+      "step": 3373
+    },
+    {
+      "epoch": 0.053984,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.536612903225807e-05,
+      "loss": 0.2068,
+      "step": 3374
+    },
+    {
+      "epoch": 0.054,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.536451612903227e-05,
+      "loss": 0.1958,
+      "step": 3375
+    },
+    {
+      "epoch": 0.054016,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.536290322580646e-05,
+      "loss": 0.1509,
+      "step": 3376
+    },
+    {
+      "epoch": 0.054032,
+      "grad_norm": 1.484375,
+      "learning_rate": 9.536129032258066e-05,
+      "loss": 0.2042,
+      "step": 3377
+    },
+    {
+      "epoch": 0.054048,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.535967741935484e-05,
+      "loss": 0.1901,
+      "step": 3378
+    },
+    {
+      "epoch": 0.054064,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.535806451612904e-05,
+      "loss": 0.1624,
+      "step": 3379
+    },
+    {
+      "epoch": 0.05408,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.535645161290323e-05,
+      "loss": 0.2392,
+      "step": 3380
+    },
+    {
+      "epoch": 0.054096,
+      "grad_norm": 1.75,
+      "learning_rate": 9.535483870967741e-05,
+      "loss": 0.1942,
+      "step": 3381
+    },
+    {
+      "epoch": 0.054112,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.535322580645161e-05,
+      "loss": 0.1665,
+      "step": 3382
+    },
+    {
+      "epoch": 0.054128,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.53516129032258e-05,
+      "loss": 0.232,
+      "step": 3383
+    },
+    {
+      "epoch": 0.054144,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.535e-05,
+      "loss": 0.1804,
+      "step": 3384
+    },
+    {
+      "epoch": 0.05416,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.53483870967742e-05,
+      "loss": 0.1321,
+      "step": 3385
+    },
+    {
+      "epoch": 0.054176,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.53467741935484e-05,
+      "loss": 0.167,
+      "step": 3386
+    },
+    {
+      "epoch": 0.054192,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.534516129032259e-05,
+      "loss": 0.1849,
+      "step": 3387
+    },
+    {
+      "epoch": 0.054208,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.534354838709679e-05,
+      "loss": 0.1834,
+      "step": 3388
+    },
+    {
+      "epoch": 0.054224,
+      "grad_norm": 1.4453125,
+      "learning_rate": 9.534193548387097e-05,
+      "loss": 0.228,
+      "step": 3389
+    },
+    {
+      "epoch": 0.05424,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.534032258064517e-05,
+      "loss": 0.1933,
+      "step": 3390
+    },
+    {
+      "epoch": 0.054256,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.533870967741936e-05,
+      "loss": 0.1486,
+      "step": 3391
+    },
+    {
+      "epoch": 0.054272,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.533709677419356e-05,
+      "loss": 0.2146,
+      "step": 3392
+    },
+    {
+      "epoch": 0.054288,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.533548387096774e-05,
+      "loss": 0.2158,
+      "step": 3393
+    },
+    {
+      "epoch": 0.054304,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.533387096774194e-05,
+      "loss": 0.2133,
+      "step": 3394
+    },
+    {
+      "epoch": 0.05432,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.533225806451613e-05,
+      "loss": 0.1807,
+      "step": 3395
+    },
+    {
+      "epoch": 0.054336,
+      "grad_norm": 1.359375,
+      "learning_rate": 9.533064516129033e-05,
+      "loss": 0.2218,
+      "step": 3396
+    },
+    {
+      "epoch": 0.054352,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.532903225806453e-05,
+      "loss": 0.171,
+      "step": 3397
+    },
+    {
+      "epoch": 0.054368,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.532741935483871e-05,
+      "loss": 0.1916,
+      "step": 3398
+    },
+    {
+      "epoch": 0.054384,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.532580645161291e-05,
+      "loss": 0.1987,
+      "step": 3399
+    },
+    {
+      "epoch": 0.0544,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.53241935483871e-05,
+      "loss": 0.1918,
+      "step": 3400
+    },
+    {
+      "epoch": 0.054416,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.53225806451613e-05,
+      "loss": 0.1936,
+      "step": 3401
+    },
+    {
+      "epoch": 0.054432,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.532096774193549e-05,
+      "loss": 0.1591,
+      "step": 3402
+    },
+    {
+      "epoch": 0.054448,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.531935483870968e-05,
+      "loss": 0.185,
+      "step": 3403
+    },
+    {
+      "epoch": 0.054464,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.531774193548387e-05,
+      "loss": 0.163,
+      "step": 3404
+    },
+    {
+      "epoch": 0.05448,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.531612903225807e-05,
+      "loss": 0.1366,
+      "step": 3405
+    },
+    {
+      "epoch": 0.054496,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.531451612903226e-05,
+      "loss": 0.2213,
+      "step": 3406
+    },
+    {
+      "epoch": 0.054512,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.531290322580646e-05,
+      "loss": 0.1651,
+      "step": 3407
+    },
+    {
+      "epoch": 0.054528,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.531129032258064e-05,
+      "loss": 0.1903,
+      "step": 3408
+    },
+    {
+      "epoch": 0.054544,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.530967741935484e-05,
+      "loss": 0.2283,
+      "step": 3409
+    },
+    {
+      "epoch": 0.05456,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.530806451612904e-05,
+      "loss": 0.1651,
+      "step": 3410
+    },
+    {
+      "epoch": 0.054576,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.530645161290324e-05,
+      "loss": 0.1477,
+      "step": 3411
+    },
+    {
+      "epoch": 0.054592,
+      "grad_norm": 1.4609375,
+      "learning_rate": 9.530483870967743e-05,
+      "loss": 0.221,
+      "step": 3412
+    },
+    {
+      "epoch": 0.054608,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.530322580645161e-05,
+      "loss": 0.2001,
+      "step": 3413
+    },
+    {
+      "epoch": 0.054624,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.530161290322581e-05,
+      "loss": 0.2184,
+      "step": 3414
+    },
+    {
+      "epoch": 0.05464,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.53e-05,
+      "loss": 0.1347,
+      "step": 3415
+    },
+    {
+      "epoch": 0.054656,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.52983870967742e-05,
+      "loss": 0.1817,
+      "step": 3416
+    },
+    {
+      "epoch": 0.054672,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.529677419354838e-05,
+      "loss": 0.1966,
+      "step": 3417
+    },
+    {
+      "epoch": 0.054688,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.529516129032258e-05,
+      "loss": 0.2183,
+      "step": 3418
+    },
+    {
+      "epoch": 0.054704,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.529354838709677e-05,
+      "loss": 0.2098,
+      "step": 3419
+    },
+    {
+      "epoch": 0.05472,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.529193548387097e-05,
+      "loss": 0.1897,
+      "step": 3420
+    },
+    {
+      "epoch": 0.054736,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.529032258064517e-05,
+      "loss": 0.1493,
+      "step": 3421
+    },
+    {
+      "epoch": 0.054752,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.528870967741937e-05,
+      "loss": 0.171,
+      "step": 3422
+    },
+    {
+      "epoch": 0.054768,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.528709677419356e-05,
+      "loss": 0.2009,
+      "step": 3423
+    },
+    {
+      "epoch": 0.054784,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.528548387096775e-05,
+      "loss": 0.1823,
+      "step": 3424
+    },
+    {
+      "epoch": 0.0548,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.528387096774194e-05,
+      "loss": 0.1655,
+      "step": 3425
+    },
+    {
+      "epoch": 0.054816,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.528225806451614e-05,
+      "loss": 0.164,
+      "step": 3426
+    },
+    {
+      "epoch": 0.054832,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.528064516129033e-05,
+      "loss": 0.2269,
+      "step": 3427
+    },
+    {
+      "epoch": 0.054848,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.527903225806451e-05,
+      "loss": 0.19,
+      "step": 3428
+    },
+    {
+      "epoch": 0.054864,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.527741935483871e-05,
+      "loss": 0.1718,
+      "step": 3429
+    },
+    {
+      "epoch": 0.05488,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.52758064516129e-05,
+      "loss": 0.1659,
+      "step": 3430
+    },
+    {
+      "epoch": 0.054896,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.52741935483871e-05,
+      "loss": 0.2216,
+      "step": 3431
+    },
+    {
+      "epoch": 0.054912,
+      "grad_norm": 1.4375,
+      "learning_rate": 9.52725806451613e-05,
+      "loss": 0.1868,
+      "step": 3432
+    },
+    {
+      "epoch": 0.054928,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.52709677419355e-05,
+      "loss": 0.1427,
+      "step": 3433
+    },
+    {
+      "epoch": 0.054944,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.526935483870968e-05,
+      "loss": 0.2076,
+      "step": 3434
+    },
+    {
+      "epoch": 0.05496,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.526774193548388e-05,
+      "loss": 0.1974,
+      "step": 3435
+    },
+    {
+      "epoch": 0.054976,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.526612903225807e-05,
+      "loss": 0.1889,
+      "step": 3436
+    },
+    {
+      "epoch": 0.054992,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.526451612903227e-05,
+      "loss": 0.1678,
+      "step": 3437
+    },
+    {
+      "epoch": 0.055008,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.526290322580645e-05,
+      "loss": 0.1595,
+      "step": 3438
+    },
+    {
+      "epoch": 0.055024,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.526129032258065e-05,
+      "loss": 0.1793,
+      "step": 3439
+    },
+    {
+      "epoch": 0.05504,
+      "grad_norm": 1.375,
+      "learning_rate": 9.525967741935484e-05,
+      "loss": 0.1703,
+      "step": 3440
+    },
+    {
+      "epoch": 0.055056,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.525806451612904e-05,
+      "loss": 0.1884,
+      "step": 3441
+    },
+    {
+      "epoch": 0.055072,
+      "grad_norm": 1.25,
+      "learning_rate": 9.525645161290323e-05,
+      "loss": 0.2299,
+      "step": 3442
+    },
+    {
+      "epoch": 0.055088,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.525483870967741e-05,
+      "loss": 0.1783,
+      "step": 3443
+    },
+    {
+      "epoch": 0.055104,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.525322580645161e-05,
+      "loss": 0.1564,
+      "step": 3444
+    },
+    {
+      "epoch": 0.05512,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.525161290322581e-05,
+      "loss": 0.1046,
+      "step": 3445
+    },
+    {
+      "epoch": 0.055136,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.525000000000001e-05,
+      "loss": 0.181,
+      "step": 3446
+    },
+    {
+      "epoch": 0.055152,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.52483870967742e-05,
+      "loss": 0.2248,
+      "step": 3447
+    },
+    {
+      "epoch": 0.055168,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.52467741935484e-05,
+      "loss": 0.2325,
+      "step": 3448
+    },
+    {
+      "epoch": 0.055184,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.524516129032258e-05,
+      "loss": 0.1996,
+      "step": 3449
+    },
+    {
+      "epoch": 0.0552,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.524354838709678e-05,
+      "loss": 0.1976,
+      "step": 3450
+    },
+    {
+      "epoch": 0.055216,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.524193548387097e-05,
+      "loss": 0.1911,
+      "step": 3451
+    },
+    {
+      "epoch": 0.055232,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.524032258064517e-05,
+      "loss": 0.2269,
+      "step": 3452
+    },
+    {
+      "epoch": 0.055248,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.523870967741935e-05,
+      "loss": 0.2194,
+      "step": 3453
+    },
+    {
+      "epoch": 0.055264,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.523709677419355e-05,
+      "loss": 0.1434,
+      "step": 3454
+    },
+    {
+      "epoch": 0.05528,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.523548387096774e-05,
+      "loss": 0.1851,
+      "step": 3455
+    },
+    {
+      "epoch": 0.055296,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.523387096774194e-05,
+      "loss": 0.1785,
+      "step": 3456
+    },
+    {
+      "epoch": 0.055312,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.523225806451614e-05,
+      "loss": 0.1916,
+      "step": 3457
+    },
+    {
+      "epoch": 0.055328,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.523064516129033e-05,
+      "loss": 0.1828,
+      "step": 3458
+    },
+    {
+      "epoch": 0.055344,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.522903225806453e-05,
+      "loss": 0.1574,
+      "step": 3459
+    },
+    {
+      "epoch": 0.05536,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.522741935483871e-05,
+      "loss": 0.199,
+      "step": 3460
+    },
+    {
+      "epoch": 0.055376,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.522580645161291e-05,
+      "loss": 0.1418,
+      "step": 3461
+    },
+    {
+      "epoch": 0.055392,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.52241935483871e-05,
+      "loss": 0.1935,
+      "step": 3462
+    },
+    {
+      "epoch": 0.055408,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.52225806451613e-05,
+      "loss": 0.2657,
+      "step": 3463
+    },
+    {
+      "epoch": 0.055424,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.522096774193548e-05,
+      "loss": 0.1727,
+      "step": 3464
+    },
+    {
+      "epoch": 0.05544,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.521935483870968e-05,
+      "loss": 0.2032,
+      "step": 3465
+    },
+    {
+      "epoch": 0.055456,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.521774193548387e-05,
+      "loss": 0.1481,
+      "step": 3466
+    },
+    {
+      "epoch": 0.055472,
+      "grad_norm": 1.5390625,
+      "learning_rate": 9.521612903225807e-05,
+      "loss": 0.1857,
+      "step": 3467
+    },
+    {
+      "epoch": 0.055488,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.521451612903225e-05,
+      "loss": 0.1518,
+      "step": 3468
+    },
+    {
+      "epoch": 0.055504,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.521290322580645e-05,
+      "loss": 0.195,
+      "step": 3469
+    },
+    {
+      "epoch": 0.05552,
+      "grad_norm": 1.0,
+      "learning_rate": 9.521129032258065e-05,
+      "loss": 0.1584,
+      "step": 3470
+    },
+    {
+      "epoch": 0.055536,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.520967741935485e-05,
+      "loss": 0.1513,
+      "step": 3471
+    },
+    {
+      "epoch": 0.055552,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.520806451612904e-05,
+      "loss": 0.184,
+      "step": 3472
+    },
+    {
+      "epoch": 0.055568,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.520645161290324e-05,
+      "loss": 0.1692,
+      "step": 3473
+    },
+    {
+      "epoch": 0.055584,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.520483870967742e-05,
+      "loss": 0.1668,
+      "step": 3474
+    },
+    {
+      "epoch": 0.0556,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.520322580645161e-05,
+      "loss": 0.1626,
+      "step": 3475
+    },
+    {
+      "epoch": 0.055616,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.520161290322581e-05,
+      "loss": 0.1899,
+      "step": 3476
+    },
+    {
+      "epoch": 0.055632,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.52e-05,
+      "loss": 0.1901,
+      "step": 3477
+    },
+    {
+      "epoch": 0.055648,
+      "grad_norm": 1.3515625,
+      "learning_rate": 9.51983870967742e-05,
+      "loss": 0.1637,
+      "step": 3478
+    },
+    {
+      "epoch": 0.055664,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.519677419354838e-05,
+      "loss": 0.1661,
+      "step": 3479
+    },
+    {
+      "epoch": 0.05568,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.519516129032258e-05,
+      "loss": 0.2348,
+      "step": 3480
+    },
+    {
+      "epoch": 0.055696,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.519354838709678e-05,
+      "loss": 0.1416,
+      "step": 3481
+    },
+    {
+      "epoch": 0.055712,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.519193548387098e-05,
+      "loss": 0.1993,
+      "step": 3482
+    },
+    {
+      "epoch": 0.055728,
+      "grad_norm": 1.296875,
+      "learning_rate": 9.519032258064517e-05,
+      "loss": 0.22,
+      "step": 3483
+    },
+    {
+      "epoch": 0.055744,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.518870967741937e-05,
+      "loss": 0.1433,
+      "step": 3484
+    },
+    {
+      "epoch": 0.05576,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.518709677419355e-05,
+      "loss": 0.1488,
+      "step": 3485
+    },
+    {
+      "epoch": 0.055776,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.518548387096775e-05,
+      "loss": 0.1799,
+      "step": 3486
+    },
+    {
+      "epoch": 0.055792,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.518387096774194e-05,
+      "loss": 0.187,
+      "step": 3487
+    },
+    {
+      "epoch": 0.055808,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.518225806451614e-05,
+      "loss": 0.1775,
+      "step": 3488
+    },
+    {
+      "epoch": 0.055824,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.518064516129032e-05,
+      "loss": 0.1592,
+      "step": 3489
+    },
+    {
+      "epoch": 0.05584,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.517903225806451e-05,
+      "loss": 0.1901,
+      "step": 3490
+    },
+    {
+      "epoch": 0.055856,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.517741935483871e-05,
+      "loss": 0.21,
+      "step": 3491
+    },
+    {
+      "epoch": 0.055872,
+      "grad_norm": 1.421875,
+      "learning_rate": 9.517580645161291e-05,
+      "loss": 0.2126,
+      "step": 3492
+    },
+    {
+      "epoch": 0.055888,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.517419354838711e-05,
+      "loss": 0.2192,
+      "step": 3493
+    },
+    {
+      "epoch": 0.055904,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.51725806451613e-05,
+      "loss": 0.1388,
+      "step": 3494
+    },
+    {
+      "epoch": 0.05592,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.51709677419355e-05,
+      "loss": 0.1968,
+      "step": 3495
+    },
+    {
+      "epoch": 0.055936,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.516935483870968e-05,
+      "loss": 0.2005,
+      "step": 3496
+    },
+    {
+      "epoch": 0.055952,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.516774193548388e-05,
+      "loss": 0.1805,
+      "step": 3497
+    },
+    {
+      "epoch": 0.055968,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.516612903225807e-05,
+      "loss": 0.176,
+      "step": 3498
+    },
+    {
+      "epoch": 0.055984,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.516451612903227e-05,
+      "loss": 0.1881,
+      "step": 3499
+    },
+    {
+      "epoch": 0.056,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.516290322580645e-05,
+      "loss": 0.1803,
+      "step": 3500
+    },
+    {
+      "epoch": 0.056016,
+      "grad_norm": 1.3125,
+      "learning_rate": 9.516129032258065e-05,
+      "loss": 0.1645,
+      "step": 3501
+    },
+    {
+      "epoch": 0.056032,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.515967741935484e-05,
+      "loss": 0.1569,
+      "step": 3502
+    },
+    {
+      "epoch": 0.056048,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.515806451612904e-05,
+      "loss": 0.1797,
+      "step": 3503
+    },
+    {
+      "epoch": 0.056064,
+      "grad_norm": 1.328125,
+      "learning_rate": 9.515645161290322e-05,
+      "loss": 0.1208,
+      "step": 3504
+    },
+    {
+      "epoch": 0.05608,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.515483870967742e-05,
+      "loss": 0.1707,
+      "step": 3505
+    },
+    {
+      "epoch": 0.056096,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.515322580645162e-05,
+      "loss": 0.2071,
+      "step": 3506
+    },
+    {
+      "epoch": 0.056112,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.515161290322581e-05,
+      "loss": 0.1824,
+      "step": 3507
+    },
+    {
+      "epoch": 0.056128,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.515000000000001e-05,
+      "loss": 0.2152,
+      "step": 3508
+    },
+    {
+      "epoch": 0.056144,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.51483870967742e-05,
+      "loss": 0.1666,
+      "step": 3509
+    },
+    {
+      "epoch": 0.05616,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.51467741935484e-05,
+      "loss": 0.1602,
+      "step": 3510
+    },
+    {
+      "epoch": 0.056176,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.514516129032258e-05,
+      "loss": 0.2069,
+      "step": 3511
+    },
+    {
+      "epoch": 0.056192,
+      "grad_norm": 1.125,
+      "learning_rate": 9.514354838709678e-05,
+      "loss": 0.1859,
+      "step": 3512
+    },
+    {
+      "epoch": 0.056208,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.514193548387097e-05,
+      "loss": 0.1594,
+      "step": 3513
+    },
+    {
+      "epoch": 0.056224,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.514032258064517e-05,
+      "loss": 0.1528,
+      "step": 3514
+    },
+    {
+      "epoch": 0.05624,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.513870967741935e-05,
+      "loss": 0.1801,
+      "step": 3515
+    },
+    {
+      "epoch": 0.056256,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.513709677419355e-05,
+      "loss": 0.177,
+      "step": 3516
+    },
+    {
+      "epoch": 0.056272,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.513548387096775e-05,
+      "loss": 0.1582,
+      "step": 3517
+    },
+    {
+      "epoch": 0.056288,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.513387096774195e-05,
+      "loss": 0.1565,
+      "step": 3518
+    },
+    {
+      "epoch": 0.056304,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.513225806451614e-05,
+      "loss": 0.1871,
+      "step": 3519
+    },
+    {
+      "epoch": 0.05632,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.513064516129034e-05,
+      "loss": 0.1546,
+      "step": 3520
+    },
+    {
+      "epoch": 0.056336,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.512903225806452e-05,
+      "loss": 0.19,
+      "step": 3521
+    },
+    {
+      "epoch": 0.056352,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.512741935483871e-05,
+      "loss": 0.1935,
+      "step": 3522
+    },
+    {
+      "epoch": 0.056368,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.512580645161291e-05,
+      "loss": 0.1618,
+      "step": 3523
+    },
+    {
+      "epoch": 0.056384,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.51241935483871e-05,
+      "loss": 0.1839,
+      "step": 3524
+    },
+    {
+      "epoch": 0.0564,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.51225806451613e-05,
+      "loss": 0.2019,
+      "step": 3525
+    },
+    {
+      "epoch": 0.056416,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.512096774193548e-05,
+      "loss": 0.2208,
+      "step": 3526
+    },
+    {
+      "epoch": 0.056432,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.511935483870968e-05,
+      "loss": 0.1731,
+      "step": 3527
+    },
+    {
+      "epoch": 0.056448,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.511774193548388e-05,
+      "loss": 0.1995,
+      "step": 3528
+    },
+    {
+      "epoch": 0.056464,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.511612903225807e-05,
+      "loss": 0.2396,
+      "step": 3529
+    },
+    {
+      "epoch": 0.05648,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.511451612903227e-05,
+      "loss": 0.1854,
+      "step": 3530
+    },
+    {
+      "epoch": 0.056496,
+      "grad_norm": 1.125,
+      "learning_rate": 9.511290322580646e-05,
+      "loss": 0.1754,
+      "step": 3531
+    },
+    {
+      "epoch": 0.056512,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.511129032258065e-05,
+      "loss": 0.1968,
+      "step": 3532
+    },
+    {
+      "epoch": 0.056528,
+      "grad_norm": 1.4375,
+      "learning_rate": 9.510967741935485e-05,
+      "loss": 0.1713,
+      "step": 3533
+    },
+    {
+      "epoch": 0.056544,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.510806451612904e-05,
+      "loss": 0.2232,
+      "step": 3534
+    },
+    {
+      "epoch": 0.05656,
+      "grad_norm": 1.328125,
+      "learning_rate": 9.510645161290324e-05,
+      "loss": 0.2106,
+      "step": 3535
+    },
+    {
+      "epoch": 0.056576,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.510483870967742e-05,
+      "loss": 0.2126,
+      "step": 3536
+    },
+    {
+      "epoch": 0.056592,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.510322580645161e-05,
+      "loss": 0.1676,
+      "step": 3537
+    },
+    {
+      "epoch": 0.056608,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.510161290322581e-05,
+      "loss": 0.2042,
+      "step": 3538
+    },
+    {
+      "epoch": 0.056624,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.51e-05,
+      "loss": 0.2299,
+      "step": 3539
+    },
+    {
+      "epoch": 0.05664,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.50983870967742e-05,
+      "loss": 0.1561,
+      "step": 3540
+    },
+    {
+      "epoch": 0.056656,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.50967741935484e-05,
+      "loss": 0.2502,
+      "step": 3541
+    },
+    {
+      "epoch": 0.056672,
+      "grad_norm": 1.125,
+      "learning_rate": 9.509516129032259e-05,
+      "loss": 0.1719,
+      "step": 3542
+    },
+    {
+      "epoch": 0.056688,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.509354838709678e-05,
+      "loss": 0.171,
+      "step": 3543
+    },
+    {
+      "epoch": 0.056704,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.509193548387098e-05,
+      "loss": 0.1867,
+      "step": 3544
+    },
+    {
+      "epoch": 0.05672,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.509032258064516e-05,
+      "loss": 0.1752,
+      "step": 3545
+    },
+    {
+      "epoch": 0.056736,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.508870967741936e-05,
+      "loss": 0.1528,
+      "step": 3546
+    },
+    {
+      "epoch": 0.056752,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.508709677419355e-05,
+      "loss": 0.2046,
+      "step": 3547
+    },
+    {
+      "epoch": 0.056768,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.508548387096775e-05,
+      "loss": 0.2068,
+      "step": 3548
+    },
+    {
+      "epoch": 0.056784,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.508387096774194e-05,
+      "loss": 0.1903,
+      "step": 3549
+    },
+    {
+      "epoch": 0.0568,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.508225806451614e-05,
+      "loss": 0.1918,
+      "step": 3550
+    },
+    {
+      "epoch": 0.056816,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.508064516129032e-05,
+      "loss": 0.181,
+      "step": 3551
+    },
+    {
+      "epoch": 0.056832,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.507903225806452e-05,
+      "loss": 0.2145,
+      "step": 3552
+    },
+    {
+      "epoch": 0.056848,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.507741935483872e-05,
+      "loss": 0.1962,
+      "step": 3553
+    },
+    {
+      "epoch": 0.056864,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.507580645161291e-05,
+      "loss": 0.2129,
+      "step": 3554
+    },
+    {
+      "epoch": 0.05688,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.507419354838711e-05,
+      "loss": 0.1867,
+      "step": 3555
+    },
+    {
+      "epoch": 0.056896,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.507258064516129e-05,
+      "loss": 0.2273,
+      "step": 3556
+    },
+    {
+      "epoch": 0.056912,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.507096774193549e-05,
+      "loss": 0.1646,
+      "step": 3557
+    },
+    {
+      "epoch": 0.056928,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.506935483870968e-05,
+      "loss": 0.2093,
+      "step": 3558
+    },
+    {
+      "epoch": 0.056944,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.506774193548388e-05,
+      "loss": 0.184,
+      "step": 3559
+    },
+    {
+      "epoch": 0.05696,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.506612903225806e-05,
+      "loss": 0.2024,
+      "step": 3560
+    },
+    {
+      "epoch": 0.056976,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.506451612903226e-05,
+      "loss": 0.1647,
+      "step": 3561
+    },
+    {
+      "epoch": 0.056992,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.506290322580645e-05,
+      "loss": 0.1939,
+      "step": 3562
+    },
+    {
+      "epoch": 0.057008,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.506129032258065e-05,
+      "loss": 0.1643,
+      "step": 3563
+    },
+    {
+      "epoch": 0.057024,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.505967741935484e-05,
+      "loss": 0.226,
+      "step": 3564
+    },
+    {
+      "epoch": 0.05704,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.505806451612904e-05,
+      "loss": 0.1635,
+      "step": 3565
+    },
+    {
+      "epoch": 0.057056,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.505645161290324e-05,
+      "loss": 0.2098,
+      "step": 3566
+    },
+    {
+      "epoch": 0.057072,
+      "grad_norm": 0.875,
+      "learning_rate": 9.505483870967742e-05,
+      "loss": 0.1887,
+      "step": 3567
+    },
+    {
+      "epoch": 0.057088,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.505322580645162e-05,
+      "loss": 0.1692,
+      "step": 3568
+    },
+    {
+      "epoch": 0.057104,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.505161290322581e-05,
+      "loss": 0.1906,
+      "step": 3569
+    },
+    {
+      "epoch": 0.05712,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.505e-05,
+      "loss": 0.188,
+      "step": 3570
+    },
+    {
+      "epoch": 0.057136,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.504838709677419e-05,
+      "loss": 0.1563,
+      "step": 3571
+    },
+    {
+      "epoch": 0.057152,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.504677419354839e-05,
+      "loss": 0.177,
+      "step": 3572
+    },
+    {
+      "epoch": 0.057168,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.504516129032258e-05,
+      "loss": 0.1226,
+      "step": 3573
+    },
+    {
+      "epoch": 0.057184,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.504354838709678e-05,
+      "loss": 0.1931,
+      "step": 3574
+    },
+    {
+      "epoch": 0.0572,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.504193548387096e-05,
+      "loss": 0.1907,
+      "step": 3575
+    },
+    {
+      "epoch": 0.057216,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.504032258064516e-05,
+      "loss": 0.1549,
+      "step": 3576
+    },
+    {
+      "epoch": 0.057232,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.503870967741936e-05,
+      "loss": 0.1708,
+      "step": 3577
+    },
+    {
+      "epoch": 0.057248,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.503709677419356e-05,
+      "loss": 0.2089,
+      "step": 3578
+    },
+    {
+      "epoch": 0.057264,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.503548387096775e-05,
+      "loss": 0.1692,
+      "step": 3579
+    },
+    {
+      "epoch": 0.05728,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.503387096774195e-05,
+      "loss": 0.2087,
+      "step": 3580
+    },
+    {
+      "epoch": 0.057296,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.503225806451613e-05,
+      "loss": 0.2011,
+      "step": 3581
+    },
+    {
+      "epoch": 0.057312,
+      "grad_norm": 1.7578125,
+      "learning_rate": 9.503064516129033e-05,
+      "loss": 0.1608,
+      "step": 3582
+    },
+    {
+      "epoch": 0.057328,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.502903225806452e-05,
+      "loss": 0.1885,
+      "step": 3583
+    },
+    {
+      "epoch": 0.057344,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.50274193548387e-05,
+      "loss": 0.1511,
+      "step": 3584
+    },
+    {
+      "epoch": 0.05736,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.50258064516129e-05,
+      "loss": 0.197,
+      "step": 3585
+    },
+    {
+      "epoch": 0.057376,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.502419354838709e-05,
+      "loss": 0.1897,
+      "step": 3586
+    },
+    {
+      "epoch": 0.057392,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.502258064516129e-05,
+      "loss": 0.1853,
+      "step": 3587
+    },
+    {
+      "epoch": 0.057408,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.502096774193549e-05,
+      "loss": 0.1765,
+      "step": 3588
+    },
+    {
+      "epoch": 0.057424,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.501935483870969e-05,
+      "loss": 0.1975,
+      "step": 3589
+    },
+    {
+      "epoch": 0.05744,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.501774193548388e-05,
+      "loss": 0.1851,
+      "step": 3590
+    },
+    {
+      "epoch": 0.057456,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.501612903225808e-05,
+      "loss": 0.1936,
+      "step": 3591
+    },
+    {
+      "epoch": 0.057472,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.501451612903226e-05,
+      "loss": 0.2256,
+      "step": 3592
+    },
+    {
+      "epoch": 0.057488,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.501290322580646e-05,
+      "loss": 0.1593,
+      "step": 3593
+    },
+    {
+      "epoch": 0.057504,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.501129032258065e-05,
+      "loss": 0.1948,
+      "step": 3594
+    },
+    {
+      "epoch": 0.05752,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.500967741935485e-05,
+      "loss": 0.1936,
+      "step": 3595
+    },
+    {
+      "epoch": 0.057536,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.500806451612903e-05,
+      "loss": 0.174,
+      "step": 3596
+    },
+    {
+      "epoch": 0.057552,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.500645161290323e-05,
+      "loss": 0.1905,
+      "step": 3597
+    },
+    {
+      "epoch": 0.057568,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.500483870967742e-05,
+      "loss": 0.1767,
+      "step": 3598
+    },
+    {
+      "epoch": 0.057584,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.50032258064516e-05,
+      "loss": 0.1654,
+      "step": 3599
+    },
+    {
+      "epoch": 0.0576,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.50016129032258e-05,
+      "loss": 0.1684,
+      "step": 3600
+    },
+    {
+      "epoch": 0.057616,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.5e-05,
+      "loss": 0.1965,
+      "step": 3601
+    },
+    {
+      "epoch": 0.057632,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.49983870967742e-05,
+      "loss": 0.1899,
+      "step": 3602
+    },
+    {
+      "epoch": 0.057648,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.499677419354839e-05,
+      "loss": 0.1725,
+      "step": 3603
+    },
+    {
+      "epoch": 0.057664,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.499516129032259e-05,
+      "loss": 0.2046,
+      "step": 3604
+    },
+    {
+      "epoch": 0.05768,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.499354838709678e-05,
+      "loss": 0.1902,
+      "step": 3605
+    },
+    {
+      "epoch": 0.057696,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.499193548387098e-05,
+      "loss": 0.2385,
+      "step": 3606
+    },
+    {
+      "epoch": 0.057712,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.499032258064516e-05,
+      "loss": 0.1903,
+      "step": 3607
+    },
+    {
+      "epoch": 0.057728,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.498870967741936e-05,
+      "loss": 0.1681,
+      "step": 3608
+    },
+    {
+      "epoch": 0.057744,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.498709677419355e-05,
+      "loss": 0.2332,
+      "step": 3609
+    },
+    {
+      "epoch": 0.05776,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.498548387096775e-05,
+      "loss": 0.1879,
+      "step": 3610
+    },
+    {
+      "epoch": 0.057776,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.498387096774193e-05,
+      "loss": 0.1789,
+      "step": 3611
+    },
+    {
+      "epoch": 0.057792,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.498225806451613e-05,
+      "loss": 0.1764,
+      "step": 3612
+    },
+    {
+      "epoch": 0.057808,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.498064516129033e-05,
+      "loss": 0.2083,
+      "step": 3613
+    },
+    {
+      "epoch": 0.057824,
+      "grad_norm": 0.875,
+      "learning_rate": 9.497903225806452e-05,
+      "loss": 0.2164,
+      "step": 3614
+    },
+    {
+      "epoch": 0.05784,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.497741935483872e-05,
+      "loss": 0.2491,
+      "step": 3615
+    },
+    {
+      "epoch": 0.057856,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.49758064516129e-05,
+      "loss": 0.1916,
+      "step": 3616
+    },
+    {
+      "epoch": 0.057872,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.49741935483871e-05,
+      "loss": 0.1492,
+      "step": 3617
+    },
+    {
+      "epoch": 0.057888,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.497258064516129e-05,
+      "loss": 0.209,
+      "step": 3618
+    },
+    {
+      "epoch": 0.057904,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.497096774193549e-05,
+      "loss": 0.1569,
+      "step": 3619
+    },
+    {
+      "epoch": 0.05792,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.496935483870968e-05,
+      "loss": 0.1817,
+      "step": 3620
+    },
+    {
+      "epoch": 0.057936,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.496774193548388e-05,
+      "loss": 0.2314,
+      "step": 3621
+    },
+    {
+      "epoch": 0.057952,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.496612903225806e-05,
+      "loss": 0.1744,
+      "step": 3622
+    },
+    {
+      "epoch": 0.057968,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.496451612903226e-05,
+      "loss": 0.1763,
+      "step": 3623
+    },
+    {
+      "epoch": 0.057984,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.496290322580645e-05,
+      "loss": 0.1855,
+      "step": 3624
+    },
+    {
+      "epoch": 0.058,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.496129032258065e-05,
+      "loss": 0.2302,
+      "step": 3625
+    },
+    {
+      "epoch": 0.058016,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.495967741935485e-05,
+      "loss": 0.1953,
+      "step": 3626
+    },
+    {
+      "epoch": 0.058032,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.495806451612905e-05,
+      "loss": 0.197,
+      "step": 3627
+    },
+    {
+      "epoch": 0.058048,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.495645161290323e-05,
+      "loss": 0.1793,
+      "step": 3628
+    },
+    {
+      "epoch": 0.058064,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.495483870967743e-05,
+      "loss": 0.1708,
+      "step": 3629
+    },
+    {
+      "epoch": 0.05808,
+      "grad_norm": 1.296875,
+      "learning_rate": 9.495322580645162e-05,
+      "loss": 0.2104,
+      "step": 3630
+    },
+    {
+      "epoch": 0.058096,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.49516129032258e-05,
+      "loss": 0.225,
+      "step": 3631
+    },
+    {
+      "epoch": 0.058112,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.495e-05,
+      "loss": 0.139,
+      "step": 3632
+    },
+    {
+      "epoch": 0.058128,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.494838709677419e-05,
+      "loss": 0.1865,
+      "step": 3633
+    },
+    {
+      "epoch": 0.058144,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.494677419354839e-05,
+      "loss": 0.1689,
+      "step": 3634
+    },
+    {
+      "epoch": 0.05816,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.494516129032258e-05,
+      "loss": 0.1662,
+      "step": 3635
+    },
+    {
+      "epoch": 0.058176,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.494354838709678e-05,
+      "loss": 0.1735,
+      "step": 3636
+    },
+    {
+      "epoch": 0.058192,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.494193548387098e-05,
+      "loss": 0.1698,
+      "step": 3637
+    },
+    {
+      "epoch": 0.058208,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.494032258064517e-05,
+      "loss": 0.159,
+      "step": 3638
+    },
+    {
+      "epoch": 0.058224,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.493870967741936e-05,
+      "loss": 0.1767,
+      "step": 3639
+    },
+    {
+      "epoch": 0.05824,
+      "grad_norm": 1.0,
+      "learning_rate": 9.493709677419356e-05,
+      "loss": 0.1685,
+      "step": 3640
+    },
+    {
+      "epoch": 0.058256,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.493548387096775e-05,
+      "loss": 0.2093,
+      "step": 3641
+    },
+    {
+      "epoch": 0.058272,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.493387096774195e-05,
+      "loss": 0.1622,
+      "step": 3642
+    },
+    {
+      "epoch": 0.058288,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.493225806451613e-05,
+      "loss": 0.1855,
+      "step": 3643
+    },
+    {
+      "epoch": 0.058304,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.493064516129033e-05,
+      "loss": 0.1657,
+      "step": 3644
+    },
+    {
+      "epoch": 0.05832,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.492903225806452e-05,
+      "loss": 0.2032,
+      "step": 3645
+    },
+    {
+      "epoch": 0.058336,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.49274193548387e-05,
+      "loss": 0.1883,
+      "step": 3646
+    },
+    {
+      "epoch": 0.058352,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.49258064516129e-05,
+      "loss": 0.192,
+      "step": 3647
+    },
+    {
+      "epoch": 0.058368,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.49241935483871e-05,
+      "loss": 0.2044,
+      "step": 3648
+    },
+    {
+      "epoch": 0.058384,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.49225806451613e-05,
+      "loss": 0.1843,
+      "step": 3649
+    },
+    {
+      "epoch": 0.0584,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.492096774193549e-05,
+      "loss": 0.1879,
+      "step": 3650
+    },
+    {
+      "epoch": 0.058416,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.491935483870969e-05,
+      "loss": 0.1704,
+      "step": 3651
+    },
+    {
+      "epoch": 0.058432,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.491774193548387e-05,
+      "loss": 0.1905,
+      "step": 3652
+    },
+    {
+      "epoch": 0.058448,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.491612903225807e-05,
+      "loss": 0.2119,
+      "step": 3653
+    },
+    {
+      "epoch": 0.058464,
+      "grad_norm": 1.859375,
+      "learning_rate": 9.491451612903226e-05,
+      "loss": 0.2264,
+      "step": 3654
+    },
+    {
+      "epoch": 0.05848,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.491290322580646e-05,
+      "loss": 0.1809,
+      "step": 3655
+    },
+    {
+      "epoch": 0.058496,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.491129032258065e-05,
+      "loss": 0.1661,
+      "step": 3656
+    },
+    {
+      "epoch": 0.058512,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.490967741935485e-05,
+      "loss": 0.2051,
+      "step": 3657
+    },
+    {
+      "epoch": 0.058528,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.490806451612903e-05,
+      "loss": 0.1735,
+      "step": 3658
+    },
+    {
+      "epoch": 0.058544,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.490645161290323e-05,
+      "loss": 0.1499,
+      "step": 3659
+    },
+    {
+      "epoch": 0.05856,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.490483870967742e-05,
+      "loss": 0.2427,
+      "step": 3660
+    },
+    {
+      "epoch": 0.058576,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.490322580645162e-05,
+      "loss": 0.1516,
+      "step": 3661
+    },
+    {
+      "epoch": 0.058592,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.490161290322582e-05,
+      "loss": 0.1541,
+      "step": 3662
+    },
+    {
+      "epoch": 0.058608,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.49e-05,
+      "loss": 0.2113,
+      "step": 3663
+    },
+    {
+      "epoch": 0.058624,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.48983870967742e-05,
+      "loss": 0.169,
+      "step": 3664
+    },
+    {
+      "epoch": 0.05864,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.489677419354839e-05,
+      "loss": 0.2052,
+      "step": 3665
+    },
+    {
+      "epoch": 0.058656,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.489516129032259e-05,
+      "loss": 0.1849,
+      "step": 3666
+    },
+    {
+      "epoch": 0.058672,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.489354838709677e-05,
+      "loss": 0.1532,
+      "step": 3667
+    },
+    {
+      "epoch": 0.058688,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.489193548387097e-05,
+      "loss": 0.2126,
+      "step": 3668
+    },
+    {
+      "epoch": 0.058704,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.489032258064516e-05,
+      "loss": 0.2046,
+      "step": 3669
+    },
+    {
+      "epoch": 0.05872,
+      "grad_norm": 1.25,
+      "learning_rate": 9.488870967741936e-05,
+      "loss": 0.1892,
+      "step": 3670
+    },
+    {
+      "epoch": 0.058736,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.488709677419355e-05,
+      "loss": 0.1395,
+      "step": 3671
+    },
+    {
+      "epoch": 0.058752,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.488548387096775e-05,
+      "loss": 0.2222,
+      "step": 3672
+    },
+    {
+      "epoch": 0.058768,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.488387096774194e-05,
+      "loss": 0.1752,
+      "step": 3673
+    },
+    {
+      "epoch": 0.058784,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.488225806451614e-05,
+      "loss": 0.2136,
+      "step": 3674
+    },
+    {
+      "epoch": 0.0588,
+      "grad_norm": 1.375,
+      "learning_rate": 9.488064516129033e-05,
+      "loss": 0.1909,
+      "step": 3675
+    },
+    {
+      "epoch": 0.058816,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.487903225806452e-05,
+      "loss": 0.1822,
+      "step": 3676
+    },
+    {
+      "epoch": 0.058832,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.487741935483872e-05,
+      "loss": 0.1565,
+      "step": 3677
+    },
+    {
+      "epoch": 0.058848,
+      "grad_norm": 1.125,
+      "learning_rate": 9.48758064516129e-05,
+      "loss": 0.2324,
+      "step": 3678
+    },
+    {
+      "epoch": 0.058864,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.48741935483871e-05,
+      "loss": 0.1949,
+      "step": 3679
+    },
+    {
+      "epoch": 0.05888,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.487258064516129e-05,
+      "loss": 0.1771,
+      "step": 3680
+    },
+    {
+      "epoch": 0.058896,
+      "grad_norm": 0.875,
+      "learning_rate": 9.487096774193549e-05,
+      "loss": 0.2247,
+      "step": 3681
+    },
+    {
+      "epoch": 0.058912,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.486935483870967e-05,
+      "loss": 0.1718,
+      "step": 3682
+    },
+    {
+      "epoch": 0.058928,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.486774193548387e-05,
+      "loss": 0.1636,
+      "step": 3683
+    },
+    {
+      "epoch": 0.058944,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.486612903225807e-05,
+      "loss": 0.204,
+      "step": 3684
+    },
+    {
+      "epoch": 0.05896,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.486451612903227e-05,
+      "loss": 0.2044,
+      "step": 3685
+    },
+    {
+      "epoch": 0.058976,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.486290322580646e-05,
+      "loss": 0.2047,
+      "step": 3686
+    },
+    {
+      "epoch": 0.058992,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.486129032258066e-05,
+      "loss": 0.1767,
+      "step": 3687
+    },
+    {
+      "epoch": 0.059008,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.485967741935484e-05,
+      "loss": 0.1809,
+      "step": 3688
+    },
+    {
+      "epoch": 0.059024,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.485806451612904e-05,
+      "loss": 0.2284,
+      "step": 3689
+    },
+    {
+      "epoch": 0.05904,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.485645161290323e-05,
+      "loss": 0.186,
+      "step": 3690
+    },
+    {
+      "epoch": 0.059056,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.485483870967743e-05,
+      "loss": 0.1482,
+      "step": 3691
+    },
+    {
+      "epoch": 0.059072,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.485322580645162e-05,
+      "loss": 0.1455,
+      "step": 3692
+    },
+    {
+      "epoch": 0.059088,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.48516129032258e-05,
+      "loss": 0.1941,
+      "step": 3693
+    },
+    {
+      "epoch": 0.059104,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.485e-05,
+      "loss": 0.1916,
+      "step": 3694
+    },
+    {
+      "epoch": 0.05912,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.484838709677419e-05,
+      "loss": 0.17,
+      "step": 3695
+    },
+    {
+      "epoch": 0.059136,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.484677419354839e-05,
+      "loss": 0.1951,
+      "step": 3696
+    },
+    {
+      "epoch": 0.059152,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.484516129032259e-05,
+      "loss": 0.1517,
+      "step": 3697
+    },
+    {
+      "epoch": 0.059168,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.484354838709679e-05,
+      "loss": 0.1398,
+      "step": 3698
+    },
+    {
+      "epoch": 0.059184,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.484193548387097e-05,
+      "loss": 0.1828,
+      "step": 3699
+    },
+    {
+      "epoch": 0.0592,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.484032258064517e-05,
+      "loss": 0.2094,
+      "step": 3700
+    },
+    {
+      "epoch": 0.059216,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.483870967741936e-05,
+      "loss": 0.1911,
+      "step": 3701
+    },
+    {
+      "epoch": 0.059232,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.483709677419356e-05,
+      "loss": 0.2129,
+      "step": 3702
+    },
+    {
+      "epoch": 0.059248,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.483548387096774e-05,
+      "loss": 0.1816,
+      "step": 3703
+    },
+    {
+      "epoch": 0.059264,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.483387096774194e-05,
+      "loss": 0.1633,
+      "step": 3704
+    },
+    {
+      "epoch": 0.05928,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.483225806451613e-05,
+      "loss": 0.1897,
+      "step": 3705
+    },
+    {
+      "epoch": 0.059296,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.483064516129033e-05,
+      "loss": 0.1912,
+      "step": 3706
+    },
+    {
+      "epoch": 0.059312,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.482903225806452e-05,
+      "loss": 0.1665,
+      "step": 3707
+    },
+    {
+      "epoch": 0.059328,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.482741935483872e-05,
+      "loss": 0.1745,
+      "step": 3708
+    },
+    {
+      "epoch": 0.059344,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.482580645161291e-05,
+      "loss": 0.2302,
+      "step": 3709
+    },
+    {
+      "epoch": 0.05936,
+      "grad_norm": 1.46875,
+      "learning_rate": 9.48241935483871e-05,
+      "loss": 0.1659,
+      "step": 3710
+    },
+    {
+      "epoch": 0.059376,
+      "grad_norm": 2.0,
+      "learning_rate": 9.48225806451613e-05,
+      "loss": 0.2223,
+      "step": 3711
+    },
+    {
+      "epoch": 0.059392,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.482096774193549e-05,
+      "loss": 0.1995,
+      "step": 3712
+    },
+    {
+      "epoch": 0.059408,
+      "grad_norm": 1.4140625,
+      "learning_rate": 9.481935483870969e-05,
+      "loss": 0.2034,
+      "step": 3713
+    },
+    {
+      "epoch": 0.059424,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.481774193548387e-05,
+      "loss": 0.1724,
+      "step": 3714
+    },
+    {
+      "epoch": 0.05944,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.481612903225807e-05,
+      "loss": 0.1816,
+      "step": 3715
+    },
+    {
+      "epoch": 0.059456,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.481451612903226e-05,
+      "loss": 0.1916,
+      "step": 3716
+    },
+    {
+      "epoch": 0.059472,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.481290322580646e-05,
+      "loss": 0.1641,
+      "step": 3717
+    },
+    {
+      "epoch": 0.059488,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.481129032258064e-05,
+      "loss": 0.1854,
+      "step": 3718
+    },
+    {
+      "epoch": 0.059504,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.480967741935484e-05,
+      "loss": 0.1922,
+      "step": 3719
+    },
+    {
+      "epoch": 0.05952,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.480806451612903e-05,
+      "loss": 0.177,
+      "step": 3720
+    },
+    {
+      "epoch": 0.059536,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.480645161290323e-05,
+      "loss": 0.1675,
+      "step": 3721
+    },
+    {
+      "epoch": 0.059552,
+      "grad_norm": 1.8046875,
+      "learning_rate": 9.480483870967743e-05,
+      "loss": 0.2013,
+      "step": 3722
+    },
+    {
+      "epoch": 0.059568,
+      "grad_norm": 1.5390625,
+      "learning_rate": 9.480322580645161e-05,
+      "loss": 0.2062,
+      "step": 3723
+    },
+    {
+      "epoch": 0.059584,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.480161290322581e-05,
+      "loss": 0.1974,
+      "step": 3724
+    },
+    {
+      "epoch": 0.0596,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.48e-05,
+      "loss": 0.1554,
+      "step": 3725
+    },
+    {
+      "epoch": 0.059616,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.47983870967742e-05,
+      "loss": 0.1841,
+      "step": 3726
+    },
+    {
+      "epoch": 0.059632,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.479677419354839e-05,
+      "loss": 0.1816,
+      "step": 3727
+    },
+    {
+      "epoch": 0.059648,
+      "grad_norm": 1.859375,
+      "learning_rate": 9.479516129032259e-05,
+      "loss": 0.1714,
+      "step": 3728
+    },
+    {
+      "epoch": 0.059664,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.479354838709677e-05,
+      "loss": 0.1917,
+      "step": 3729
+    },
+    {
+      "epoch": 0.05968,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.479193548387097e-05,
+      "loss": 0.1713,
+      "step": 3730
+    },
+    {
+      "epoch": 0.059696,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.479032258064516e-05,
+      "loss": 0.1859,
+      "step": 3731
+    },
+    {
+      "epoch": 0.059712,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.478870967741936e-05,
+      "loss": 0.1499,
+      "step": 3732
+    },
+    {
+      "epoch": 0.059728,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.478709677419356e-05,
+      "loss": 0.1969,
+      "step": 3733
+    },
+    {
+      "epoch": 0.059744,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.478548387096776e-05,
+      "loss": 0.155,
+      "step": 3734
+    },
+    {
+      "epoch": 0.05976,
+      "grad_norm": 0.75,
+      "learning_rate": 9.478387096774194e-05,
+      "loss": 0.1958,
+      "step": 3735
+    },
+    {
+      "epoch": 0.059776,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.478225806451614e-05,
+      "loss": 0.1742,
+      "step": 3736
+    },
+    {
+      "epoch": 0.059792,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.478064516129033e-05,
+      "loss": 0.1682,
+      "step": 3737
+    },
+    {
+      "epoch": 0.059808,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.477903225806451e-05,
+      "loss": 0.1599,
+      "step": 3738
+    },
+    {
+      "epoch": 0.059824,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.477741935483871e-05,
+      "loss": 0.1381,
+      "step": 3739
+    },
+    {
+      "epoch": 0.05984,
+      "grad_norm": 1.46875,
+      "learning_rate": 9.47758064516129e-05,
+      "loss": 0.2187,
+      "step": 3740
+    },
+    {
+      "epoch": 0.059856,
+      "grad_norm": 0.75,
+      "learning_rate": 9.47741935483871e-05,
+      "loss": 0.1627,
+      "step": 3741
+    },
+    {
+      "epoch": 0.059872,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.477258064516129e-05,
+      "loss": 0.1604,
+      "step": 3742
+    },
+    {
+      "epoch": 0.059888,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.477096774193549e-05,
+      "loss": 0.14,
+      "step": 3743
+    },
+    {
+      "epoch": 0.059904,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.476935483870968e-05,
+      "loss": 0.2245,
+      "step": 3744
+    },
+    {
+      "epoch": 0.05992,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.476774193548388e-05,
+      "loss": 0.1689,
+      "step": 3745
+    },
+    {
+      "epoch": 0.059936,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.476612903225807e-05,
+      "loss": 0.1814,
+      "step": 3746
+    },
+    {
+      "epoch": 0.059952,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.476451612903227e-05,
+      "loss": 0.2279,
+      "step": 3747
+    },
+    {
+      "epoch": 0.059968,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.476290322580646e-05,
+      "loss": 0.2161,
+      "step": 3748
+    },
+    {
+      "epoch": 0.059984,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.476129032258066e-05,
+      "loss": 0.2082,
+      "step": 3749
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.475967741935484e-05,
+      "loss": 0.1345,
+      "step": 3750
+    },
+    {
+      "epoch": 0.060016,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.475806451612904e-05,
+      "loss": 0.1783,
+      "step": 3751
+    },
+    {
+      "epoch": 0.060032,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.475645161290323e-05,
+      "loss": 0.1756,
+      "step": 3752
+    },
+    {
+      "epoch": 0.060048,
+      "grad_norm": 1.4140625,
+      "learning_rate": 9.475483870967743e-05,
+      "loss": 0.2194,
+      "step": 3753
+    },
+    {
+      "epoch": 0.060064,
+      "grad_norm": 1.125,
+      "learning_rate": 9.475322580645161e-05,
+      "loss": 0.175,
+      "step": 3754
+    },
+    {
+      "epoch": 0.06008,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.47516129032258e-05,
+      "loss": 0.1946,
+      "step": 3755
+    },
+    {
+      "epoch": 0.060096,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.475e-05,
+      "loss": 0.211,
+      "step": 3756
+    },
+    {
+      "epoch": 0.060112,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.47483870967742e-05,
+      "loss": 0.191,
+      "step": 3757
+    },
+    {
+      "epoch": 0.060128,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.47467741935484e-05,
+      "loss": 0.1976,
+      "step": 3758
+    },
+    {
+      "epoch": 0.060144,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.474516129032258e-05,
+      "loss": 0.1804,
+      "step": 3759
+    },
+    {
+      "epoch": 0.06016,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.474354838709678e-05,
+      "loss": 0.1727,
+      "step": 3760
+    },
+    {
+      "epoch": 0.060176,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.474193548387097e-05,
+      "loss": 0.1398,
+      "step": 3761
+    },
+    {
+      "epoch": 0.060192,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.474032258064517e-05,
+      "loss": 0.1728,
+      "step": 3762
+    },
+    {
+      "epoch": 0.060208,
+      "grad_norm": 1.875,
+      "learning_rate": 9.473870967741936e-05,
+      "loss": 0.1995,
+      "step": 3763
+    },
+    {
+      "epoch": 0.060224,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.473709677419356e-05,
+      "loss": 0.1956,
+      "step": 3764
+    },
+    {
+      "epoch": 0.06024,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.473548387096774e-05,
+      "loss": 0.2543,
+      "step": 3765
+    },
+    {
+      "epoch": 0.060256,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.473387096774194e-05,
+      "loss": 0.1986,
+      "step": 3766
+    },
+    {
+      "epoch": 0.060272,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.473225806451613e-05,
+      "loss": 0.2208,
+      "step": 3767
+    },
+    {
+      "epoch": 0.060288,
+      "grad_norm": 1.46875,
+      "learning_rate": 9.473064516129033e-05,
+      "loss": 0.1977,
+      "step": 3768
+    },
+    {
+      "epoch": 0.060304,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.472903225806453e-05,
+      "loss": 0.1776,
+      "step": 3769
+    },
+    {
+      "epoch": 0.06032,
+      "grad_norm": 1.8203125,
+      "learning_rate": 9.472741935483871e-05,
+      "loss": 0.2054,
+      "step": 3770
+    },
+    {
+      "epoch": 0.060336,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.472580645161291e-05,
+      "loss": 0.1907,
+      "step": 3771
+    },
+    {
+      "epoch": 0.060352,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.47241935483871e-05,
+      "loss": 0.1509,
+      "step": 3772
+    },
+    {
+      "epoch": 0.060368,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.47225806451613e-05,
+      "loss": 0.1557,
+      "step": 3773
+    },
+    {
+      "epoch": 0.060384,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.472096774193548e-05,
+      "loss": 0.2038,
+      "step": 3774
+    },
+    {
+      "epoch": 0.0604,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.471935483870968e-05,
+      "loss": 0.1638,
+      "step": 3775
+    },
+    {
+      "epoch": 0.060416,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.471774193548387e-05,
+      "loss": 0.213,
+      "step": 3776
+    },
+    {
+      "epoch": 0.060432,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.471612903225807e-05,
+      "loss": 0.1739,
+      "step": 3777
+    },
+    {
+      "epoch": 0.060448,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.471451612903226e-05,
+      "loss": 0.1602,
+      "step": 3778
+    },
+    {
+      "epoch": 0.060464,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.471290322580646e-05,
+      "loss": 0.1918,
+      "step": 3779
+    },
+    {
+      "epoch": 0.06048,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.471129032258065e-05,
+      "loss": 0.1528,
+      "step": 3780
+    },
+    {
+      "epoch": 0.060496,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.470967741935484e-05,
+      "loss": 0.1834,
+      "step": 3781
+    },
+    {
+      "epoch": 0.060512,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.470806451612904e-05,
+      "loss": 0.2108,
+      "step": 3782
+    },
+    {
+      "epoch": 0.060528,
+      "grad_norm": 1.4609375,
+      "learning_rate": 9.470645161290324e-05,
+      "loss": 0.1869,
+      "step": 3783
+    },
+    {
+      "epoch": 0.060544,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.470483870967743e-05,
+      "loss": 0.1419,
+      "step": 3784
+    },
+    {
+      "epoch": 0.06056,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.470322580645161e-05,
+      "loss": 0.1534,
+      "step": 3785
+    },
+    {
+      "epoch": 0.060576,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.470161290322581e-05,
+      "loss": 0.1566,
+      "step": 3786
+    },
+    {
+      "epoch": 0.060592,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.47e-05,
+      "loss": 0.1606,
+      "step": 3787
+    },
+    {
+      "epoch": 0.060608,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.46983870967742e-05,
+      "loss": 0.211,
+      "step": 3788
+    },
+    {
+      "epoch": 0.060624,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.469677419354838e-05,
+      "loss": 0.2087,
+      "step": 3789
+    },
+    {
+      "epoch": 0.06064,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.469516129032258e-05,
+      "loss": 0.15,
+      "step": 3790
+    },
+    {
+      "epoch": 0.060656,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.469354838709677e-05,
+      "loss": 0.2135,
+      "step": 3791
+    },
+    {
+      "epoch": 0.060672,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.469193548387097e-05,
+      "loss": 0.1513,
+      "step": 3792
+    },
+    {
+      "epoch": 0.060688,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.469032258064517e-05,
+      "loss": 0.2174,
+      "step": 3793
+    },
+    {
+      "epoch": 0.060704,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.468870967741937e-05,
+      "loss": 0.1804,
+      "step": 3794
+    },
+    {
+      "epoch": 0.06072,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.468709677419355e-05,
+      "loss": 0.1805,
+      "step": 3795
+    },
+    {
+      "epoch": 0.060736,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.468548387096775e-05,
+      "loss": 0.1664,
+      "step": 3796
+    },
+    {
+      "epoch": 0.060752,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.468387096774194e-05,
+      "loss": 0.1929,
+      "step": 3797
+    },
+    {
+      "epoch": 0.060768,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.468225806451614e-05,
+      "loss": 0.1549,
+      "step": 3798
+    },
+    {
+      "epoch": 0.060784,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.468064516129033e-05,
+      "loss": 0.1976,
+      "step": 3799
+    },
+    {
+      "epoch": 0.0608,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.467903225806453e-05,
+      "loss": 0.1972,
+      "step": 3800
+    },
+    {
+      "epoch": 0.060816,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.467741935483871e-05,
+      "loss": 0.1828,
+      "step": 3801
+    },
+    {
+      "epoch": 0.060832,
+      "grad_norm": 0.625,
+      "learning_rate": 9.46758064516129e-05,
+      "loss": 0.1717,
+      "step": 3802
+    },
+    {
+      "epoch": 0.060848,
+      "grad_norm": 1.125,
+      "learning_rate": 9.46741935483871e-05,
+      "loss": 0.1977,
+      "step": 3803
+    },
+    {
+      "epoch": 0.060864,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.46725806451613e-05,
+      "loss": 0.1643,
+      "step": 3804
+    },
+    {
+      "epoch": 0.06088,
+      "grad_norm": 1.328125,
+      "learning_rate": 9.46709677419355e-05,
+      "loss": 0.1552,
+      "step": 3805
+    },
+    {
+      "epoch": 0.060896,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.466935483870968e-05,
+      "loss": 0.1979,
+      "step": 3806
+    },
+    {
+      "epoch": 0.060912,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.466774193548388e-05,
+      "loss": 0.189,
+      "step": 3807
+    },
+    {
+      "epoch": 0.060928,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.466612903225807e-05,
+      "loss": 0.1729,
+      "step": 3808
+    },
+    {
+      "epoch": 0.060944,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.466451612903227e-05,
+      "loss": 0.1794,
+      "step": 3809
+    },
+    {
+      "epoch": 0.06096,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.466290322580645e-05,
+      "loss": 0.2047,
+      "step": 3810
+    },
+    {
+      "epoch": 0.060976,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.466129032258065e-05,
+      "loss": 0.1672,
+      "step": 3811
+    },
+    {
+      "epoch": 0.060992,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.465967741935484e-05,
+      "loss": 0.2125,
+      "step": 3812
+    },
+    {
+      "epoch": 0.061008,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.465806451612904e-05,
+      "loss": 0.1742,
+      "step": 3813
+    },
+    {
+      "epoch": 0.061024,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.465645161290323e-05,
+      "loss": 0.2197,
+      "step": 3814
+    },
+    {
+      "epoch": 0.06104,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.465483870967742e-05,
+      "loss": 0.1546,
+      "step": 3815
+    },
+    {
+      "epoch": 0.061056,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.465322580645161e-05,
+      "loss": 0.2141,
+      "step": 3816
+    },
+    {
+      "epoch": 0.061072,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.465161290322581e-05,
+      "loss": 0.1586,
+      "step": 3817
+    },
+    {
+      "epoch": 0.061088,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.465000000000001e-05,
+      "loss": 0.1949,
+      "step": 3818
+    },
+    {
+      "epoch": 0.061104,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.46483870967742e-05,
+      "loss": 0.1948,
+      "step": 3819
+    },
+    {
+      "epoch": 0.06112,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.46467741935484e-05,
+      "loss": 0.2057,
+      "step": 3820
+    },
+    {
+      "epoch": 0.061136,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.464516129032258e-05,
+      "loss": 0.2197,
+      "step": 3821
+    },
+    {
+      "epoch": 0.061152,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.464354838709678e-05,
+      "loss": 0.1802,
+      "step": 3822
+    },
+    {
+      "epoch": 0.061168,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.464193548387097e-05,
+      "loss": 0.2015,
+      "step": 3823
+    },
+    {
+      "epoch": 0.061184,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.464032258064517e-05,
+      "loss": 0.1654,
+      "step": 3824
+    },
+    {
+      "epoch": 0.0612,
+      "grad_norm": 0.490234375,
+      "learning_rate": 9.463870967741935e-05,
+      "loss": 0.1637,
+      "step": 3825
+    },
+    {
+      "epoch": 0.061216,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.463709677419355e-05,
+      "loss": 0.187,
+      "step": 3826
+    },
+    {
+      "epoch": 0.061232,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.463548387096774e-05,
+      "loss": 0.1801,
+      "step": 3827
+    },
+    {
+      "epoch": 0.061248,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.463387096774194e-05,
+      "loss": 0.1651,
+      "step": 3828
+    },
+    {
+      "epoch": 0.061264,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.463225806451614e-05,
+      "loss": 0.1698,
+      "step": 3829
+    },
+    {
+      "epoch": 0.06128,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.463064516129034e-05,
+      "loss": 0.1985,
+      "step": 3830
+    },
+    {
+      "epoch": 0.061296,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.462903225806452e-05,
+      "loss": 0.1459,
+      "step": 3831
+    },
+    {
+      "epoch": 0.061312,
+      "grad_norm": 0.75,
+      "learning_rate": 9.462741935483871e-05,
+      "loss": 0.1785,
+      "step": 3832
+    },
+    {
+      "epoch": 0.061328,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.462580645161291e-05,
+      "loss": 0.185,
+      "step": 3833
+    },
+    {
+      "epoch": 0.061344,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.46241935483871e-05,
+      "loss": 0.1851,
+      "step": 3834
+    },
+    {
+      "epoch": 0.06136,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.46225806451613e-05,
+      "loss": 0.2044,
+      "step": 3835
+    },
+    {
+      "epoch": 0.061376,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.462096774193548e-05,
+      "loss": 0.202,
+      "step": 3836
+    },
+    {
+      "epoch": 0.061392,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.461935483870968e-05,
+      "loss": 0.1943,
+      "step": 3837
+    },
+    {
+      "epoch": 0.061408,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.461774193548387e-05,
+      "loss": 0.184,
+      "step": 3838
+    },
+    {
+      "epoch": 0.061424,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.461612903225807e-05,
+      "loss": 0.1909,
+      "step": 3839
+    },
+    {
+      "epoch": 0.06144,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.461451612903227e-05,
+      "loss": 0.1366,
+      "step": 3840
+    },
+    {
+      "epoch": 0.061456,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.461290322580647e-05,
+      "loss": 0.1902,
+      "step": 3841
+    },
+    {
+      "epoch": 0.061472,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.461129032258065e-05,
+      "loss": 0.1989,
+      "step": 3842
+    },
+    {
+      "epoch": 0.061488,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.460967741935485e-05,
+      "loss": 0.1845,
+      "step": 3843
+    },
+    {
+      "epoch": 0.061504,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.460806451612904e-05,
+      "loss": 0.1727,
+      "step": 3844
+    },
+    {
+      "epoch": 0.06152,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.460645161290324e-05,
+      "loss": 0.1995,
+      "step": 3845
+    },
+    {
+      "epoch": 0.061536,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.460483870967742e-05,
+      "loss": 0.1641,
+      "step": 3846
+    },
+    {
+      "epoch": 0.061552,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.460322580645161e-05,
+      "loss": 0.2055,
+      "step": 3847
+    },
+    {
+      "epoch": 0.061568,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.460161290322581e-05,
+      "loss": 0.1754,
+      "step": 3848
+    },
+    {
+      "epoch": 0.061584,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.46e-05,
+      "loss": 0.2149,
+      "step": 3849
+    },
+    {
+      "epoch": 0.0616,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.45983870967742e-05,
+      "loss": 0.1532,
+      "step": 3850
+    },
+    {
+      "epoch": 0.061616,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.459677419354838e-05,
+      "loss": 0.2005,
+      "step": 3851
+    },
+    {
+      "epoch": 0.061632,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.459516129032258e-05,
+      "loss": 0.2126,
+      "step": 3852
+    },
+    {
+      "epoch": 0.061648,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.459354838709678e-05,
+      "loss": 0.1978,
+      "step": 3853
+    },
+    {
+      "epoch": 0.061664,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.459193548387098e-05,
+      "loss": 0.1604,
+      "step": 3854
+    },
+    {
+      "epoch": 0.06168,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.459032258064517e-05,
+      "loss": 0.1839,
+      "step": 3855
+    },
+    {
+      "epoch": 0.061696,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.458870967741937e-05,
+      "loss": 0.1709,
+      "step": 3856
+    },
+    {
+      "epoch": 0.061712,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.458709677419355e-05,
+      "loss": 0.1413,
+      "step": 3857
+    },
+    {
+      "epoch": 0.061728,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.458548387096775e-05,
+      "loss": 0.1996,
+      "step": 3858
+    },
+    {
+      "epoch": 0.061744,
+      "grad_norm": 1.296875,
+      "learning_rate": 9.458387096774194e-05,
+      "loss": 0.1956,
+      "step": 3859
+    },
+    {
+      "epoch": 0.06176,
+      "grad_norm": 1.8125,
+      "learning_rate": 9.458225806451614e-05,
+      "loss": 0.1754,
+      "step": 3860
+    },
+    {
+      "epoch": 0.061776,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.458064516129032e-05,
+      "loss": 0.1856,
+      "step": 3861
+    },
+    {
+      "epoch": 0.061792,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.457903225806452e-05,
+      "loss": 0.1954,
+      "step": 3862
+    },
+    {
+      "epoch": 0.061808,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.457741935483871e-05,
+      "loss": 0.2055,
+      "step": 3863
+    },
+    {
+      "epoch": 0.061824,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.457580645161291e-05,
+      "loss": 0.2162,
+      "step": 3864
+    },
+    {
+      "epoch": 0.06184,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.457419354838711e-05,
+      "loss": 0.2104,
+      "step": 3865
+    },
+    {
+      "epoch": 0.061856,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.45725806451613e-05,
+      "loss": 0.1464,
+      "step": 3866
+    },
+    {
+      "epoch": 0.061872,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.45709677419355e-05,
+      "loss": 0.1367,
+      "step": 3867
+    },
+    {
+      "epoch": 0.061888,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.456935483870968e-05,
+      "loss": 0.1962,
+      "step": 3868
+    },
+    {
+      "epoch": 0.061904,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.456774193548388e-05,
+      "loss": 0.1582,
+      "step": 3869
+    },
+    {
+      "epoch": 0.06192,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.456612903225807e-05,
+      "loss": 0.1749,
+      "step": 3870
+    },
+    {
+      "epoch": 0.061936,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.456451612903227e-05,
+      "loss": 0.1731,
+      "step": 3871
+    },
+    {
+      "epoch": 0.061952,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.456290322580645e-05,
+      "loss": 0.1595,
+      "step": 3872
+    },
+    {
+      "epoch": 0.061968,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.456129032258065e-05,
+      "loss": 0.1415,
+      "step": 3873
+    },
+    {
+      "epoch": 0.061984,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.455967741935484e-05,
+      "loss": 0.2027,
+      "step": 3874
+    },
+    {
+      "epoch": 0.062,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.455806451612904e-05,
+      "loss": 0.184,
+      "step": 3875
+    },
+    {
+      "epoch": 0.062016,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.455645161290322e-05,
+      "loss": 0.1793,
+      "step": 3876
+    },
+    {
+      "epoch": 0.062032,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.455483870967742e-05,
+      "loss": 0.166,
+      "step": 3877
+    },
+    {
+      "epoch": 0.062048,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.455322580645162e-05,
+      "loss": 0.199,
+      "step": 3878
+    },
+    {
+      "epoch": 0.062064,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.455161290322581e-05,
+      "loss": 0.2266,
+      "step": 3879
+    },
+    {
+      "epoch": 0.06208,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.455000000000001e-05,
+      "loss": 0.1626,
+      "step": 3880
+    },
+    {
+      "epoch": 0.062096,
+      "grad_norm": 0.875,
+      "learning_rate": 9.45483870967742e-05,
+      "loss": 0.1765,
+      "step": 3881
+    },
+    {
+      "epoch": 0.062112,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.45467741935484e-05,
+      "loss": 0.1769,
+      "step": 3882
+    },
+    {
+      "epoch": 0.062128,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.454516129032258e-05,
+      "loss": 0.1735,
+      "step": 3883
+    },
+    {
+      "epoch": 0.062144,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.454354838709678e-05,
+      "loss": 0.1836,
+      "step": 3884
+    },
+    {
+      "epoch": 0.06216,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.454193548387097e-05,
+      "loss": 0.2073,
+      "step": 3885
+    },
+    {
+      "epoch": 0.062176,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.454032258064516e-05,
+      "loss": 0.1729,
+      "step": 3886
+    },
+    {
+      "epoch": 0.062192,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.453870967741935e-05,
+      "loss": 0.1666,
+      "step": 3887
+    },
+    {
+      "epoch": 0.062208,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.453709677419355e-05,
+      "loss": 0.1835,
+      "step": 3888
+    },
+    {
+      "epoch": 0.062224,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.453548387096775e-05,
+      "loss": 0.192,
+      "step": 3889
+    },
+    {
+      "epoch": 0.06224,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.453387096774195e-05,
+      "loss": 0.1539,
+      "step": 3890
+    },
+    {
+      "epoch": 0.062256,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.453225806451614e-05,
+      "loss": 0.1693,
+      "step": 3891
+    },
+    {
+      "epoch": 0.062272,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.453064516129034e-05,
+      "loss": 0.1609,
+      "step": 3892
+    },
+    {
+      "epoch": 0.062288,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.452903225806452e-05,
+      "loss": 0.1894,
+      "step": 3893
+    },
+    {
+      "epoch": 0.062304,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.452741935483871e-05,
+      "loss": 0.2074,
+      "step": 3894
+    },
+    {
+      "epoch": 0.06232,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.452580645161291e-05,
+      "loss": 0.1637,
+      "step": 3895
+    },
+    {
+      "epoch": 0.062336,
+      "grad_norm": 1.3359375,
+      "learning_rate": 9.45241935483871e-05,
+      "loss": 0.2201,
+      "step": 3896
+    },
+    {
+      "epoch": 0.062352,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.452258064516129e-05,
+      "loss": 0.15,
+      "step": 3897
+    },
+    {
+      "epoch": 0.062368,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.452096774193548e-05,
+      "loss": 0.1619,
+      "step": 3898
+    },
+    {
+      "epoch": 0.062384,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.451935483870968e-05,
+      "loss": 0.2026,
+      "step": 3899
+    },
+    {
+      "epoch": 0.0624,
+      "grad_norm": 1.390625,
+      "learning_rate": 9.451774193548388e-05,
+      "loss": 0.1668,
+      "step": 3900
+    },
+    {
+      "epoch": 0.062416,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.451612903225808e-05,
+      "loss": 0.1575,
+      "step": 3901
+    },
+    {
+      "epoch": 0.062432,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.451451612903226e-05,
+      "loss": 0.1955,
+      "step": 3902
+    },
+    {
+      "epoch": 0.062448,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.451290322580646e-05,
+      "loss": 0.1821,
+      "step": 3903
+    },
+    {
+      "epoch": 0.062464,
+      "grad_norm": 1.8515625,
+      "learning_rate": 9.451129032258065e-05,
+      "loss": 0.2086,
+      "step": 3904
+    },
+    {
+      "epoch": 0.06248,
+      "grad_norm": 1.8046875,
+      "learning_rate": 9.450967741935485e-05,
+      "loss": 0.1961,
+      "step": 3905
+    },
+    {
+      "epoch": 0.062496,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.450806451612904e-05,
+      "loss": 0.1829,
+      "step": 3906
+    },
+    {
+      "epoch": 0.062512,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.450645161290324e-05,
+      "loss": 0.219,
+      "step": 3907
+    },
+    {
+      "epoch": 0.062528,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.450483870967742e-05,
+      "loss": 0.1764,
+      "step": 3908
+    },
+    {
+      "epoch": 0.062544,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.450322580645162e-05,
+      "loss": 0.194,
+      "step": 3909
+    },
+    {
+      "epoch": 0.06256,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.450161290322581e-05,
+      "loss": 0.19,
+      "step": 3910
+    },
+    {
+      "epoch": 0.062576,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.449999999999999e-05,
+      "loss": 0.157,
+      "step": 3911
+    },
+    {
+      "epoch": 0.062592,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.449838709677419e-05,
+      "loss": 0.1958,
+      "step": 3912
+    },
+    {
+      "epoch": 0.062608,
+      "grad_norm": 1.359375,
+      "learning_rate": 9.449677419354839e-05,
+      "loss": 0.1801,
+      "step": 3913
+    },
+    {
+      "epoch": 0.062624,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.449516129032259e-05,
+      "loss": 0.1672,
+      "step": 3914
+    },
+    {
+      "epoch": 0.06264,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.449354838709678e-05,
+      "loss": 0.2635,
+      "step": 3915
+    },
+    {
+      "epoch": 0.062656,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.449193548387098e-05,
+      "loss": 0.1669,
+      "step": 3916
+    },
+    {
+      "epoch": 0.062672,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.449032258064516e-05,
+      "loss": 0.178,
+      "step": 3917
+    },
+    {
+      "epoch": 0.062688,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.448870967741936e-05,
+      "loss": 0.179,
+      "step": 3918
+    },
+    {
+      "epoch": 0.062704,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.448709677419355e-05,
+      "loss": 0.1718,
+      "step": 3919
+    },
+    {
+      "epoch": 0.06272,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.448548387096775e-05,
+      "loss": 0.1762,
+      "step": 3920
+    },
+    {
+      "epoch": 0.062736,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.448387096774194e-05,
+      "loss": 0.2465,
+      "step": 3921
+    },
+    {
+      "epoch": 0.062752,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.448225806451613e-05,
+      "loss": 0.1632,
+      "step": 3922
+    },
+    {
+      "epoch": 0.062768,
+      "grad_norm": 1.0,
+      "learning_rate": 9.448064516129032e-05,
+      "loss": 0.2349,
+      "step": 3923
+    },
+    {
+      "epoch": 0.062784,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.447903225806452e-05,
+      "loss": 0.1565,
+      "step": 3924
+    },
+    {
+      "epoch": 0.0628,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.447741935483872e-05,
+      "loss": 0.1964,
+      "step": 3925
+    },
+    {
+      "epoch": 0.062816,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.44758064516129e-05,
+      "loss": 0.2098,
+      "step": 3926
+    },
+    {
+      "epoch": 0.062832,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.44741935483871e-05,
+      "loss": 0.1871,
+      "step": 3927
+    },
+    {
+      "epoch": 0.062848,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.447258064516129e-05,
+      "loss": 0.1429,
+      "step": 3928
+    },
+    {
+      "epoch": 0.062864,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.447096774193549e-05,
+      "loss": 0.1684,
+      "step": 3929
+    },
+    {
+      "epoch": 0.06288,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.446935483870968e-05,
+      "loss": 0.1731,
+      "step": 3930
+    },
+    {
+      "epoch": 0.062896,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.446774193548388e-05,
+      "loss": 0.1674,
+      "step": 3931
+    },
+    {
+      "epoch": 0.062912,
+      "grad_norm": 1.5,
+      "learning_rate": 9.446612903225806e-05,
+      "loss": 0.209,
+      "step": 3932
+    },
+    {
+      "epoch": 0.062928,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.446451612903226e-05,
+      "loss": 0.2116,
+      "step": 3933
+    },
+    {
+      "epoch": 0.062944,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.446290322580645e-05,
+      "loss": 0.1748,
+      "step": 3934
+    },
+    {
+      "epoch": 0.06296,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.446129032258065e-05,
+      "loss": 0.1756,
+      "step": 3935
+    },
+    {
+      "epoch": 0.062976,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.445967741935485e-05,
+      "loss": 0.1867,
+      "step": 3936
+    },
+    {
+      "epoch": 0.062992,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.445806451612905e-05,
+      "loss": 0.2114,
+      "step": 3937
+    },
+    {
+      "epoch": 0.063008,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.445645161290323e-05,
+      "loss": 0.1768,
+      "step": 3938
+    },
+    {
+      "epoch": 0.063024,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.445483870967743e-05,
+      "loss": 0.1756,
+      "step": 3939
+    },
+    {
+      "epoch": 0.06304,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.445322580645162e-05,
+      "loss": 0.185,
+      "step": 3940
+    },
+    {
+      "epoch": 0.063056,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.44516129032258e-05,
+      "loss": 0.2184,
+      "step": 3941
+    },
+    {
+      "epoch": 0.063072,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.445e-05,
+      "loss": 0.2056,
+      "step": 3942
+    },
+    {
+      "epoch": 0.063088,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.444838709677419e-05,
+      "loss": 0.1915,
+      "step": 3943
+    },
+    {
+      "epoch": 0.063104,
+      "grad_norm": 1.3359375,
+      "learning_rate": 9.444677419354839e-05,
+      "loss": 0.207,
+      "step": 3944
+    },
+    {
+      "epoch": 0.06312,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.444516129032258e-05,
+      "loss": 0.2125,
+      "step": 3945
+    },
+    {
+      "epoch": 0.063136,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.444354838709678e-05,
+      "loss": 0.1847,
+      "step": 3946
+    },
+    {
+      "epoch": 0.063152,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.444193548387096e-05,
+      "loss": 0.1813,
+      "step": 3947
+    },
+    {
+      "epoch": 0.063168,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.444032258064516e-05,
+      "loss": 0.1555,
+      "step": 3948
+    },
+    {
+      "epoch": 0.063184,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.443870967741936e-05,
+      "loss": 0.1578,
+      "step": 3949
+    },
+    {
+      "epoch": 0.0632,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.443709677419356e-05,
+      "loss": 0.1647,
+      "step": 3950
+    },
+    {
+      "epoch": 0.063216,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.443548387096775e-05,
+      "loss": 0.1717,
+      "step": 3951
+    },
+    {
+      "epoch": 0.063232,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.443387096774195e-05,
+      "loss": 0.1801,
+      "step": 3952
+    },
+    {
+      "epoch": 0.063248,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.443225806451613e-05,
+      "loss": 0.1967,
+      "step": 3953
+    },
+    {
+      "epoch": 0.063264,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.443064516129033e-05,
+      "loss": 0.1829,
+      "step": 3954
+    },
+    {
+      "epoch": 0.06328,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.442903225806452e-05,
+      "loss": 0.2066,
+      "step": 3955
+    },
+    {
+      "epoch": 0.063296,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.44274193548387e-05,
+      "loss": 0.1892,
+      "step": 3956
+    },
+    {
+      "epoch": 0.063312,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.44258064516129e-05,
+      "loss": 0.2113,
+      "step": 3957
+    },
+    {
+      "epoch": 0.063328,
+      "grad_norm": 0.875,
+      "learning_rate": 9.442419354838709e-05,
+      "loss": 0.1544,
+      "step": 3958
+    },
+    {
+      "epoch": 0.063344,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.442258064516129e-05,
+      "loss": 0.2131,
+      "step": 3959
+    },
+    {
+      "epoch": 0.06336,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.442096774193549e-05,
+      "loss": 0.1406,
+      "step": 3960
+    },
+    {
+      "epoch": 0.063376,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.441935483870969e-05,
+      "loss": 0.1607,
+      "step": 3961
+    },
+    {
+      "epoch": 0.063392,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.441774193548388e-05,
+      "loss": 0.15,
+      "step": 3962
+    },
+    {
+      "epoch": 0.063408,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.441612903225808e-05,
+      "loss": 0.1678,
+      "step": 3963
+    },
+    {
+      "epoch": 0.063424,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.441451612903226e-05,
+      "loss": 0.186,
+      "step": 3964
+    },
+    {
+      "epoch": 0.06344,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.441290322580646e-05,
+      "loss": 0.1597,
+      "step": 3965
+    },
+    {
+      "epoch": 0.063456,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.441129032258065e-05,
+      "loss": 0.1671,
+      "step": 3966
+    },
+    {
+      "epoch": 0.063472,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.440967741935485e-05,
+      "loss": 0.1747,
+      "step": 3967
+    },
+    {
+      "epoch": 0.063488,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.440806451612903e-05,
+      "loss": 0.185,
+      "step": 3968
+    },
+    {
+      "epoch": 0.063504,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.440645161290323e-05,
+      "loss": 0.1872,
+      "step": 3969
+    },
+    {
+      "epoch": 0.06352,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.440483870967742e-05,
+      "loss": 0.2251,
+      "step": 3970
+    },
+    {
+      "epoch": 0.063536,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.440322580645162e-05,
+      "loss": 0.1898,
+      "step": 3971
+    },
+    {
+      "epoch": 0.063552,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.44016129032258e-05,
+      "loss": 0.1708,
+      "step": 3972
+    },
+    {
+      "epoch": 0.063568,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.44e-05,
+      "loss": 0.2076,
+      "step": 3973
+    },
+    {
+      "epoch": 0.063584,
+      "grad_norm": 1.375,
+      "learning_rate": 9.43983870967742e-05,
+      "loss": 0.2299,
+      "step": 3974
+    },
+    {
+      "epoch": 0.0636,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.439677419354839e-05,
+      "loss": 0.1638,
+      "step": 3975
+    },
+    {
+      "epoch": 0.063616,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.439516129032259e-05,
+      "loss": 0.1345,
+      "step": 3976
+    },
+    {
+      "epoch": 0.063632,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.439354838709678e-05,
+      "loss": 0.1956,
+      "step": 3977
+    },
+    {
+      "epoch": 0.063648,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.439193548387098e-05,
+      "loss": 0.185,
+      "step": 3978
+    },
+    {
+      "epoch": 0.063664,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.439032258064516e-05,
+      "loss": 0.198,
+      "step": 3979
+    },
+    {
+      "epoch": 0.06368,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.438870967741936e-05,
+      "loss": 0.182,
+      "step": 3980
+    },
+    {
+      "epoch": 0.063696,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.438709677419355e-05,
+      "loss": 0.1874,
+      "step": 3981
+    },
+    {
+      "epoch": 0.063712,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.438548387096775e-05,
+      "loss": 0.2083,
+      "step": 3982
+    },
+    {
+      "epoch": 0.063728,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.438387096774193e-05,
+      "loss": 0.2091,
+      "step": 3983
+    },
+    {
+      "epoch": 0.063744,
+      "grad_norm": 1.375,
+      "learning_rate": 9.438225806451613e-05,
+      "loss": 0.2094,
+      "step": 3984
+    },
+    {
+      "epoch": 0.06376,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.438064516129033e-05,
+      "loss": 0.1594,
+      "step": 3985
+    },
+    {
+      "epoch": 0.063776,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.437903225806453e-05,
+      "loss": 0.1893,
+      "step": 3986
+    },
+    {
+      "epoch": 0.063792,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.437741935483872e-05,
+      "loss": 0.1633,
+      "step": 3987
+    },
+    {
+      "epoch": 0.063808,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.43758064516129e-05,
+      "loss": 0.2005,
+      "step": 3988
+    },
+    {
+      "epoch": 0.063824,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.43741935483871e-05,
+      "loss": 0.1923,
+      "step": 3989
+    },
+    {
+      "epoch": 0.06384,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.437258064516129e-05,
+      "loss": 0.1761,
+      "step": 3990
+    },
+    {
+      "epoch": 0.063856,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.437096774193549e-05,
+      "loss": 0.1851,
+      "step": 3991
+    },
+    {
+      "epoch": 0.063872,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.436935483870968e-05,
+      "loss": 0.188,
+      "step": 3992
+    },
+    {
+      "epoch": 0.063888,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.436774193548387e-05,
+      "loss": 0.176,
+      "step": 3993
+    },
+    {
+      "epoch": 0.063904,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.436612903225806e-05,
+      "loss": 0.1928,
+      "step": 3994
+    },
+    {
+      "epoch": 0.06392,
+      "grad_norm": 1.4296875,
+      "learning_rate": 9.436451612903226e-05,
+      "loss": 0.2359,
+      "step": 3995
+    },
+    {
+      "epoch": 0.063936,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.436290322580646e-05,
+      "loss": 0.1431,
+      "step": 3996
+    },
+    {
+      "epoch": 0.063952,
+      "grad_norm": 1.5703125,
+      "learning_rate": 9.436129032258066e-05,
+      "loss": 0.2064,
+      "step": 3997
+    },
+    {
+      "epoch": 0.063968,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.435967741935485e-05,
+      "loss": 0.2037,
+      "step": 3998
+    },
+    {
+      "epoch": 0.063984,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.435806451612905e-05,
+      "loss": 0.1969,
+      "step": 3999
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.435645161290323e-05,
+      "loss": 0.1644,
+      "step": 4000
+    },
+    {
+      "epoch": 0.064016,
+      "grad_norm": 1.328125,
+      "learning_rate": 9.435483870967743e-05,
+      "loss": 0.2575,
+      "step": 4001
+    },
+    {
+      "epoch": 0.064032,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.435322580645162e-05,
+      "loss": 0.1753,
+      "step": 4002
+    },
+    {
+      "epoch": 0.064048,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.43516129032258e-05,
+      "loss": 0.1959,
+      "step": 4003
+    },
+    {
+      "epoch": 0.064064,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.435e-05,
+      "loss": 0.1613,
+      "step": 4004
+    },
+    {
+      "epoch": 0.06408,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.434838709677419e-05,
+      "loss": 0.206,
+      "step": 4005
+    },
+    {
+      "epoch": 0.064096,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.434677419354839e-05,
+      "loss": 0.2054,
+      "step": 4006
+    },
+    {
+      "epoch": 0.064112,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.434516129032257e-05,
+      "loss": 0.1724,
+      "step": 4007
+    },
+    {
+      "epoch": 0.064128,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.434354838709677e-05,
+      "loss": 0.1959,
+      "step": 4008
+    },
+    {
+      "epoch": 0.064144,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.434193548387097e-05,
+      "loss": 0.1526,
+      "step": 4009
+    },
+    {
+      "epoch": 0.06416,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.434032258064517e-05,
+      "loss": 0.1479,
+      "step": 4010
+    },
+    {
+      "epoch": 0.064176,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.433870967741936e-05,
+      "loss": 0.1588,
+      "step": 4011
+    },
+    {
+      "epoch": 0.064192,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.433709677419356e-05,
+      "loss": 0.1616,
+      "step": 4012
+    },
+    {
+      "epoch": 0.064208,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.433548387096775e-05,
+      "loss": 0.1768,
+      "step": 4013
+    },
+    {
+      "epoch": 0.064224,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.433387096774195e-05,
+      "loss": 0.2293,
+      "step": 4014
+    },
+    {
+      "epoch": 0.06424,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.433225806451613e-05,
+      "loss": 0.2168,
+      "step": 4015
+    },
+    {
+      "epoch": 0.064256,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.433064516129033e-05,
+      "loss": 0.1768,
+      "step": 4016
+    },
+    {
+      "epoch": 0.064272,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.432903225806452e-05,
+      "loss": 0.1752,
+      "step": 4017
+    },
+    {
+      "epoch": 0.064288,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.432741935483872e-05,
+      "loss": 0.1727,
+      "step": 4018
+    },
+    {
+      "epoch": 0.064304,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.43258064516129e-05,
+      "loss": 0.2083,
+      "step": 4019
+    },
+    {
+      "epoch": 0.06432,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.43241935483871e-05,
+      "loss": 0.2386,
+      "step": 4020
+    },
+    {
+      "epoch": 0.064336,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.43225806451613e-05,
+      "loss": 0.2765,
+      "step": 4021
+    },
+    {
+      "epoch": 0.064352,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.432096774193549e-05,
+      "loss": 0.171,
+      "step": 4022
+    },
+    {
+      "epoch": 0.064368,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.431935483870969e-05,
+      "loss": 0.1786,
+      "step": 4023
+    },
+    {
+      "epoch": 0.064384,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.431774193548387e-05,
+      "loss": 0.1654,
+      "step": 4024
+    },
+    {
+      "epoch": 0.0644,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.431612903225807e-05,
+      "loss": 0.2044,
+      "step": 4025
+    },
+    {
+      "epoch": 0.064416,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.431451612903226e-05,
+      "loss": 0.2164,
+      "step": 4026
+    },
+    {
+      "epoch": 0.064432,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.431290322580646e-05,
+      "loss": 0.2256,
+      "step": 4027
+    },
+    {
+      "epoch": 0.064448,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.431129032258065e-05,
+      "loss": 0.1734,
+      "step": 4028
+    },
+    {
+      "epoch": 0.064464,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.430967741935484e-05,
+      "loss": 0.1774,
+      "step": 4029
+    },
+    {
+      "epoch": 0.06448,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.430806451612903e-05,
+      "loss": 0.1608,
+      "step": 4030
+    },
+    {
+      "epoch": 0.064496,
+      "grad_norm": 1.421875,
+      "learning_rate": 9.430645161290323e-05,
+      "loss": 0.1579,
+      "step": 4031
+    },
+    {
+      "epoch": 0.064512,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.430483870967743e-05,
+      "loss": 0.1942,
+      "step": 4032
+    },
+    {
+      "epoch": 0.064528,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.430322580645162e-05,
+      "loss": 0.125,
+      "step": 4033
+    },
+    {
+      "epoch": 0.064544,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.430161290322582e-05,
+      "loss": 0.1678,
+      "step": 4034
+    },
+    {
+      "epoch": 0.06456,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.43e-05,
+      "loss": 0.1602,
+      "step": 4035
+    },
+    {
+      "epoch": 0.064576,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.42983870967742e-05,
+      "loss": 0.1786,
+      "step": 4036
+    },
+    {
+      "epoch": 0.064592,
+      "grad_norm": 1.3515625,
+      "learning_rate": 9.429677419354839e-05,
+      "loss": 0.1634,
+      "step": 4037
+    },
+    {
+      "epoch": 0.064608,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.429516129032259e-05,
+      "loss": 0.1976,
+      "step": 4038
+    },
+    {
+      "epoch": 0.064624,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.429354838709677e-05,
+      "loss": 0.1673,
+      "step": 4039
+    },
+    {
+      "epoch": 0.06464,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.429193548387097e-05,
+      "loss": 0.1943,
+      "step": 4040
+    },
+    {
+      "epoch": 0.064656,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.429032258064516e-05,
+      "loss": 0.1874,
+      "step": 4041
+    },
+    {
+      "epoch": 0.064672,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.428870967741936e-05,
+      "loss": 0.1781,
+      "step": 4042
+    },
+    {
+      "epoch": 0.064688,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.428709677419354e-05,
+      "loss": 0.1573,
+      "step": 4043
+    },
+    {
+      "epoch": 0.064704,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.428548387096774e-05,
+      "loss": 0.1659,
+      "step": 4044
+    },
+    {
+      "epoch": 0.06472,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.428387096774194e-05,
+      "loss": 0.1342,
+      "step": 4045
+    },
+    {
+      "epoch": 0.064736,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.428225806451614e-05,
+      "loss": 0.2052,
+      "step": 4046
+    },
+    {
+      "epoch": 0.064752,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.428064516129033e-05,
+      "loss": 0.2094,
+      "step": 4047
+    },
+    {
+      "epoch": 0.064768,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.427903225806453e-05,
+      "loss": 0.139,
+      "step": 4048
+    },
+    {
+      "epoch": 0.064784,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.427741935483872e-05,
+      "loss": 0.1759,
+      "step": 4049
+    },
+    {
+      "epoch": 0.0648,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.42758064516129e-05,
+      "loss": 0.1651,
+      "step": 4050
+    },
+    {
+      "epoch": 0.064816,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.42741935483871e-05,
+      "loss": 0.1575,
+      "step": 4051
+    },
+    {
+      "epoch": 0.064832,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.427258064516129e-05,
+      "loss": 0.2137,
+      "step": 4052
+    },
+    {
+      "epoch": 0.064848,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.427096774193549e-05,
+      "loss": 0.1729,
+      "step": 4053
+    },
+    {
+      "epoch": 0.064864,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.426935483870967e-05,
+      "loss": 0.1628,
+      "step": 4054
+    },
+    {
+      "epoch": 0.06488,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.426774193548387e-05,
+      "loss": 0.1486,
+      "step": 4055
+    },
+    {
+      "epoch": 0.064896,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.426612903225807e-05,
+      "loss": 0.1639,
+      "step": 4056
+    },
+    {
+      "epoch": 0.064912,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.426451612903227e-05,
+      "loss": 0.1925,
+      "step": 4057
+    },
+    {
+      "epoch": 0.064928,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.426290322580646e-05,
+      "loss": 0.155,
+      "step": 4058
+    },
+    {
+      "epoch": 0.064944,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.426129032258066e-05,
+      "loss": 0.1689,
+      "step": 4059
+    },
+    {
+      "epoch": 0.06496,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.425967741935484e-05,
+      "loss": 0.1772,
+      "step": 4060
+    },
+    {
+      "epoch": 0.064976,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.425806451612904e-05,
+      "loss": 0.1427,
+      "step": 4061
+    },
+    {
+      "epoch": 0.064992,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.425645161290323e-05,
+      "loss": 0.1535,
+      "step": 4062
+    },
+    {
+      "epoch": 0.065008,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.425483870967743e-05,
+      "loss": 0.1907,
+      "step": 4063
+    },
+    {
+      "epoch": 0.065024,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.425322580645161e-05,
+      "loss": 0.1767,
+      "step": 4064
+    },
+    {
+      "epoch": 0.06504,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.42516129032258e-05,
+      "loss": 0.1755,
+      "step": 4065
+    },
+    {
+      "epoch": 0.065056,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.425e-05,
+      "loss": 0.1739,
+      "step": 4066
+    },
+    {
+      "epoch": 0.065072,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.424838709677419e-05,
+      "loss": 0.2038,
+      "step": 4067
+    },
+    {
+      "epoch": 0.065088,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.424677419354839e-05,
+      "loss": 0.1887,
+      "step": 4068
+    },
+    {
+      "epoch": 0.065104,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.424516129032259e-05,
+      "loss": 0.2033,
+      "step": 4069
+    },
+    {
+      "epoch": 0.06512,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.424354838709679e-05,
+      "loss": 0.1882,
+      "step": 4070
+    },
+    {
+      "epoch": 0.065136,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.424193548387097e-05,
+      "loss": 0.1828,
+      "step": 4071
+    },
+    {
+      "epoch": 0.065152,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.424032258064517e-05,
+      "loss": 0.1584,
+      "step": 4072
+    },
+    {
+      "epoch": 0.065168,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.423870967741936e-05,
+      "loss": 0.178,
+      "step": 4073
+    },
+    {
+      "epoch": 0.065184,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.423709677419356e-05,
+      "loss": 0.1931,
+      "step": 4074
+    },
+    {
+      "epoch": 0.0652,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.423548387096774e-05,
+      "loss": 0.1882,
+      "step": 4075
+    },
+    {
+      "epoch": 0.065216,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.423387096774194e-05,
+      "loss": 0.1514,
+      "step": 4076
+    },
+    {
+      "epoch": 0.065232,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.423225806451613e-05,
+      "loss": 0.1693,
+      "step": 4077
+    },
+    {
+      "epoch": 0.065248,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.423064516129033e-05,
+      "loss": 0.2341,
+      "step": 4078
+    },
+    {
+      "epoch": 0.065264,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.422903225806451e-05,
+      "loss": 0.1558,
+      "step": 4079
+    },
+    {
+      "epoch": 0.06528,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.422741935483871e-05,
+      "loss": 0.1552,
+      "step": 4080
+    },
+    {
+      "epoch": 0.065296,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.422580645161291e-05,
+      "loss": 0.2009,
+      "step": 4081
+    },
+    {
+      "epoch": 0.065312,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.42241935483871e-05,
+      "loss": 0.2308,
+      "step": 4082
+    },
+    {
+      "epoch": 0.065328,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.42225806451613e-05,
+      "loss": 0.1646,
+      "step": 4083
+    },
+    {
+      "epoch": 0.065344,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.422096774193549e-05,
+      "loss": 0.1544,
+      "step": 4084
+    },
+    {
+      "epoch": 0.06536,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.421935483870969e-05,
+      "loss": 0.1583,
+      "step": 4085
+    },
+    {
+      "epoch": 0.065376,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.421774193548387e-05,
+      "loss": 0.1552,
+      "step": 4086
+    },
+    {
+      "epoch": 0.065392,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.421612903225807e-05,
+      "loss": 0.1872,
+      "step": 4087
+    },
+    {
+      "epoch": 0.065408,
+      "grad_norm": 1.4296875,
+      "learning_rate": 9.421451612903226e-05,
+      "loss": 0.1508,
+      "step": 4088
+    },
+    {
+      "epoch": 0.065424,
+      "grad_norm": 1.375,
+      "learning_rate": 9.421290322580646e-05,
+      "loss": 0.1732,
+      "step": 4089
+    },
+    {
+      "epoch": 0.06544,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.421129032258064e-05,
+      "loss": 0.2067,
+      "step": 4090
+    },
+    {
+      "epoch": 0.065456,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.420967741935484e-05,
+      "loss": 0.1686,
+      "step": 4091
+    },
+    {
+      "epoch": 0.065472,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.420806451612904e-05,
+      "loss": 0.2068,
+      "step": 4092
+    },
+    {
+      "epoch": 0.065488,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.420645161290324e-05,
+      "loss": 0.1488,
+      "step": 4093
+    },
+    {
+      "epoch": 0.065504,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.420483870967743e-05,
+      "loss": 0.2164,
+      "step": 4094
+    },
+    {
+      "epoch": 0.06552,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.420322580645163e-05,
+      "loss": 0.1711,
+      "step": 4095
+    },
+    {
+      "epoch": 0.065536,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.420161290322581e-05,
+      "loss": 0.153,
+      "step": 4096
+    },
+    {
+      "epoch": 0.065552,
+      "grad_norm": 1.5,
+      "learning_rate": 9.42e-05,
+      "loss": 0.2149,
+      "step": 4097
+    },
+    {
+      "epoch": 0.065568,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.41983870967742e-05,
+      "loss": 0.2173,
+      "step": 4098
+    },
+    {
+      "epoch": 0.065584,
+      "grad_norm": 1.375,
+      "learning_rate": 9.419677419354839e-05,
+      "loss": 0.1858,
+      "step": 4099
+    },
+    {
+      "epoch": 0.0656,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.419516129032258e-05,
+      "loss": 0.2045,
+      "step": 4100
+    },
+    {
+      "epoch": 0.065616,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.419354838709677e-05,
+      "loss": 0.1909,
+      "step": 4101
+    },
+    {
+      "epoch": 0.065632,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.419193548387097e-05,
+      "loss": 0.1801,
+      "step": 4102
+    },
+    {
+      "epoch": 0.065648,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.419032258064516e-05,
+      "loss": 0.1756,
+      "step": 4103
+    },
+    {
+      "epoch": 0.065664,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.418870967741936e-05,
+      "loss": 0.1851,
+      "step": 4104
+    },
+    {
+      "epoch": 0.06568,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.418709677419356e-05,
+      "loss": 0.2052,
+      "step": 4105
+    },
+    {
+      "epoch": 0.065696,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.418548387096776e-05,
+      "loss": 0.2012,
+      "step": 4106
+    },
+    {
+      "epoch": 0.065712,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.418387096774194e-05,
+      "loss": 0.2412,
+      "step": 4107
+    },
+    {
+      "epoch": 0.065728,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.418225806451614e-05,
+      "loss": 0.1707,
+      "step": 4108
+    },
+    {
+      "epoch": 0.065744,
+      "grad_norm": 1.421875,
+      "learning_rate": 9.418064516129033e-05,
+      "loss": 0.1459,
+      "step": 4109
+    },
+    {
+      "epoch": 0.06576,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.417903225806453e-05,
+      "loss": 0.2188,
+      "step": 4110
+    },
+    {
+      "epoch": 0.065776,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.417741935483871e-05,
+      "loss": 0.1689,
+      "step": 4111
+    },
+    {
+      "epoch": 0.065792,
+      "grad_norm": 1.359375,
+      "learning_rate": 9.41758064516129e-05,
+      "loss": 0.223,
+      "step": 4112
+    },
+    {
+      "epoch": 0.065808,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.41741935483871e-05,
+      "loss": 0.1777,
+      "step": 4113
+    },
+    {
+      "epoch": 0.065824,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.417258064516128e-05,
+      "loss": 0.1627,
+      "step": 4114
+    },
+    {
+      "epoch": 0.06584,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.417096774193548e-05,
+      "loss": 0.1886,
+      "step": 4115
+    },
+    {
+      "epoch": 0.065856,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.416935483870968e-05,
+      "loss": 0.1919,
+      "step": 4116
+    },
+    {
+      "epoch": 0.065872,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.416774193548388e-05,
+      "loss": 0.2483,
+      "step": 4117
+    },
+    {
+      "epoch": 0.065888,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.416612903225807e-05,
+      "loss": 0.1593,
+      "step": 4118
+    },
+    {
+      "epoch": 0.065904,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.416451612903227e-05,
+      "loss": 0.1582,
+      "step": 4119
+    },
+    {
+      "epoch": 0.06592,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.416290322580646e-05,
+      "loss": 0.1956,
+      "step": 4120
+    },
+    {
+      "epoch": 0.065936,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.416129032258065e-05,
+      "loss": 0.1811,
+      "step": 4121
+    },
+    {
+      "epoch": 0.065952,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.415967741935484e-05,
+      "loss": 0.1431,
+      "step": 4122
+    },
+    {
+      "epoch": 0.065968,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.415806451612904e-05,
+      "loss": 0.2013,
+      "step": 4123
+    },
+    {
+      "epoch": 0.065984,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.415645161290323e-05,
+      "loss": 0.1959,
+      "step": 4124
+    },
+    {
+      "epoch": 0.066,
+      "grad_norm": 1.3515625,
+      "learning_rate": 9.415483870967743e-05,
+      "loss": 0.1821,
+      "step": 4125
+    },
+    {
+      "epoch": 0.066016,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.415322580645161e-05,
+      "loss": 0.1872,
+      "step": 4126
+    },
+    {
+      "epoch": 0.066032,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.415161290322581e-05,
+      "loss": 0.1742,
+      "step": 4127
+    },
+    {
+      "epoch": 0.066048,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.415e-05,
+      "loss": 0.1718,
+      "step": 4128
+    },
+    {
+      "epoch": 0.066064,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.41483870967742e-05,
+      "loss": 0.179,
+      "step": 4129
+    },
+    {
+      "epoch": 0.06608,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.41467741935484e-05,
+      "loss": 0.1998,
+      "step": 4130
+    },
+    {
+      "epoch": 0.066096,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.414516129032258e-05,
+      "loss": 0.1861,
+      "step": 4131
+    },
+    {
+      "epoch": 0.066112,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.414354838709678e-05,
+      "loss": 0.1546,
+      "step": 4132
+    },
+    {
+      "epoch": 0.066128,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.414193548387097e-05,
+      "loss": 0.186,
+      "step": 4133
+    },
+    {
+      "epoch": 0.066144,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.414032258064517e-05,
+      "loss": 0.1694,
+      "step": 4134
+    },
+    {
+      "epoch": 0.06616,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.413870967741935e-05,
+      "loss": 0.1826,
+      "step": 4135
+    },
+    {
+      "epoch": 0.066176,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.413709677419355e-05,
+      "loss": 0.1538,
+      "step": 4136
+    },
+    {
+      "epoch": 0.066192,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.413548387096774e-05,
+      "loss": 0.2155,
+      "step": 4137
+    },
+    {
+      "epoch": 0.066208,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.413387096774194e-05,
+      "loss": 0.2216,
+      "step": 4138
+    },
+    {
+      "epoch": 0.066224,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.413225806451613e-05,
+      "loss": 0.1704,
+      "step": 4139
+    },
+    {
+      "epoch": 0.06624,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.413064516129033e-05,
+      "loss": 0.2068,
+      "step": 4140
+    },
+    {
+      "epoch": 0.066256,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.412903225806453e-05,
+      "loss": 0.185,
+      "step": 4141
+    },
+    {
+      "epoch": 0.066272,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.412741935483873e-05,
+      "loss": 0.1907,
+      "step": 4142
+    },
+    {
+      "epoch": 0.066288,
+      "grad_norm": 1.3125,
+      "learning_rate": 9.412580645161291e-05,
+      "loss": 0.2148,
+      "step": 4143
+    },
+    {
+      "epoch": 0.066304,
+      "grad_norm": 0.75,
+      "learning_rate": 9.41241935483871e-05,
+      "loss": 0.1775,
+      "step": 4144
+    },
+    {
+      "epoch": 0.06632,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.41225806451613e-05,
+      "loss": 0.1533,
+      "step": 4145
+    },
+    {
+      "epoch": 0.066336,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.412096774193548e-05,
+      "loss": 0.1927,
+      "step": 4146
+    },
+    {
+      "epoch": 0.066352,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.411935483870968e-05,
+      "loss": 0.1853,
+      "step": 4147
+    },
+    {
+      "epoch": 0.066368,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.411774193548387e-05,
+      "loss": 0.198,
+      "step": 4148
+    },
+    {
+      "epoch": 0.066384,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.411612903225807e-05,
+      "loss": 0.1435,
+      "step": 4149
+    },
+    {
+      "epoch": 0.0664,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.411451612903225e-05,
+      "loss": 0.2319,
+      "step": 4150
+    },
+    {
+      "epoch": 0.066416,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.411290322580645e-05,
+      "loss": 0.1678,
+      "step": 4151
+    },
+    {
+      "epoch": 0.066432,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.411129032258065e-05,
+      "loss": 0.1645,
+      "step": 4152
+    },
+    {
+      "epoch": 0.066448,
+      "grad_norm": 1.7578125,
+      "learning_rate": 9.410967741935485e-05,
+      "loss": 0.1772,
+      "step": 4153
+    },
+    {
+      "epoch": 0.066464,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.410806451612904e-05,
+      "loss": 0.1952,
+      "step": 4154
+    },
+    {
+      "epoch": 0.06648,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.410645161290324e-05,
+      "loss": 0.1787,
+      "step": 4155
+    },
+    {
+      "epoch": 0.066496,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.410483870967743e-05,
+      "loss": 0.1878,
+      "step": 4156
+    },
+    {
+      "epoch": 0.066512,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.410322580645162e-05,
+      "loss": 0.1614,
+      "step": 4157
+    },
+    {
+      "epoch": 0.066528,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.410161290322581e-05,
+      "loss": 0.2038,
+      "step": 4158
+    },
+    {
+      "epoch": 0.066544,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.41e-05,
+      "loss": 0.1834,
+      "step": 4159
+    },
+    {
+      "epoch": 0.06656,
+      "grad_norm": 0.625,
+      "learning_rate": 9.40983870967742e-05,
+      "loss": 0.1759,
+      "step": 4160
+    },
+    {
+      "epoch": 0.066576,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.409677419354838e-05,
+      "loss": 0.1912,
+      "step": 4161
+    },
+    {
+      "epoch": 0.066592,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.409516129032258e-05,
+      "loss": 0.1753,
+      "step": 4162
+    },
+    {
+      "epoch": 0.066608,
+      "grad_norm": 0.625,
+      "learning_rate": 9.409354838709677e-05,
+      "loss": 0.188,
+      "step": 4163
+    },
+    {
+      "epoch": 0.066624,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.409193548387097e-05,
+      "loss": 0.1973,
+      "step": 4164
+    },
+    {
+      "epoch": 0.06664,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.409032258064517e-05,
+      "loss": 0.1864,
+      "step": 4165
+    },
+    {
+      "epoch": 0.066656,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.408870967741937e-05,
+      "loss": 0.2044,
+      "step": 4166
+    },
+    {
+      "epoch": 0.066672,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.408709677419355e-05,
+      "loss": 0.1737,
+      "step": 4167
+    },
+    {
+      "epoch": 0.066688,
+      "grad_norm": 1.5078125,
+      "learning_rate": 9.408548387096775e-05,
+      "loss": 0.182,
+      "step": 4168
+    },
+    {
+      "epoch": 0.066704,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.408387096774194e-05,
+      "loss": 0.186,
+      "step": 4169
+    },
+    {
+      "epoch": 0.06672,
+      "grad_norm": 1.4375,
+      "learning_rate": 9.408225806451614e-05,
+      "loss": 0.2534,
+      "step": 4170
+    },
+    {
+      "epoch": 0.066736,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.408064516129032e-05,
+      "loss": 0.2023,
+      "step": 4171
+    },
+    {
+      "epoch": 0.066752,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.407903225806452e-05,
+      "loss": 0.1629,
+      "step": 4172
+    },
+    {
+      "epoch": 0.066768,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.407741935483871e-05,
+      "loss": 0.189,
+      "step": 4173
+    },
+    {
+      "epoch": 0.066784,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.40758064516129e-05,
+      "loss": 0.1662,
+      "step": 4174
+    },
+    {
+      "epoch": 0.0668,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.40741935483871e-05,
+      "loss": 0.1726,
+      "step": 4175
+    },
+    {
+      "epoch": 0.066816,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.40725806451613e-05,
+      "loss": 0.1781,
+      "step": 4176
+    },
+    {
+      "epoch": 0.066832,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.40709677419355e-05,
+      "loss": 0.1474,
+      "step": 4177
+    },
+    {
+      "epoch": 0.066848,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.406935483870968e-05,
+      "loss": 0.1916,
+      "step": 4178
+    },
+    {
+      "epoch": 0.066864,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.406774193548388e-05,
+      "loss": 0.2129,
+      "step": 4179
+    },
+    {
+      "epoch": 0.06688,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.406612903225807e-05,
+      "loss": 0.1796,
+      "step": 4180
+    },
+    {
+      "epoch": 0.066896,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.406451612903227e-05,
+      "loss": 0.196,
+      "step": 4181
+    },
+    {
+      "epoch": 0.066912,
+      "grad_norm": 1.4375,
+      "learning_rate": 9.406290322580645e-05,
+      "loss": 0.2385,
+      "step": 4182
+    },
+    {
+      "epoch": 0.066928,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.406129032258065e-05,
+      "loss": 0.1476,
+      "step": 4183
+    },
+    {
+      "epoch": 0.066944,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.405967741935484e-05,
+      "loss": 0.1873,
+      "step": 4184
+    },
+    {
+      "epoch": 0.06696,
+      "grad_norm": 1.125,
+      "learning_rate": 9.405806451612904e-05,
+      "loss": 0.1902,
+      "step": 4185
+    },
+    {
+      "epoch": 0.066976,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.405645161290322e-05,
+      "loss": 0.1416,
+      "step": 4186
+    },
+    {
+      "epoch": 0.066992,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.405483870967742e-05,
+      "loss": 0.1737,
+      "step": 4187
+    },
+    {
+      "epoch": 0.067008,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.405322580645162e-05,
+      "loss": 0.2056,
+      "step": 4188
+    },
+    {
+      "epoch": 0.067024,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.405161290322582e-05,
+      "loss": 0.1641,
+      "step": 4189
+    },
+    {
+      "epoch": 0.06704,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.405000000000001e-05,
+      "loss": 0.2196,
+      "step": 4190
+    },
+    {
+      "epoch": 0.067056,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.40483870967742e-05,
+      "loss": 0.2038,
+      "step": 4191
+    },
+    {
+      "epoch": 0.067072,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.40467741935484e-05,
+      "loss": 0.1876,
+      "step": 4192
+    },
+    {
+      "epoch": 0.067088,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.404516129032258e-05,
+      "loss": 0.1757,
+      "step": 4193
+    },
+    {
+      "epoch": 0.067104,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.404354838709678e-05,
+      "loss": 0.1511,
+      "step": 4194
+    },
+    {
+      "epoch": 0.06712,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.404193548387097e-05,
+      "loss": 0.1925,
+      "step": 4195
+    },
+    {
+      "epoch": 0.067136,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.404032258064517e-05,
+      "loss": 0.1638,
+      "step": 4196
+    },
+    {
+      "epoch": 0.067152,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.403870967741935e-05,
+      "loss": 0.1706,
+      "step": 4197
+    },
+    {
+      "epoch": 0.067168,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.403709677419355e-05,
+      "loss": 0.1821,
+      "step": 4198
+    },
+    {
+      "epoch": 0.067184,
+      "grad_norm": 0.49609375,
+      "learning_rate": 9.403548387096774e-05,
+      "loss": 0.1522,
+      "step": 4199
+    },
+    {
+      "epoch": 0.0672,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.403387096774194e-05,
+      "loss": 0.1742,
+      "step": 4200
+    },
+    {
+      "epoch": 0.067216,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.403225806451614e-05,
+      "loss": 0.1639,
+      "step": 4201
+    },
+    {
+      "epoch": 0.067232,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.403064516129034e-05,
+      "loss": 0.1909,
+      "step": 4202
+    },
+    {
+      "epoch": 0.067248,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.402903225806452e-05,
+      "loss": 0.235,
+      "step": 4203
+    },
+    {
+      "epoch": 0.067264,
+      "grad_norm": 1.3125,
+      "learning_rate": 9.402741935483872e-05,
+      "loss": 0.1739,
+      "step": 4204
+    },
+    {
+      "epoch": 0.06728,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.402580645161291e-05,
+      "loss": 0.1686,
+      "step": 4205
+    },
+    {
+      "epoch": 0.067296,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.40241935483871e-05,
+      "loss": 0.1854,
+      "step": 4206
+    },
+    {
+      "epoch": 0.067312,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.40225806451613e-05,
+      "loss": 0.1897,
+      "step": 4207
+    },
+    {
+      "epoch": 0.067328,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.402096774193548e-05,
+      "loss": 0.2003,
+      "step": 4208
+    },
+    {
+      "epoch": 0.067344,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.401935483870968e-05,
+      "loss": 0.1636,
+      "step": 4209
+    },
+    {
+      "epoch": 0.06736,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.401774193548387e-05,
+      "loss": 0.2688,
+      "step": 4210
+    },
+    {
+      "epoch": 0.067376,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.401612903225807e-05,
+      "loss": 0.211,
+      "step": 4211
+    },
+    {
+      "epoch": 0.067392,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.401451612903227e-05,
+      "loss": 0.2069,
+      "step": 4212
+    },
+    {
+      "epoch": 0.067408,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.401290322580647e-05,
+      "loss": 0.2139,
+      "step": 4213
+    },
+    {
+      "epoch": 0.067424,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.401129032258065e-05,
+      "loss": 0.151,
+      "step": 4214
+    },
+    {
+      "epoch": 0.06744,
+      "grad_norm": 1.4296875,
+      "learning_rate": 9.400967741935485e-05,
+      "loss": 0.2013,
+      "step": 4215
+    },
+    {
+      "epoch": 0.067456,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.400806451612904e-05,
+      "loss": 0.2164,
+      "step": 4216
+    },
+    {
+      "epoch": 0.067472,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.400645161290324e-05,
+      "loss": 0.1923,
+      "step": 4217
+    },
+    {
+      "epoch": 0.067488,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.400483870967742e-05,
+      "loss": 0.1984,
+      "step": 4218
+    },
+    {
+      "epoch": 0.067504,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.400322580645162e-05,
+      "loss": 0.1762,
+      "step": 4219
+    },
+    {
+      "epoch": 0.06752,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.400161290322581e-05,
+      "loss": 0.1926,
+      "step": 4220
+    },
+    {
+      "epoch": 0.067536,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.4e-05,
+      "loss": 0.1439,
+      "step": 4221
+    },
+    {
+      "epoch": 0.067552,
+      "grad_norm": 1.453125,
+      "learning_rate": 9.39983870967742e-05,
+      "loss": 0.2253,
+      "step": 4222
+    },
+    {
+      "epoch": 0.067568,
+      "grad_norm": 1.34375,
+      "learning_rate": 9.39967741935484e-05,
+      "loss": 0.1774,
+      "step": 4223
+    },
+    {
+      "epoch": 0.067584,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.399516129032258e-05,
+      "loss": 0.1676,
+      "step": 4224
+    },
+    {
+      "epoch": 0.0676,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.399354838709678e-05,
+      "loss": 0.1971,
+      "step": 4225
+    },
+    {
+      "epoch": 0.067616,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.399193548387098e-05,
+      "loss": 0.2343,
+      "step": 4226
+    },
+    {
+      "epoch": 0.067632,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.399032258064517e-05,
+      "loss": 0.1926,
+      "step": 4227
+    },
+    {
+      "epoch": 0.067648,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.398870967741936e-05,
+      "loss": 0.1937,
+      "step": 4228
+    },
+    {
+      "epoch": 0.067664,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.398709677419355e-05,
+      "loss": 0.1513,
+      "step": 4229
+    },
+    {
+      "epoch": 0.06768,
+      "grad_norm": 0.875,
+      "learning_rate": 9.398548387096775e-05,
+      "loss": 0.1704,
+      "step": 4230
+    },
+    {
+      "epoch": 0.067696,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.398387096774194e-05,
+      "loss": 0.19,
+      "step": 4231
+    },
+    {
+      "epoch": 0.067712,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.398225806451614e-05,
+      "loss": 0.1966,
+      "step": 4232
+    },
+    {
+      "epoch": 0.067728,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.398064516129032e-05,
+      "loss": 0.1754,
+      "step": 4233
+    },
+    {
+      "epoch": 0.067744,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.397903225806452e-05,
+      "loss": 0.1651,
+      "step": 4234
+    },
+    {
+      "epoch": 0.06776,
+      "grad_norm": 1.25,
+      "learning_rate": 9.397741935483871e-05,
+      "loss": 0.1682,
+      "step": 4235
+    },
+    {
+      "epoch": 0.067776,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.397580645161291e-05,
+      "loss": 0.1981,
+      "step": 4236
+    },
+    {
+      "epoch": 0.067792,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.397419354838711e-05,
+      "loss": 0.1399,
+      "step": 4237
+    },
+    {
+      "epoch": 0.067808,
+      "grad_norm": 0.875,
+      "learning_rate": 9.39725806451613e-05,
+      "loss": 0.1939,
+      "step": 4238
+    },
+    {
+      "epoch": 0.067824,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.397096774193549e-05,
+      "loss": 0.1714,
+      "step": 4239
+    },
+    {
+      "epoch": 0.06784,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.396935483870968e-05,
+      "loss": 0.2062,
+      "step": 4240
+    },
+    {
+      "epoch": 0.067856,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.396774193548388e-05,
+      "loss": 0.1985,
+      "step": 4241
+    },
+    {
+      "epoch": 0.067872,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.396612903225806e-05,
+      "loss": 0.1766,
+      "step": 4242
+    },
+    {
+      "epoch": 0.067888,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.396451612903226e-05,
+      "loss": 0.1911,
+      "step": 4243
+    },
+    {
+      "epoch": 0.067904,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.396290322580645e-05,
+      "loss": 0.1815,
+      "step": 4244
+    },
+    {
+      "epoch": 0.06792,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.396129032258065e-05,
+      "loss": 0.1987,
+      "step": 4245
+    },
+    {
+      "epoch": 0.067936,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.395967741935484e-05,
+      "loss": 0.1898,
+      "step": 4246
+    },
+    {
+      "epoch": 0.067952,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.395806451612904e-05,
+      "loss": 0.2021,
+      "step": 4247
+    },
+    {
+      "epoch": 0.067968,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.395645161290324e-05,
+      "loss": 0.206,
+      "step": 4248
+    },
+    {
+      "epoch": 0.067984,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.395483870967744e-05,
+      "loss": 0.1671,
+      "step": 4249
+    },
+    {
+      "epoch": 0.068,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.395322580645162e-05,
+      "loss": 0.211,
+      "step": 4250
+    },
+    {
+      "epoch": 0.068016,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.395161290322582e-05,
+      "loss": 0.2425,
+      "step": 4251
+    },
+    {
+      "epoch": 0.068032,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.395000000000001e-05,
+      "loss": 0.2346,
+      "step": 4252
+    },
+    {
+      "epoch": 0.068048,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.394838709677419e-05,
+      "loss": 0.1733,
+      "step": 4253
+    },
+    {
+      "epoch": 0.068064,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.394677419354839e-05,
+      "loss": 0.1626,
+      "step": 4254
+    },
+    {
+      "epoch": 0.06808,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.394516129032258e-05,
+      "loss": 0.1445,
+      "step": 4255
+    },
+    {
+      "epoch": 0.068096,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.394354838709678e-05,
+      "loss": 0.2092,
+      "step": 4256
+    },
+    {
+      "epoch": 0.068112,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.394193548387096e-05,
+      "loss": 0.1582,
+      "step": 4257
+    },
+    {
+      "epoch": 0.068128,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.394032258064516e-05,
+      "loss": 0.1734,
+      "step": 4258
+    },
+    {
+      "epoch": 0.068144,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.393870967741935e-05,
+      "loss": 0.1612,
+      "step": 4259
+    },
+    {
+      "epoch": 0.06816,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.393709677419355e-05,
+      "loss": 0.164,
+      "step": 4260
+    },
+    {
+      "epoch": 0.068176,
+      "grad_norm": 1.125,
+      "learning_rate": 9.393548387096775e-05,
+      "loss": 0.1962,
+      "step": 4261
+    },
+    {
+      "epoch": 0.068192,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.393387096774195e-05,
+      "loss": 0.1593,
+      "step": 4262
+    },
+    {
+      "epoch": 0.068208,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.393225806451613e-05,
+      "loss": 0.1533,
+      "step": 4263
+    },
+    {
+      "epoch": 0.068224,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.393064516129033e-05,
+      "loss": 0.1683,
+      "step": 4264
+    },
+    {
+      "epoch": 0.06824,
+      "grad_norm": 1.3984375,
+      "learning_rate": 9.392903225806452e-05,
+      "loss": 0.1768,
+      "step": 4265
+    },
+    {
+      "epoch": 0.068256,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.392741935483872e-05,
+      "loss": 0.1576,
+      "step": 4266
+    },
+    {
+      "epoch": 0.068272,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.39258064516129e-05,
+      "loss": 0.2049,
+      "step": 4267
+    },
+    {
+      "epoch": 0.068288,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.392419354838709e-05,
+      "loss": 0.164,
+      "step": 4268
+    },
+    {
+      "epoch": 0.068304,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.392258064516129e-05,
+      "loss": 0.1873,
+      "step": 4269
+    },
+    {
+      "epoch": 0.06832,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.392096774193548e-05,
+      "loss": 0.1822,
+      "step": 4270
+    },
+    {
+      "epoch": 0.068336,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.391935483870968e-05,
+      "loss": 0.1529,
+      "step": 4271
+    },
+    {
+      "epoch": 0.068352,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.391774193548388e-05,
+      "loss": 0.1803,
+      "step": 4272
+    },
+    {
+      "epoch": 0.068368,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.391612903225808e-05,
+      "loss": 0.2703,
+      "step": 4273
+    },
+    {
+      "epoch": 0.068384,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.391451612903226e-05,
+      "loss": 0.1338,
+      "step": 4274
+    },
+    {
+      "epoch": 0.0684,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.391290322580646e-05,
+      "loss": 0.1558,
+      "step": 4275
+    },
+    {
+      "epoch": 0.068416,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.391129032258065e-05,
+      "loss": 0.1996,
+      "step": 4276
+    },
+    {
+      "epoch": 0.068432,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.390967741935485e-05,
+      "loss": 0.2086,
+      "step": 4277
+    },
+    {
+      "epoch": 0.068448,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.390806451612903e-05,
+      "loss": 0.1555,
+      "step": 4278
+    },
+    {
+      "epoch": 0.068464,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.390645161290323e-05,
+      "loss": 0.1524,
+      "step": 4279
+    },
+    {
+      "epoch": 0.06848,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.390483870967742e-05,
+      "loss": 0.1961,
+      "step": 4280
+    },
+    {
+      "epoch": 0.068496,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.390322580645162e-05,
+      "loss": 0.1545,
+      "step": 4281
+    },
+    {
+      "epoch": 0.068512,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.39016129032258e-05,
+      "loss": 0.159,
+      "step": 4282
+    },
+    {
+      "epoch": 0.068528,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.39e-05,
+      "loss": 0.1873,
+      "step": 4283
+    },
+    {
+      "epoch": 0.068544,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.38983870967742e-05,
+      "loss": 0.1698,
+      "step": 4284
+    },
+    {
+      "epoch": 0.06856,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.389677419354839e-05,
+      "loss": 0.1749,
+      "step": 4285
+    },
+    {
+      "epoch": 0.068576,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.389516129032259e-05,
+      "loss": 0.2094,
+      "step": 4286
+    },
+    {
+      "epoch": 0.068592,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.389354838709678e-05,
+      "loss": 0.16,
+      "step": 4287
+    },
+    {
+      "epoch": 0.068608,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.389193548387098e-05,
+      "loss": 0.1802,
+      "step": 4288
+    },
+    {
+      "epoch": 0.068624,
+      "grad_norm": 0.625,
+      "learning_rate": 9.389032258064516e-05,
+      "loss": 0.1427,
+      "step": 4289
+    },
+    {
+      "epoch": 0.06864,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.388870967741936e-05,
+      "loss": 0.1872,
+      "step": 4290
+    },
+    {
+      "epoch": 0.068656,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.388709677419355e-05,
+      "loss": 0.1894,
+      "step": 4291
+    },
+    {
+      "epoch": 0.068672,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.388548387096775e-05,
+      "loss": 0.2049,
+      "step": 4292
+    },
+    {
+      "epoch": 0.068688,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.388387096774193e-05,
+      "loss": 0.1657,
+      "step": 4293
+    },
+    {
+      "epoch": 0.068704,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.388225806451613e-05,
+      "loss": 0.1632,
+      "step": 4294
+    },
+    {
+      "epoch": 0.06872,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.388064516129032e-05,
+      "loss": 0.1935,
+      "step": 4295
+    },
+    {
+      "epoch": 0.068736,
+      "grad_norm": 1.0,
+      "learning_rate": 9.387903225806452e-05,
+      "loss": 0.1591,
+      "step": 4296
+    },
+    {
+      "epoch": 0.068752,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.387741935483872e-05,
+      "loss": 0.1721,
+      "step": 4297
+    },
+    {
+      "epoch": 0.068768,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.387580645161292e-05,
+      "loss": 0.1875,
+      "step": 4298
+    },
+    {
+      "epoch": 0.068784,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.38741935483871e-05,
+      "loss": 0.1667,
+      "step": 4299
+    },
+    {
+      "epoch": 0.0688,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.387258064516129e-05,
+      "loss": 0.1645,
+      "step": 4300
+    },
+    {
+      "epoch": 0.068816,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.387096774193549e-05,
+      "loss": 0.1597,
+      "step": 4301
+    },
+    {
+      "epoch": 0.068832,
+      "grad_norm": 1.5390625,
+      "learning_rate": 9.386935483870968e-05,
+      "loss": 0.1747,
+      "step": 4302
+    },
+    {
+      "epoch": 0.068848,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.386774193548388e-05,
+      "loss": 0.1441,
+      "step": 4303
+    },
+    {
+      "epoch": 0.068864,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.386612903225806e-05,
+      "loss": 0.1817,
+      "step": 4304
+    },
+    {
+      "epoch": 0.06888,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.386451612903226e-05,
+      "loss": 0.1505,
+      "step": 4305
+    },
+    {
+      "epoch": 0.068896,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.386290322580645e-05,
+      "loss": 0.1634,
+      "step": 4306
+    },
+    {
+      "epoch": 0.068912,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.386129032258065e-05,
+      "loss": 0.2073,
+      "step": 4307
+    },
+    {
+      "epoch": 0.068928,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.385967741935485e-05,
+      "loss": 0.2143,
+      "step": 4308
+    },
+    {
+      "epoch": 0.068944,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.385806451612905e-05,
+      "loss": 0.1778,
+      "step": 4309
+    },
+    {
+      "epoch": 0.06896,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.385645161290323e-05,
+      "loss": 0.1904,
+      "step": 4310
+    },
+    {
+      "epoch": 0.068976,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.385483870967743e-05,
+      "loss": 0.1375,
+      "step": 4311
+    },
+    {
+      "epoch": 0.068992,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.385322580645162e-05,
+      "loss": 0.1612,
+      "step": 4312
+    },
+    {
+      "epoch": 0.069008,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.385161290322582e-05,
+      "loss": 0.2173,
+      "step": 4313
+    },
+    {
+      "epoch": 0.069024,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.385e-05,
+      "loss": 0.1821,
+      "step": 4314
+    },
+    {
+      "epoch": 0.06904,
+      "grad_norm": 1.328125,
+      "learning_rate": 9.384838709677419e-05,
+      "loss": 0.2108,
+      "step": 4315
+    },
+    {
+      "epoch": 0.069056,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.384677419354839e-05,
+      "loss": 0.1485,
+      "step": 4316
+    },
+    {
+      "epoch": 0.069072,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.384516129032258e-05,
+      "loss": 0.187,
+      "step": 4317
+    },
+    {
+      "epoch": 0.069088,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.384354838709678e-05,
+      "loss": 0.1694,
+      "step": 4318
+    },
+    {
+      "epoch": 0.069104,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.384193548387096e-05,
+      "loss": 0.1653,
+      "step": 4319
+    },
+    {
+      "epoch": 0.06912,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.384032258064516e-05,
+      "loss": 0.1897,
+      "step": 4320
+    },
+    {
+      "epoch": 0.069136,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.383870967741936e-05,
+      "loss": 0.1778,
+      "step": 4321
+    },
+    {
+      "epoch": 0.069152,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.383709677419356e-05,
+      "loss": 0.1981,
+      "step": 4322
+    },
+    {
+      "epoch": 0.069168,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.383548387096775e-05,
+      "loss": 0.1785,
+      "step": 4323
+    },
+    {
+      "epoch": 0.069184,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.383387096774195e-05,
+      "loss": 0.1642,
+      "step": 4324
+    },
+    {
+      "epoch": 0.0692,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.383225806451613e-05,
+      "loss": 0.1858,
+      "step": 4325
+    },
+    {
+      "epoch": 0.069216,
+      "grad_norm": 1.0,
+      "learning_rate": 9.383064516129033e-05,
+      "loss": 0.1916,
+      "step": 4326
+    },
+    {
+      "epoch": 0.069232,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.382903225806452e-05,
+      "loss": 0.2054,
+      "step": 4327
+    },
+    {
+      "epoch": 0.069248,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.382741935483872e-05,
+      "loss": 0.1601,
+      "step": 4328
+    },
+    {
+      "epoch": 0.069264,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.38258064516129e-05,
+      "loss": 0.1709,
+      "step": 4329
+    },
+    {
+      "epoch": 0.06928,
+      "grad_norm": 1.125,
+      "learning_rate": 9.382419354838709e-05,
+      "loss": 0.1685,
+      "step": 4330
+    },
+    {
+      "epoch": 0.069296,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.382258064516129e-05,
+      "loss": 0.1932,
+      "step": 4331
+    },
+    {
+      "epoch": 0.069312,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.382096774193549e-05,
+      "loss": 0.2069,
+      "step": 4332
+    },
+    {
+      "epoch": 0.069328,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.381935483870969e-05,
+      "loss": 0.1812,
+      "step": 4333
+    },
+    {
+      "epoch": 0.069344,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.381774193548388e-05,
+      "loss": 0.1836,
+      "step": 4334
+    },
+    {
+      "epoch": 0.06936,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.381612903225807e-05,
+      "loss": 0.1766,
+      "step": 4335
+    },
+    {
+      "epoch": 0.069376,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.381451612903226e-05,
+      "loss": 0.1661,
+      "step": 4336
+    },
+    {
+      "epoch": 0.069392,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.381290322580646e-05,
+      "loss": 0.1672,
+      "step": 4337
+    },
+    {
+      "epoch": 0.069408,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.381129032258065e-05,
+      "loss": 0.1441,
+      "step": 4338
+    },
+    {
+      "epoch": 0.069424,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.380967741935485e-05,
+      "loss": 0.2262,
+      "step": 4339
+    },
+    {
+      "epoch": 0.06944,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.380806451612903e-05,
+      "loss": 0.1484,
+      "step": 4340
+    },
+    {
+      "epoch": 0.069456,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.380645161290323e-05,
+      "loss": 0.172,
+      "step": 4341
+    },
+    {
+      "epoch": 0.069472,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.380483870967742e-05,
+      "loss": 0.1939,
+      "step": 4342
+    },
+    {
+      "epoch": 0.069488,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.380322580645162e-05,
+      "loss": 0.2012,
+      "step": 4343
+    },
+    {
+      "epoch": 0.069504,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.380161290322582e-05,
+      "loss": 0.1971,
+      "step": 4344
+    },
+    {
+      "epoch": 0.06952,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.38e-05,
+      "loss": 0.2057,
+      "step": 4345
+    },
+    {
+      "epoch": 0.069536,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.37983870967742e-05,
+      "loss": 0.1869,
+      "step": 4346
+    },
+    {
+      "epoch": 0.069552,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.379677419354839e-05,
+      "loss": 0.1954,
+      "step": 4347
+    },
+    {
+      "epoch": 0.069568,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.379516129032259e-05,
+      "loss": 0.1951,
+      "step": 4348
+    },
+    {
+      "epoch": 0.069584,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.379354838709677e-05,
+      "loss": 0.2255,
+      "step": 4349
+    },
+    {
+      "epoch": 0.0696,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.379193548387097e-05,
+      "loss": 0.2033,
+      "step": 4350
+    },
+    {
+      "epoch": 0.069616,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.379032258064516e-05,
+      "loss": 0.16,
+      "step": 4351
+    },
+    {
+      "epoch": 0.069632,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.378870967741936e-05,
+      "loss": 0.1631,
+      "step": 4352
+    },
+    {
+      "epoch": 0.069648,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.378709677419355e-05,
+      "loss": 0.1322,
+      "step": 4353
+    },
+    {
+      "epoch": 0.069664,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.378548387096775e-05,
+      "loss": 0.1567,
+      "step": 4354
+    },
+    {
+      "epoch": 0.06968,
+      "grad_norm": 1.5078125,
+      "learning_rate": 9.378387096774193e-05,
+      "loss": 0.1737,
+      "step": 4355
+    },
+    {
+      "epoch": 0.069696,
+      "grad_norm": 1.25,
+      "learning_rate": 9.378225806451613e-05,
+      "loss": 0.2103,
+      "step": 4356
+    },
+    {
+      "epoch": 0.069712,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.378064516129033e-05,
+      "loss": 0.2004,
+      "step": 4357
+    },
+    {
+      "epoch": 0.069728,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.377903225806453e-05,
+      "loss": 0.2031,
+      "step": 4358
+    },
+    {
+      "epoch": 0.069744,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.377741935483872e-05,
+      "loss": 0.1902,
+      "step": 4359
+    },
+    {
+      "epoch": 0.06976,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.377580645161292e-05,
+      "loss": 0.1422,
+      "step": 4360
+    },
+    {
+      "epoch": 0.069776,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.37741935483871e-05,
+      "loss": 0.1809,
+      "step": 4361
+    },
+    {
+      "epoch": 0.069792,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.377258064516129e-05,
+      "loss": 0.1969,
+      "step": 4362
+    },
+    {
+      "epoch": 0.069808,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.377096774193549e-05,
+      "loss": 0.1835,
+      "step": 4363
+    },
+    {
+      "epoch": 0.069824,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.376935483870967e-05,
+      "loss": 0.1747,
+      "step": 4364
+    },
+    {
+      "epoch": 0.06984,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.376774193548387e-05,
+      "loss": 0.2205,
+      "step": 4365
+    },
+    {
+      "epoch": 0.069856,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.376612903225806e-05,
+      "loss": 0.2102,
+      "step": 4366
+    },
+    {
+      "epoch": 0.069872,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.376451612903226e-05,
+      "loss": 0.1867,
+      "step": 4367
+    },
+    {
+      "epoch": 0.069888,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.376290322580646e-05,
+      "loss": 0.1398,
+      "step": 4368
+    },
+    {
+      "epoch": 0.069904,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.376129032258066e-05,
+      "loss": 0.1947,
+      "step": 4369
+    },
+    {
+      "epoch": 0.06992,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.375967741935484e-05,
+      "loss": 0.2051,
+      "step": 4370
+    },
+    {
+      "epoch": 0.069936,
+      "grad_norm": 1.40625,
+      "learning_rate": 9.375806451612904e-05,
+      "loss": 0.1917,
+      "step": 4371
+    },
+    {
+      "epoch": 0.069952,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.375645161290323e-05,
+      "loss": 0.2061,
+      "step": 4372
+    },
+    {
+      "epoch": 0.069968,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.375483870967743e-05,
+      "loss": 0.1967,
+      "step": 4373
+    },
+    {
+      "epoch": 0.069984,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.375322580645162e-05,
+      "loss": 0.1616,
+      "step": 4374
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.875,
+      "learning_rate": 9.375161290322582e-05,
+      "loss": 0.1853,
+      "step": 4375
+    },
+    {
+      "epoch": 0.070016,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.375e-05,
+      "loss": 0.1843,
+      "step": 4376
+    },
+    {
+      "epoch": 0.070032,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.374838709677419e-05,
+      "loss": 0.1871,
+      "step": 4377
+    },
+    {
+      "epoch": 0.070048,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.374677419354839e-05,
+      "loss": 0.2112,
+      "step": 4378
+    },
+    {
+      "epoch": 0.070064,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.374516129032259e-05,
+      "loss": 0.1359,
+      "step": 4379
+    },
+    {
+      "epoch": 0.07008,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.374354838709677e-05,
+      "loss": 0.1588,
+      "step": 4380
+    },
+    {
+      "epoch": 0.070096,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.374193548387097e-05,
+      "loss": 0.1765,
+      "step": 4381
+    },
+    {
+      "epoch": 0.070112,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.374032258064517e-05,
+      "loss": 0.1701,
+      "step": 4382
+    },
+    {
+      "epoch": 0.070128,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.373870967741936e-05,
+      "loss": 0.1683,
+      "step": 4383
+    },
+    {
+      "epoch": 0.070144,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.373709677419356e-05,
+      "loss": 0.1785,
+      "step": 4384
+    },
+    {
+      "epoch": 0.07016,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.373548387096774e-05,
+      "loss": 0.13,
+      "step": 4385
+    },
+    {
+      "epoch": 0.070176,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.373387096774194e-05,
+      "loss": 0.1531,
+      "step": 4386
+    },
+    {
+      "epoch": 0.070192,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.373225806451613e-05,
+      "loss": 0.157,
+      "step": 4387
+    },
+    {
+      "epoch": 0.070208,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.373064516129033e-05,
+      "loss": 0.1675,
+      "step": 4388
+    },
+    {
+      "epoch": 0.070224,
+      "grad_norm": 0.875,
+      "learning_rate": 9.372903225806452e-05,
+      "loss": 0.1959,
+      "step": 4389
+    },
+    {
+      "epoch": 0.07024,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.372741935483872e-05,
+      "loss": 0.1781,
+      "step": 4390
+    },
+    {
+      "epoch": 0.070256,
+      "grad_norm": 1.4375,
+      "learning_rate": 9.37258064516129e-05,
+      "loss": 0.1916,
+      "step": 4391
+    },
+    {
+      "epoch": 0.070272,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.37241935483871e-05,
+      "loss": 0.161,
+      "step": 4392
+    },
+    {
+      "epoch": 0.070288,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.37225806451613e-05,
+      "loss": 0.1662,
+      "step": 4393
+    },
+    {
+      "epoch": 0.070304,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.372096774193549e-05,
+      "loss": 0.2058,
+      "step": 4394
+    },
+    {
+      "epoch": 0.07032,
+      "grad_norm": 1.46875,
+      "learning_rate": 9.371935483870969e-05,
+      "loss": 0.1806,
+      "step": 4395
+    },
+    {
+      "epoch": 0.070336,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.371774193548387e-05,
+      "loss": 0.1784,
+      "step": 4396
+    },
+    {
+      "epoch": 0.070352,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.371612903225807e-05,
+      "loss": 0.1853,
+      "step": 4397
+    },
+    {
+      "epoch": 0.070368,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.371451612903226e-05,
+      "loss": 0.1488,
+      "step": 4398
+    },
+    {
+      "epoch": 0.070384,
+      "grad_norm": 0.875,
+      "learning_rate": 9.371290322580646e-05,
+      "loss": 0.2069,
+      "step": 4399
+    },
+    {
+      "epoch": 0.0704,
+      "grad_norm": 1.0,
+      "learning_rate": 9.371129032258064e-05,
+      "loss": 0.2079,
+      "step": 4400
+    },
+    {
+      "epoch": 0.070416,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.370967741935484e-05,
+      "loss": 0.2197,
+      "step": 4401
+    },
+    {
+      "epoch": 0.070432,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.370806451612903e-05,
+      "loss": 0.1127,
+      "step": 4402
+    },
+    {
+      "epoch": 0.070448,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.370645161290323e-05,
+      "loss": 0.207,
+      "step": 4403
+    },
+    {
+      "epoch": 0.070464,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.370483870967743e-05,
+      "loss": 0.1969,
+      "step": 4404
+    },
+    {
+      "epoch": 0.07048,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.370322580645163e-05,
+      "loss": 0.1382,
+      "step": 4405
+    },
+    {
+      "epoch": 0.070496,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.370161290322581e-05,
+      "loss": 0.2109,
+      "step": 4406
+    },
+    {
+      "epoch": 0.070512,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.370000000000001e-05,
+      "loss": 0.1845,
+      "step": 4407
+    },
+    {
+      "epoch": 0.070528,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.36983870967742e-05,
+      "loss": 0.1489,
+      "step": 4408
+    },
+    {
+      "epoch": 0.070544,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.369677419354839e-05,
+      "loss": 0.1734,
+      "step": 4409
+    },
+    {
+      "epoch": 0.07056,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.369516129032259e-05,
+      "loss": 0.2131,
+      "step": 4410
+    },
+    {
+      "epoch": 0.070576,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.369354838709677e-05,
+      "loss": 0.1437,
+      "step": 4411
+    },
+    {
+      "epoch": 0.070592,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.369193548387097e-05,
+      "loss": 0.1638,
+      "step": 4412
+    },
+    {
+      "epoch": 0.070608,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.369032258064516e-05,
+      "loss": 0.1406,
+      "step": 4413
+    },
+    {
+      "epoch": 0.070624,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.368870967741936e-05,
+      "loss": 0.1683,
+      "step": 4414
+    },
+    {
+      "epoch": 0.07064,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.368709677419354e-05,
+      "loss": 0.1743,
+      "step": 4415
+    },
+    {
+      "epoch": 0.070656,
+      "grad_norm": 1.25,
+      "learning_rate": 9.368548387096774e-05,
+      "loss": 0.2132,
+      "step": 4416
+    },
+    {
+      "epoch": 0.070672,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.368387096774194e-05,
+      "loss": 0.1973,
+      "step": 4417
+    },
+    {
+      "epoch": 0.070688,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.368225806451614e-05,
+      "loss": 0.1713,
+      "step": 4418
+    },
+    {
+      "epoch": 0.070704,
+      "grad_norm": 1.0,
+      "learning_rate": 9.368064516129033e-05,
+      "loss": 0.1639,
+      "step": 4419
+    },
+    {
+      "epoch": 0.07072,
+      "grad_norm": 0.75,
+      "learning_rate": 9.367903225806453e-05,
+      "loss": 0.1468,
+      "step": 4420
+    },
+    {
+      "epoch": 0.070736,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.367741935483871e-05,
+      "loss": 0.2049,
+      "step": 4421
+    },
+    {
+      "epoch": 0.070752,
+      "grad_norm": 1.3828125,
+      "learning_rate": 9.367580645161291e-05,
+      "loss": 0.181,
+      "step": 4422
+    },
+    {
+      "epoch": 0.070768,
+      "grad_norm": 1.4375,
+      "learning_rate": 9.36741935483871e-05,
+      "loss": 0.2431,
+      "step": 4423
+    },
+    {
+      "epoch": 0.070784,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.367258064516129e-05,
+      "loss": 0.1789,
+      "step": 4424
+    },
+    {
+      "epoch": 0.0708,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.367096774193549e-05,
+      "loss": 0.1442,
+      "step": 4425
+    },
+    {
+      "epoch": 0.070816,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.366935483870967e-05,
+      "loss": 0.1458,
+      "step": 4426
+    },
+    {
+      "epoch": 0.070832,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.366774193548387e-05,
+      "loss": 0.1741,
+      "step": 4427
+    },
+    {
+      "epoch": 0.070848,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.366612903225807e-05,
+      "loss": 0.1761,
+      "step": 4428
+    },
+    {
+      "epoch": 0.070864,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.366451612903227e-05,
+      "loss": 0.1748,
+      "step": 4429
+    },
+    {
+      "epoch": 0.07088,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.366290322580646e-05,
+      "loss": 0.1597,
+      "step": 4430
+    },
+    {
+      "epoch": 0.070896,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.366129032258066e-05,
+      "loss": 0.1633,
+      "step": 4431
+    },
+    {
+      "epoch": 0.070912,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.365967741935484e-05,
+      "loss": 0.2082,
+      "step": 4432
+    },
+    {
+      "epoch": 0.070928,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.365806451612904e-05,
+      "loss": 0.1778,
+      "step": 4433
+    },
+    {
+      "epoch": 0.070944,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.365645161290323e-05,
+      "loss": 0.1496,
+      "step": 4434
+    },
+    {
+      "epoch": 0.07096,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.365483870967743e-05,
+      "loss": 0.2008,
+      "step": 4435
+    },
+    {
+      "epoch": 0.070976,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.365322580645161e-05,
+      "loss": 0.204,
+      "step": 4436
+    },
+    {
+      "epoch": 0.070992,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.365161290322581e-05,
+      "loss": 0.1489,
+      "step": 4437
+    },
+    {
+      "epoch": 0.071008,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.365e-05,
+      "loss": 0.1492,
+      "step": 4438
+    },
+    {
+      "epoch": 0.071024,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.36483870967742e-05,
+      "loss": 0.1708,
+      "step": 4439
+    },
+    {
+      "epoch": 0.07104,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.36467741935484e-05,
+      "loss": 0.1806,
+      "step": 4440
+    },
+    {
+      "epoch": 0.071056,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.364516129032258e-05,
+      "loss": 0.2073,
+      "step": 4441
+    },
+    {
+      "epoch": 0.071072,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.364354838709678e-05,
+      "loss": 0.1327,
+      "step": 4442
+    },
+    {
+      "epoch": 0.071088,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.364193548387097e-05,
+      "loss": 0.1554,
+      "step": 4443
+    },
+    {
+      "epoch": 0.071104,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.364032258064517e-05,
+      "loss": 0.1753,
+      "step": 4444
+    },
+    {
+      "epoch": 0.07112,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.363870967741936e-05,
+      "loss": 0.2036,
+      "step": 4445
+    },
+    {
+      "epoch": 0.071136,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.363709677419356e-05,
+      "loss": 0.1856,
+      "step": 4446
+    },
+    {
+      "epoch": 0.071152,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.363548387096774e-05,
+      "loss": 0.2038,
+      "step": 4447
+    },
+    {
+      "epoch": 0.071168,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.363387096774194e-05,
+      "loss": 0.233,
+      "step": 4448
+    },
+    {
+      "epoch": 0.071184,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.363225806451613e-05,
+      "loss": 0.1529,
+      "step": 4449
+    },
+    {
+      "epoch": 0.0712,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.363064516129033e-05,
+      "loss": 0.1857,
+      "step": 4450
+    },
+    {
+      "epoch": 0.071216,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.362903225806451e-05,
+      "loss": 0.1768,
+      "step": 4451
+    },
+    {
+      "epoch": 0.071232,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.362741935483871e-05,
+      "loss": 0.1758,
+      "step": 4452
+    },
+    {
+      "epoch": 0.071248,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.362580645161291e-05,
+      "loss": 0.2229,
+      "step": 4453
+    },
+    {
+      "epoch": 0.071264,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.36241935483871e-05,
+      "loss": 0.1625,
+      "step": 4454
+    },
+    {
+      "epoch": 0.07128,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.36225806451613e-05,
+      "loss": 0.2075,
+      "step": 4455
+    },
+    {
+      "epoch": 0.071296,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.362096774193548e-05,
+      "loss": 0.1529,
+      "step": 4456
+    },
+    {
+      "epoch": 0.071312,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.361935483870968e-05,
+      "loss": 0.1679,
+      "step": 4457
+    },
+    {
+      "epoch": 0.071328,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.361774193548387e-05,
+      "loss": 0.1715,
+      "step": 4458
+    },
+    {
+      "epoch": 0.071344,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.361612903225807e-05,
+      "loss": 0.1538,
+      "step": 4459
+    },
+    {
+      "epoch": 0.07136,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.361451612903226e-05,
+      "loss": 0.2023,
+      "step": 4460
+    },
+    {
+      "epoch": 0.071376,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.361290322580646e-05,
+      "loss": 0.1777,
+      "step": 4461
+    },
+    {
+      "epoch": 0.071392,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.361129032258064e-05,
+      "loss": 0.2179,
+      "step": 4462
+    },
+    {
+      "epoch": 0.071408,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.360967741935484e-05,
+      "loss": 0.2347,
+      "step": 4463
+    },
+    {
+      "epoch": 0.071424,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.360806451612904e-05,
+      "loss": 0.1734,
+      "step": 4464
+    },
+    {
+      "epoch": 0.07144,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.360645161290324e-05,
+      "loss": 0.1861,
+      "step": 4465
+    },
+    {
+      "epoch": 0.071456,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.360483870967743e-05,
+      "loss": 0.2177,
+      "step": 4466
+    },
+    {
+      "epoch": 0.071472,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.360322580645163e-05,
+      "loss": 0.2088,
+      "step": 4467
+    },
+    {
+      "epoch": 0.071488,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.360161290322581e-05,
+      "loss": 0.219,
+      "step": 4468
+    },
+    {
+      "epoch": 0.071504,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.360000000000001e-05,
+      "loss": 0.2024,
+      "step": 4469
+    },
+    {
+      "epoch": 0.07152,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.35983870967742e-05,
+      "loss": 0.1763,
+      "step": 4470
+    },
+    {
+      "epoch": 0.071536,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.359677419354838e-05,
+      "loss": 0.1731,
+      "step": 4471
+    },
+    {
+      "epoch": 0.071552,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.359516129032258e-05,
+      "loss": 0.1546,
+      "step": 4472
+    },
+    {
+      "epoch": 0.071568,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.359354838709677e-05,
+      "loss": 0.1563,
+      "step": 4473
+    },
+    {
+      "epoch": 0.071584,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.359193548387097e-05,
+      "loss": 0.2306,
+      "step": 4474
+    },
+    {
+      "epoch": 0.0716,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.359032258064517e-05,
+      "loss": 0.1362,
+      "step": 4475
+    },
+    {
+      "epoch": 0.071616,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.358870967741936e-05,
+      "loss": 0.2109,
+      "step": 4476
+    },
+    {
+      "epoch": 0.071632,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.358709677419355e-05,
+      "loss": 0.1625,
+      "step": 4477
+    },
+    {
+      "epoch": 0.071648,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.358548387096775e-05,
+      "loss": 0.1635,
+      "step": 4478
+    },
+    {
+      "epoch": 0.071664,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.358387096774194e-05,
+      "loss": 0.167,
+      "step": 4479
+    },
+    {
+      "epoch": 0.07168,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.358225806451614e-05,
+      "loss": 0.1758,
+      "step": 4480
+    },
+    {
+      "epoch": 0.071696,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.358064516129033e-05,
+      "loss": 0.1619,
+      "step": 4481
+    },
+    {
+      "epoch": 0.071712,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.357903225806453e-05,
+      "loss": 0.1693,
+      "step": 4482
+    },
+    {
+      "epoch": 0.071728,
+      "grad_norm": 2.0,
+      "learning_rate": 9.357741935483871e-05,
+      "loss": 0.2066,
+      "step": 4483
+    },
+    {
+      "epoch": 0.071744,
+      "grad_norm": 1.5234375,
+      "learning_rate": 9.357580645161291e-05,
+      "loss": 0.1999,
+      "step": 4484
+    },
+    {
+      "epoch": 0.07176,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.35741935483871e-05,
+      "loss": 0.169,
+      "step": 4485
+    },
+    {
+      "epoch": 0.071776,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.357258064516128e-05,
+      "loss": 0.1792,
+      "step": 4486
+    },
+    {
+      "epoch": 0.071792,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.357096774193548e-05,
+      "loss": 0.2401,
+      "step": 4487
+    },
+    {
+      "epoch": 0.071808,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.356935483870968e-05,
+      "loss": 0.2355,
+      "step": 4488
+    },
+    {
+      "epoch": 0.071824,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.356774193548388e-05,
+      "loss": 0.1873,
+      "step": 4489
+    },
+    {
+      "epoch": 0.07184,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.356612903225807e-05,
+      "loss": 0.2339,
+      "step": 4490
+    },
+    {
+      "epoch": 0.071856,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.356451612903227e-05,
+      "loss": 0.2063,
+      "step": 4491
+    },
+    {
+      "epoch": 0.071872,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.356290322580645e-05,
+      "loss": 0.1688,
+      "step": 4492
+    },
+    {
+      "epoch": 0.071888,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.356129032258065e-05,
+      "loss": 0.2028,
+      "step": 4493
+    },
+    {
+      "epoch": 0.071904,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.355967741935484e-05,
+      "loss": 0.176,
+      "step": 4494
+    },
+    {
+      "epoch": 0.07192,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.355806451612904e-05,
+      "loss": 0.1905,
+      "step": 4495
+    },
+    {
+      "epoch": 0.071936,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.355645161290323e-05,
+      "loss": 0.1917,
+      "step": 4496
+    },
+    {
+      "epoch": 0.071952,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.355483870967743e-05,
+      "loss": 0.2038,
+      "step": 4497
+    },
+    {
+      "epoch": 0.071968,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.355322580645161e-05,
+      "loss": 0.2156,
+      "step": 4498
+    },
+    {
+      "epoch": 0.071984,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.355161290322581e-05,
+      "loss": 0.2089,
+      "step": 4499
+    },
+    {
+      "epoch": 0.072,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.355000000000001e-05,
+      "loss": 0.2278,
+      "step": 4500
+    },
+    {
+      "epoch": 0.072016,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.35483870967742e-05,
+      "loss": 0.2314,
+      "step": 4501
+    },
+    {
+      "epoch": 0.072032,
+      "grad_norm": 1.3828125,
+      "learning_rate": 9.35467741935484e-05,
+      "loss": 0.2024,
+      "step": 4502
+    },
+    {
+      "epoch": 0.072048,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.354516129032258e-05,
+      "loss": 0.1784,
+      "step": 4503
+    },
+    {
+      "epoch": 0.072064,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.354354838709678e-05,
+      "loss": 0.1765,
+      "step": 4504
+    },
+    {
+      "epoch": 0.07208,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.354193548387097e-05,
+      "loss": 0.1864,
+      "step": 4505
+    },
+    {
+      "epoch": 0.072096,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.354032258064517e-05,
+      "loss": 0.1884,
+      "step": 4506
+    },
+    {
+      "epoch": 0.072112,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.353870967741935e-05,
+      "loss": 0.1905,
+      "step": 4507
+    },
+    {
+      "epoch": 0.072128,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.353709677419355e-05,
+      "loss": 0.2015,
+      "step": 4508
+    },
+    {
+      "epoch": 0.072144,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.353548387096774e-05,
+      "loss": 0.1735,
+      "step": 4509
+    },
+    {
+      "epoch": 0.07216,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.353387096774194e-05,
+      "loss": 0.2059,
+      "step": 4510
+    },
+    {
+      "epoch": 0.072176,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.353225806451613e-05,
+      "loss": 0.1782,
+      "step": 4511
+    },
+    {
+      "epoch": 0.072192,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.353064516129032e-05,
+      "loss": 0.1763,
+      "step": 4512
+    },
+    {
+      "epoch": 0.072208,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.352903225806452e-05,
+      "loss": 0.2022,
+      "step": 4513
+    },
+    {
+      "epoch": 0.072224,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.352741935483872e-05,
+      "loss": 0.1671,
+      "step": 4514
+    },
+    {
+      "epoch": 0.07224,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.352580645161291e-05,
+      "loss": 0.174,
+      "step": 4515
+    },
+    {
+      "epoch": 0.072256,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.352419354838711e-05,
+      "loss": 0.1933,
+      "step": 4516
+    },
+    {
+      "epoch": 0.072272,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.35225806451613e-05,
+      "loss": 0.1947,
+      "step": 4517
+    },
+    {
+      "epoch": 0.072288,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.352096774193548e-05,
+      "loss": 0.2249,
+      "step": 4518
+    },
+    {
+      "epoch": 0.072304,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.351935483870968e-05,
+      "loss": 0.1998,
+      "step": 4519
+    },
+    {
+      "epoch": 0.07232,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.351774193548387e-05,
+      "loss": 0.1745,
+      "step": 4520
+    },
+    {
+      "epoch": 0.072336,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.351612903225807e-05,
+      "loss": 0.2467,
+      "step": 4521
+    },
+    {
+      "epoch": 0.072352,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.351451612903225e-05,
+      "loss": 0.1533,
+      "step": 4522
+    },
+    {
+      "epoch": 0.072368,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.351290322580645e-05,
+      "loss": 0.181,
+      "step": 4523
+    },
+    {
+      "epoch": 0.072384,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.351129032258065e-05,
+      "loss": 0.1433,
+      "step": 4524
+    },
+    {
+      "epoch": 0.0724,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.350967741935485e-05,
+      "loss": 0.174,
+      "step": 4525
+    },
+    {
+      "epoch": 0.072416,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.350806451612904e-05,
+      "loss": 0.1872,
+      "step": 4526
+    },
+    {
+      "epoch": 0.072432,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.350645161290324e-05,
+      "loss": 0.1842,
+      "step": 4527
+    },
+    {
+      "epoch": 0.072448,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.350483870967742e-05,
+      "loss": 0.1453,
+      "step": 4528
+    },
+    {
+      "epoch": 0.072464,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.350322580645162e-05,
+      "loss": 0.1589,
+      "step": 4529
+    },
+    {
+      "epoch": 0.07248,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.350161290322581e-05,
+      "loss": 0.1702,
+      "step": 4530
+    },
+    {
+      "epoch": 0.072496,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.350000000000001e-05,
+      "loss": 0.2119,
+      "step": 4531
+    },
+    {
+      "epoch": 0.072512,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.34983870967742e-05,
+      "loss": 0.192,
+      "step": 4532
+    },
+    {
+      "epoch": 0.072528,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.349677419354838e-05,
+      "loss": 0.1992,
+      "step": 4533
+    },
+    {
+      "epoch": 0.072544,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.349516129032258e-05,
+      "loss": 0.1851,
+      "step": 4534
+    },
+    {
+      "epoch": 0.07256,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.349354838709678e-05,
+      "loss": 0.1724,
+      "step": 4535
+    },
+    {
+      "epoch": 0.072576,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.349193548387098e-05,
+      "loss": 0.179,
+      "step": 4536
+    },
+    {
+      "epoch": 0.072592,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.349032258064517e-05,
+      "loss": 0.1679,
+      "step": 4537
+    },
+    {
+      "epoch": 0.072608,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.348870967741937e-05,
+      "loss": 0.1947,
+      "step": 4538
+    },
+    {
+      "epoch": 0.072624,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.348709677419355e-05,
+      "loss": 0.1537,
+      "step": 4539
+    },
+    {
+      "epoch": 0.07264,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.348548387096775e-05,
+      "loss": 0.1495,
+      "step": 4540
+    },
+    {
+      "epoch": 0.072656,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.348387096774194e-05,
+      "loss": 0.1879,
+      "step": 4541
+    },
+    {
+      "epoch": 0.072672,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.348225806451614e-05,
+      "loss": 0.1825,
+      "step": 4542
+    },
+    {
+      "epoch": 0.072688,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.348064516129032e-05,
+      "loss": 0.1895,
+      "step": 4543
+    },
+    {
+      "epoch": 0.072704,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.347903225806452e-05,
+      "loss": 0.1515,
+      "step": 4544
+    },
+    {
+      "epoch": 0.07272,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.347741935483871e-05,
+      "loss": 0.204,
+      "step": 4545
+    },
+    {
+      "epoch": 0.072736,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.347580645161291e-05,
+      "loss": 0.1793,
+      "step": 4546
+    },
+    {
+      "epoch": 0.072752,
+      "grad_norm": 0.75,
+      "learning_rate": 9.34741935483871e-05,
+      "loss": 0.2165,
+      "step": 4547
+    },
+    {
+      "epoch": 0.072768,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.34725806451613e-05,
+      "loss": 0.245,
+      "step": 4548
+    },
+    {
+      "epoch": 0.072784,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.34709677419355e-05,
+      "loss": 0.1993,
+      "step": 4549
+    },
+    {
+      "epoch": 0.0728,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.346935483870968e-05,
+      "loss": 0.1493,
+      "step": 4550
+    },
+    {
+      "epoch": 0.072816,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.346774193548388e-05,
+      "loss": 0.1381,
+      "step": 4551
+    },
+    {
+      "epoch": 0.072832,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.346612903225807e-05,
+      "loss": 0.1754,
+      "step": 4552
+    },
+    {
+      "epoch": 0.072848,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.346451612903227e-05,
+      "loss": 0.1488,
+      "step": 4553
+    },
+    {
+      "epoch": 0.072864,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.346290322580645e-05,
+      "loss": 0.2236,
+      "step": 4554
+    },
+    {
+      "epoch": 0.07288,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.346129032258065e-05,
+      "loss": 0.1527,
+      "step": 4555
+    },
+    {
+      "epoch": 0.072896,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.345967741935484e-05,
+      "loss": 0.1602,
+      "step": 4556
+    },
+    {
+      "epoch": 0.072912,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.345806451612904e-05,
+      "loss": 0.1836,
+      "step": 4557
+    },
+    {
+      "epoch": 0.072928,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.345645161290322e-05,
+      "loss": 0.1942,
+      "step": 4558
+    },
+    {
+      "epoch": 0.072944,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.345483870967742e-05,
+      "loss": 0.173,
+      "step": 4559
+    },
+    {
+      "epoch": 0.07296,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.345322580645162e-05,
+      "loss": 0.1606,
+      "step": 4560
+    },
+    {
+      "epoch": 0.072976,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.345161290322582e-05,
+      "loss": 0.2234,
+      "step": 4561
+    },
+    {
+      "epoch": 0.072992,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.345000000000001e-05,
+      "loss": 0.1717,
+      "step": 4562
+    },
+    {
+      "epoch": 0.073008,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.34483870967742e-05,
+      "loss": 0.1693,
+      "step": 4563
+    },
+    {
+      "epoch": 0.073024,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.34467741935484e-05,
+      "loss": 0.2314,
+      "step": 4564
+    },
+    {
+      "epoch": 0.07304,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.344516129032258e-05,
+      "loss": 0.1517,
+      "step": 4565
+    },
+    {
+      "epoch": 0.073056,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.344354838709678e-05,
+      "loss": 0.2096,
+      "step": 4566
+    },
+    {
+      "epoch": 0.073072,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.344193548387097e-05,
+      "loss": 0.189,
+      "step": 4567
+    },
+    {
+      "epoch": 0.073088,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.344032258064517e-05,
+      "loss": 0.1903,
+      "step": 4568
+    },
+    {
+      "epoch": 0.073104,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.343870967741935e-05,
+      "loss": 0.1492,
+      "step": 4569
+    },
+    {
+      "epoch": 0.07312,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.343709677419355e-05,
+      "loss": 0.1787,
+      "step": 4570
+    },
+    {
+      "epoch": 0.073136,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.343548387096774e-05,
+      "loss": 0.199,
+      "step": 4571
+    },
+    {
+      "epoch": 0.073152,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.343387096774194e-05,
+      "loss": 0.144,
+      "step": 4572
+    },
+    {
+      "epoch": 0.073168,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.343225806451614e-05,
+      "loss": 0.176,
+      "step": 4573
+    },
+    {
+      "epoch": 0.073184,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.343064516129034e-05,
+      "loss": 0.1669,
+      "step": 4574
+    },
+    {
+      "epoch": 0.0732,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.342903225806452e-05,
+      "loss": 0.1979,
+      "step": 4575
+    },
+    {
+      "epoch": 0.073216,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.342741935483872e-05,
+      "loss": 0.2013,
+      "step": 4576
+    },
+    {
+      "epoch": 0.073232,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.342580645161291e-05,
+      "loss": 0.1988,
+      "step": 4577
+    },
+    {
+      "epoch": 0.073248,
+      "grad_norm": 1.125,
+      "learning_rate": 9.342419354838711e-05,
+      "loss": 0.1875,
+      "step": 4578
+    },
+    {
+      "epoch": 0.073264,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.34225806451613e-05,
+      "loss": 0.1825,
+      "step": 4579
+    },
+    {
+      "epoch": 0.07328,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.342096774193548e-05,
+      "loss": 0.1871,
+      "step": 4580
+    },
+    {
+      "epoch": 0.073296,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.341935483870968e-05,
+      "loss": 0.2228,
+      "step": 4581
+    },
+    {
+      "epoch": 0.073312,
+      "grad_norm": 1.34375,
+      "learning_rate": 9.341774193548387e-05,
+      "loss": 0.1823,
+      "step": 4582
+    },
+    {
+      "epoch": 0.073328,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.341612903225806e-05,
+      "loss": 0.1632,
+      "step": 4583
+    },
+    {
+      "epoch": 0.073344,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.341451612903226e-05,
+      "loss": 0.188,
+      "step": 4584
+    },
+    {
+      "epoch": 0.07336,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.341290322580646e-05,
+      "loss": 0.2005,
+      "step": 4585
+    },
+    {
+      "epoch": 0.073376,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.341129032258065e-05,
+      "loss": 0.1827,
+      "step": 4586
+    },
+    {
+      "epoch": 0.073392,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.340967741935485e-05,
+      "loss": 0.2071,
+      "step": 4587
+    },
+    {
+      "epoch": 0.073408,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.340806451612904e-05,
+      "loss": 0.1989,
+      "step": 4588
+    },
+    {
+      "epoch": 0.073424,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.340645161290324e-05,
+      "loss": 0.1822,
+      "step": 4589
+    },
+    {
+      "epoch": 0.07344,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.340483870967742e-05,
+      "loss": 0.155,
+      "step": 4590
+    },
+    {
+      "epoch": 0.073456,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.340322580645162e-05,
+      "loss": 0.2017,
+      "step": 4591
+    },
+    {
+      "epoch": 0.073472,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.340161290322581e-05,
+      "loss": 0.1786,
+      "step": 4592
+    },
+    {
+      "epoch": 0.073488,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.340000000000001e-05,
+      "loss": 0.1726,
+      "step": 4593
+    },
+    {
+      "epoch": 0.073504,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.339838709677419e-05,
+      "loss": 0.1993,
+      "step": 4594
+    },
+    {
+      "epoch": 0.07352,
+      "grad_norm": 0.875,
+      "learning_rate": 9.339677419354839e-05,
+      "loss": 0.2094,
+      "step": 4595
+    },
+    {
+      "epoch": 0.073536,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.339516129032259e-05,
+      "loss": 0.1922,
+      "step": 4596
+    },
+    {
+      "epoch": 0.073552,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.339354838709678e-05,
+      "loss": 0.1659,
+      "step": 4597
+    },
+    {
+      "epoch": 0.073568,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.339193548387098e-05,
+      "loss": 0.2211,
+      "step": 4598
+    },
+    {
+      "epoch": 0.073584,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.339032258064516e-05,
+      "loss": 0.1936,
+      "step": 4599
+    },
+    {
+      "epoch": 0.0736,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.338870967741936e-05,
+      "loss": 0.1874,
+      "step": 4600
+    },
+    {
+      "epoch": 0.073616,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.338709677419355e-05,
+      "loss": 0.1904,
+      "step": 4601
+    },
+    {
+      "epoch": 0.073632,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.338548387096775e-05,
+      "loss": 0.1364,
+      "step": 4602
+    },
+    {
+      "epoch": 0.073648,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.338387096774194e-05,
+      "loss": 0.1902,
+      "step": 4603
+    },
+    {
+      "epoch": 0.073664,
+      "grad_norm": 0.75,
+      "learning_rate": 9.338225806451614e-05,
+      "loss": 0.1782,
+      "step": 4604
+    },
+    {
+      "epoch": 0.07368,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.338064516129032e-05,
+      "loss": 0.1856,
+      "step": 4605
+    },
+    {
+      "epoch": 0.073696,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.337903225806452e-05,
+      "loss": 0.1617,
+      "step": 4606
+    },
+    {
+      "epoch": 0.073712,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.337741935483871e-05,
+      "loss": 0.1636,
+      "step": 4607
+    },
+    {
+      "epoch": 0.073728,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.33758064516129e-05,
+      "loss": 0.1817,
+      "step": 4608
+    },
+    {
+      "epoch": 0.073744,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.33741935483871e-05,
+      "loss": 0.1948,
+      "step": 4609
+    },
+    {
+      "epoch": 0.07376,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.337258064516129e-05,
+      "loss": 0.1723,
+      "step": 4610
+    },
+    {
+      "epoch": 0.073776,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.337096774193549e-05,
+      "loss": 0.1741,
+      "step": 4611
+    },
+    {
+      "epoch": 0.073792,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.336935483870968e-05,
+      "loss": 0.1652,
+      "step": 4612
+    },
+    {
+      "epoch": 0.073808,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.336774193548388e-05,
+      "loss": 0.1853,
+      "step": 4613
+    },
+    {
+      "epoch": 0.073824,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.336612903225806e-05,
+      "loss": 0.1697,
+      "step": 4614
+    },
+    {
+      "epoch": 0.07384,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.336451612903226e-05,
+      "loss": 0.1495,
+      "step": 4615
+    },
+    {
+      "epoch": 0.073856,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.336290322580645e-05,
+      "loss": 0.1647,
+      "step": 4616
+    },
+    {
+      "epoch": 0.073872,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.336129032258065e-05,
+      "loss": 0.2278,
+      "step": 4617
+    },
+    {
+      "epoch": 0.073888,
+      "grad_norm": 0.75,
+      "learning_rate": 9.335967741935484e-05,
+      "loss": 0.154,
+      "step": 4618
+    },
+    {
+      "epoch": 0.073904,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.335806451612903e-05,
+      "loss": 0.2128,
+      "step": 4619
+    },
+    {
+      "epoch": 0.07392,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.335645161290323e-05,
+      "loss": 0.1823,
+      "step": 4620
+    },
+    {
+      "epoch": 0.073936,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.335483870967743e-05,
+      "loss": 0.2009,
+      "step": 4621
+    },
+    {
+      "epoch": 0.073952,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.335322580645162e-05,
+      "loss": 0.1963,
+      "step": 4622
+    },
+    {
+      "epoch": 0.073968,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.335161290322582e-05,
+      "loss": 0.1692,
+      "step": 4623
+    },
+    {
+      "epoch": 0.073984,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.335e-05,
+      "loss": 0.2426,
+      "step": 4624
+    },
+    {
+      "epoch": 0.074,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.334838709677419e-05,
+      "loss": 0.1941,
+      "step": 4625
+    },
+    {
+      "epoch": 0.074016,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.334677419354839e-05,
+      "loss": 0.1595,
+      "step": 4626
+    },
+    {
+      "epoch": 0.074032,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.334516129032258e-05,
+      "loss": 0.1581,
+      "step": 4627
+    },
+    {
+      "epoch": 0.074048,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.334354838709678e-05,
+      "loss": 0.1735,
+      "step": 4628
+    },
+    {
+      "epoch": 0.074064,
+      "grad_norm": 1.4921875,
+      "learning_rate": 9.334193548387096e-05,
+      "loss": 0.1462,
+      "step": 4629
+    },
+    {
+      "epoch": 0.07408,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.334032258064516e-05,
+      "loss": 0.1988,
+      "step": 4630
+    },
+    {
+      "epoch": 0.074096,
+      "grad_norm": 1.4453125,
+      "learning_rate": 9.333870967741936e-05,
+      "loss": 0.1879,
+      "step": 4631
+    },
+    {
+      "epoch": 0.074112,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.333709677419355e-05,
+      "loss": 0.2199,
+      "step": 4632
+    },
+    {
+      "epoch": 0.074128,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.333548387096775e-05,
+      "loss": 0.1624,
+      "step": 4633
+    },
+    {
+      "epoch": 0.074144,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.333387096774195e-05,
+      "loss": 0.1633,
+      "step": 4634
+    },
+    {
+      "epoch": 0.07416,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.333225806451613e-05,
+      "loss": 0.2088,
+      "step": 4635
+    },
+    {
+      "epoch": 0.074176,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.333064516129033e-05,
+      "loss": 0.236,
+      "step": 4636
+    },
+    {
+      "epoch": 0.074192,
+      "grad_norm": 0.875,
+      "learning_rate": 9.332903225806452e-05,
+      "loss": 0.1493,
+      "step": 4637
+    },
+    {
+      "epoch": 0.074208,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.332741935483872e-05,
+      "loss": 0.16,
+      "step": 4638
+    },
+    {
+      "epoch": 0.074224,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.33258064516129e-05,
+      "loss": 0.1636,
+      "step": 4639
+    },
+    {
+      "epoch": 0.07424,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.33241935483871e-05,
+      "loss": 0.1596,
+      "step": 4640
+    },
+    {
+      "epoch": 0.074256,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.332258064516129e-05,
+      "loss": 0.2003,
+      "step": 4641
+    },
+    {
+      "epoch": 0.074272,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.332096774193548e-05,
+      "loss": 0.1649,
+      "step": 4642
+    },
+    {
+      "epoch": 0.074288,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.331935483870968e-05,
+      "loss": 0.1689,
+      "step": 4643
+    },
+    {
+      "epoch": 0.074304,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.331774193548388e-05,
+      "loss": 0.2065,
+      "step": 4644
+    },
+    {
+      "epoch": 0.07432,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.331612903225808e-05,
+      "loss": 0.1727,
+      "step": 4645
+    },
+    {
+      "epoch": 0.074336,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.331451612903226e-05,
+      "loss": 0.1815,
+      "step": 4646
+    },
+    {
+      "epoch": 0.074352,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.331290322580646e-05,
+      "loss": 0.1802,
+      "step": 4647
+    },
+    {
+      "epoch": 0.074368,
+      "grad_norm": 0.875,
+      "learning_rate": 9.331129032258065e-05,
+      "loss": 0.1531,
+      "step": 4648
+    },
+    {
+      "epoch": 0.074384,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.330967741935485e-05,
+      "loss": 0.1891,
+      "step": 4649
+    },
+    {
+      "epoch": 0.0744,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.330806451612903e-05,
+      "loss": 0.1562,
+      "step": 4650
+    },
+    {
+      "epoch": 0.074416,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.330645161290323e-05,
+      "loss": 0.1991,
+      "step": 4651
+    },
+    {
+      "epoch": 0.074432,
+      "grad_norm": 1.453125,
+      "learning_rate": 9.330483870967742e-05,
+      "loss": 0.2165,
+      "step": 4652
+    },
+    {
+      "epoch": 0.074448,
+      "grad_norm": 1.0,
+      "learning_rate": 9.330322580645162e-05,
+      "loss": 0.1784,
+      "step": 4653
+    },
+    {
+      "epoch": 0.074464,
+      "grad_norm": 0.75,
+      "learning_rate": 9.33016129032258e-05,
+      "loss": 0.1589,
+      "step": 4654
+    },
+    {
+      "epoch": 0.07448,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.33e-05,
+      "loss": 0.1707,
+      "step": 4655
+    },
+    {
+      "epoch": 0.074496,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.32983870967742e-05,
+      "loss": 0.1543,
+      "step": 4656
+    },
+    {
+      "epoch": 0.074512,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.329677419354839e-05,
+      "loss": 0.1491,
+      "step": 4657
+    },
+    {
+      "epoch": 0.074528,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.329516129032259e-05,
+      "loss": 0.1553,
+      "step": 4658
+    },
+    {
+      "epoch": 0.074544,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.329354838709678e-05,
+      "loss": 0.2013,
+      "step": 4659
+    },
+    {
+      "epoch": 0.07456,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.329193548387098e-05,
+      "loss": 0.2157,
+      "step": 4660
+    },
+    {
+      "epoch": 0.074576,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.329032258064516e-05,
+      "loss": 0.1668,
+      "step": 4661
+    },
+    {
+      "epoch": 0.074592,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.328870967741936e-05,
+      "loss": 0.1732,
+      "step": 4662
+    },
+    {
+      "epoch": 0.074608,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.328709677419355e-05,
+      "loss": 0.1707,
+      "step": 4663
+    },
+    {
+      "epoch": 0.074624,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.328548387096775e-05,
+      "loss": 0.2123,
+      "step": 4664
+    },
+    {
+      "epoch": 0.07464,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.328387096774193e-05,
+      "loss": 0.1599,
+      "step": 4665
+    },
+    {
+      "epoch": 0.074656,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.328225806451613e-05,
+      "loss": 0.1399,
+      "step": 4666
+    },
+    {
+      "epoch": 0.074672,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.328064516129032e-05,
+      "loss": 0.1929,
+      "step": 4667
+    },
+    {
+      "epoch": 0.074688,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.327903225806452e-05,
+      "loss": 0.2044,
+      "step": 4668
+    },
+    {
+      "epoch": 0.074704,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.327741935483872e-05,
+      "loss": 0.1387,
+      "step": 4669
+    },
+    {
+      "epoch": 0.07472,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.327580645161292e-05,
+      "loss": 0.1473,
+      "step": 4670
+    },
+    {
+      "epoch": 0.074736,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.32741935483871e-05,
+      "loss": 0.1751,
+      "step": 4671
+    },
+    {
+      "epoch": 0.074752,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.327258064516129e-05,
+      "loss": 0.1989,
+      "step": 4672
+    },
+    {
+      "epoch": 0.074768,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.327096774193549e-05,
+      "loss": 0.1955,
+      "step": 4673
+    },
+    {
+      "epoch": 0.074784,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.326935483870968e-05,
+      "loss": 0.1862,
+      "step": 4674
+    },
+    {
+      "epoch": 0.0748,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.326774193548388e-05,
+      "loss": 0.1821,
+      "step": 4675
+    },
+    {
+      "epoch": 0.074816,
+      "grad_norm": 0.875,
+      "learning_rate": 9.326612903225806e-05,
+      "loss": 0.1621,
+      "step": 4676
+    },
+    {
+      "epoch": 0.074832,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.326451612903226e-05,
+      "loss": 0.1848,
+      "step": 4677
+    },
+    {
+      "epoch": 0.074848,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.326290322580645e-05,
+      "loss": 0.1467,
+      "step": 4678
+    },
+    {
+      "epoch": 0.074864,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.326129032258065e-05,
+      "loss": 0.1638,
+      "step": 4679
+    },
+    {
+      "epoch": 0.07488,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.325967741935485e-05,
+      "loss": 0.1715,
+      "step": 4680
+    },
+    {
+      "epoch": 0.074896,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.325806451612905e-05,
+      "loss": 0.1966,
+      "step": 4681
+    },
+    {
+      "epoch": 0.074912,
+      "grad_norm": 0.875,
+      "learning_rate": 9.325645161290323e-05,
+      "loss": 0.2078,
+      "step": 4682
+    },
+    {
+      "epoch": 0.074928,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.325483870967743e-05,
+      "loss": 0.1444,
+      "step": 4683
+    },
+    {
+      "epoch": 0.074944,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.325322580645162e-05,
+      "loss": 0.1652,
+      "step": 4684
+    },
+    {
+      "epoch": 0.07496,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.325161290322582e-05,
+      "loss": 0.1685,
+      "step": 4685
+    },
+    {
+      "epoch": 0.074976,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.325e-05,
+      "loss": 0.1957,
+      "step": 4686
+    },
+    {
+      "epoch": 0.074992,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.32483870967742e-05,
+      "loss": 0.1846,
+      "step": 4687
+    },
+    {
+      "epoch": 0.075008,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.324677419354839e-05,
+      "loss": 0.2039,
+      "step": 4688
+    },
+    {
+      "epoch": 0.075024,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.324516129032258e-05,
+      "loss": 0.1726,
+      "step": 4689
+    },
+    {
+      "epoch": 0.07504,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.324354838709677e-05,
+      "loss": 0.1926,
+      "step": 4690
+    },
+    {
+      "epoch": 0.075056,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.324193548387097e-05,
+      "loss": 0.182,
+      "step": 4691
+    },
+    {
+      "epoch": 0.075072,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.324032258064517e-05,
+      "loss": 0.1714,
+      "step": 4692
+    },
+    {
+      "epoch": 0.075088,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.323870967741936e-05,
+      "loss": 0.1954,
+      "step": 4693
+    },
+    {
+      "epoch": 0.075104,
+      "grad_norm": 1.359375,
+      "learning_rate": 9.323709677419356e-05,
+      "loss": 0.1825,
+      "step": 4694
+    },
+    {
+      "epoch": 0.07512,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.323548387096775e-05,
+      "loss": 0.2363,
+      "step": 4695
+    },
+    {
+      "epoch": 0.075136,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.323387096774195e-05,
+      "loss": 0.2315,
+      "step": 4696
+    },
+    {
+      "epoch": 0.075152,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.323225806451613e-05,
+      "loss": 0.2203,
+      "step": 4697
+    },
+    {
+      "epoch": 0.075168,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.323064516129033e-05,
+      "loss": 0.1363,
+      "step": 4698
+    },
+    {
+      "epoch": 0.075184,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.322903225806452e-05,
+      "loss": 0.1862,
+      "step": 4699
+    },
+    {
+      "epoch": 0.0752,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.322741935483872e-05,
+      "loss": 0.1647,
+      "step": 4700
+    },
+    {
+      "epoch": 0.075216,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.32258064516129e-05,
+      "loss": 0.1772,
+      "step": 4701
+    },
+    {
+      "epoch": 0.075232,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.32241935483871e-05,
+      "loss": 0.2046,
+      "step": 4702
+    },
+    {
+      "epoch": 0.075248,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.322258064516129e-05,
+      "loss": 0.1837,
+      "step": 4703
+    },
+    {
+      "epoch": 0.075264,
+      "grad_norm": 0.50390625,
+      "learning_rate": 9.322096774193549e-05,
+      "loss": 0.142,
+      "step": 4704
+    },
+    {
+      "epoch": 0.07528,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.321935483870969e-05,
+      "loss": 0.1593,
+      "step": 4705
+    },
+    {
+      "epoch": 0.075296,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.321774193548387e-05,
+      "loss": 0.1908,
+      "step": 4706
+    },
+    {
+      "epoch": 0.075312,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.321612903225807e-05,
+      "loss": 0.2418,
+      "step": 4707
+    },
+    {
+      "epoch": 0.075328,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.321451612903226e-05,
+      "loss": 0.1985,
+      "step": 4708
+    },
+    {
+      "epoch": 0.075344,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.321290322580646e-05,
+      "loss": 0.156,
+      "step": 4709
+    },
+    {
+      "epoch": 0.07536,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.321129032258065e-05,
+      "loss": 0.2155,
+      "step": 4710
+    },
+    {
+      "epoch": 0.075376,
+      "grad_norm": 1.25,
+      "learning_rate": 9.320967741935485e-05,
+      "loss": 0.1856,
+      "step": 4711
+    },
+    {
+      "epoch": 0.075392,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.320806451612903e-05,
+      "loss": 0.1527,
+      "step": 4712
+    },
+    {
+      "epoch": 0.075408,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.320645161290323e-05,
+      "loss": 0.1764,
+      "step": 4713
+    },
+    {
+      "epoch": 0.075424,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.320483870967742e-05,
+      "loss": 0.1818,
+      "step": 4714
+    },
+    {
+      "epoch": 0.07544,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.320322580645162e-05,
+      "loss": 0.2455,
+      "step": 4715
+    },
+    {
+      "epoch": 0.075456,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.320161290322582e-05,
+      "loss": 0.2275,
+      "step": 4716
+    },
+    {
+      "epoch": 0.075472,
+      "grad_norm": 0.50390625,
+      "learning_rate": 9.320000000000002e-05,
+      "loss": 0.1375,
+      "step": 4717
+    },
+    {
+      "epoch": 0.075488,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.31983870967742e-05,
+      "loss": 0.2169,
+      "step": 4718
+    },
+    {
+      "epoch": 0.075504,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.319677419354839e-05,
+      "loss": 0.1941,
+      "step": 4719
+    },
+    {
+      "epoch": 0.07552,
+      "grad_norm": 1.3125,
+      "learning_rate": 9.319516129032259e-05,
+      "loss": 0.2286,
+      "step": 4720
+    },
+    {
+      "epoch": 0.075536,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.319354838709677e-05,
+      "loss": 0.1974,
+      "step": 4721
+    },
+    {
+      "epoch": 0.075552,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.319193548387097e-05,
+      "loss": 0.184,
+      "step": 4722
+    },
+    {
+      "epoch": 0.075568,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.319032258064516e-05,
+      "loss": 0.1945,
+      "step": 4723
+    },
+    {
+      "epoch": 0.075584,
+      "grad_norm": 0.43359375,
+      "learning_rate": 9.318870967741936e-05,
+      "loss": 0.1473,
+      "step": 4724
+    },
+    {
+      "epoch": 0.0756,
+      "grad_norm": 1.8125,
+      "learning_rate": 9.318709677419354e-05,
+      "loss": 0.2267,
+      "step": 4725
+    },
+    {
+      "epoch": 0.075616,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.318548387096774e-05,
+      "loss": 0.1627,
+      "step": 4726
+    },
+    {
+      "epoch": 0.075632,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.318387096774194e-05,
+      "loss": 0.1557,
+      "step": 4727
+    },
+    {
+      "epoch": 0.075648,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.318225806451613e-05,
+      "loss": 0.1975,
+      "step": 4728
+    },
+    {
+      "epoch": 0.075664,
+      "grad_norm": 1.5703125,
+      "learning_rate": 9.318064516129033e-05,
+      "loss": 0.1878,
+      "step": 4729
+    },
+    {
+      "epoch": 0.07568,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.317903225806453e-05,
+      "loss": 0.2087,
+      "step": 4730
+    },
+    {
+      "epoch": 0.075696,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.317741935483872e-05,
+      "loss": 0.2388,
+      "step": 4731
+    },
+    {
+      "epoch": 0.075712,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.317580645161292e-05,
+      "loss": 0.1859,
+      "step": 4732
+    },
+    {
+      "epoch": 0.075728,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.31741935483871e-05,
+      "loss": 0.1494,
+      "step": 4733
+    },
+    {
+      "epoch": 0.075744,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.317258064516129e-05,
+      "loss": 0.1714,
+      "step": 4734
+    },
+    {
+      "epoch": 0.07576,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.317096774193549e-05,
+      "loss": 0.1579,
+      "step": 4735
+    },
+    {
+      "epoch": 0.075776,
+      "grad_norm": 1.0,
+      "learning_rate": 9.316935483870967e-05,
+      "loss": 0.2376,
+      "step": 4736
+    },
+    {
+      "epoch": 0.075792,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.316774193548387e-05,
+      "loss": 0.1663,
+      "step": 4737
+    },
+    {
+      "epoch": 0.075808,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.316612903225806e-05,
+      "loss": 0.1582,
+      "step": 4738
+    },
+    {
+      "epoch": 0.075824,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.316451612903226e-05,
+      "loss": 0.1658,
+      "step": 4739
+    },
+    {
+      "epoch": 0.07584,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.316290322580646e-05,
+      "loss": 0.1641,
+      "step": 4740
+    },
+    {
+      "epoch": 0.075856,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.316129032258066e-05,
+      "loss": 0.1643,
+      "step": 4741
+    },
+    {
+      "epoch": 0.075872,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.315967741935484e-05,
+      "loss": 0.1467,
+      "step": 4742
+    },
+    {
+      "epoch": 0.075888,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.315806451612904e-05,
+      "loss": 0.1622,
+      "step": 4743
+    },
+    {
+      "epoch": 0.075904,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.315645161290323e-05,
+      "loss": 0.167,
+      "step": 4744
+    },
+    {
+      "epoch": 0.07592,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.315483870967743e-05,
+      "loss": 0.191,
+      "step": 4745
+    },
+    {
+      "epoch": 0.075936,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.315322580645162e-05,
+      "loss": 0.2011,
+      "step": 4746
+    },
+    {
+      "epoch": 0.075952,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.315161290322581e-05,
+      "loss": 0.2075,
+      "step": 4747
+    },
+    {
+      "epoch": 0.075968,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.315e-05,
+      "loss": 0.1612,
+      "step": 4748
+    },
+    {
+      "epoch": 0.075984,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.31483870967742e-05,
+      "loss": 0.2041,
+      "step": 4749
+    },
+    {
+      "epoch": 0.076,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.314677419354839e-05,
+      "loss": 0.2054,
+      "step": 4750
+    },
+    {
+      "epoch": 0.076016,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.314516129032259e-05,
+      "loss": 0.2365,
+      "step": 4751
+    },
+    {
+      "epoch": 0.076032,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.314354838709679e-05,
+      "loss": 0.1401,
+      "step": 4752
+    },
+    {
+      "epoch": 0.076048,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.314193548387097e-05,
+      "loss": 0.2147,
+      "step": 4753
+    },
+    {
+      "epoch": 0.076064,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.314032258064517e-05,
+      "loss": 0.1823,
+      "step": 4754
+    },
+    {
+      "epoch": 0.07608,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.313870967741936e-05,
+      "loss": 0.2285,
+      "step": 4755
+    },
+    {
+      "epoch": 0.076096,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.313709677419356e-05,
+      "loss": 0.1708,
+      "step": 4756
+    },
+    {
+      "epoch": 0.076112,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.313548387096774e-05,
+      "loss": 0.1576,
+      "step": 4757
+    },
+    {
+      "epoch": 0.076128,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.313387096774194e-05,
+      "loss": 0.1992,
+      "step": 4758
+    },
+    {
+      "epoch": 0.076144,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.313225806451613e-05,
+      "loss": 0.1929,
+      "step": 4759
+    },
+    {
+      "epoch": 0.07616,
+      "grad_norm": 0.625,
+      "learning_rate": 9.313064516129033e-05,
+      "loss": 0.1505,
+      "step": 4760
+    },
+    {
+      "epoch": 0.076176,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.312903225806451e-05,
+      "loss": 0.1728,
+      "step": 4761
+    },
+    {
+      "epoch": 0.076192,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.312741935483871e-05,
+      "loss": 0.1763,
+      "step": 4762
+    },
+    {
+      "epoch": 0.076208,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.31258064516129e-05,
+      "loss": 0.1945,
+      "step": 4763
+    },
+    {
+      "epoch": 0.076224,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.31241935483871e-05,
+      "loss": 0.1673,
+      "step": 4764
+    },
+    {
+      "epoch": 0.07624,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.31225806451613e-05,
+      "loss": 0.1842,
+      "step": 4765
+    },
+    {
+      "epoch": 0.076256,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.312096774193549e-05,
+      "loss": 0.1824,
+      "step": 4766
+    },
+    {
+      "epoch": 0.076272,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.311935483870969e-05,
+      "loss": 0.1722,
+      "step": 4767
+    },
+    {
+      "epoch": 0.076288,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.311774193548387e-05,
+      "loss": 0.1608,
+      "step": 4768
+    },
+    {
+      "epoch": 0.076304,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.311612903225807e-05,
+      "loss": 0.162,
+      "step": 4769
+    },
+    {
+      "epoch": 0.07632,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.311451612903226e-05,
+      "loss": 0.1653,
+      "step": 4770
+    },
+    {
+      "epoch": 0.076336,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.311290322580646e-05,
+      "loss": 0.2163,
+      "step": 4771
+    },
+    {
+      "epoch": 0.076352,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.311129032258064e-05,
+      "loss": 0.1677,
+      "step": 4772
+    },
+    {
+      "epoch": 0.076368,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.310967741935484e-05,
+      "loss": 0.2009,
+      "step": 4773
+    },
+    {
+      "epoch": 0.076384,
+      "grad_norm": 0.625,
+      "learning_rate": 9.310806451612903e-05,
+      "loss": 0.1761,
+      "step": 4774
+    },
+    {
+      "epoch": 0.0764,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.310645161290323e-05,
+      "loss": 0.1478,
+      "step": 4775
+    },
+    {
+      "epoch": 0.076416,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.310483870967743e-05,
+      "loss": 0.1888,
+      "step": 4776
+    },
+    {
+      "epoch": 0.076432,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.310322580645163e-05,
+      "loss": 0.1774,
+      "step": 4777
+    },
+    {
+      "epoch": 0.076448,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.310161290322581e-05,
+      "loss": 0.1959,
+      "step": 4778
+    },
+    {
+      "epoch": 0.076464,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.310000000000001e-05,
+      "loss": 0.1645,
+      "step": 4779
+    },
+    {
+      "epoch": 0.07648,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.30983870967742e-05,
+      "loss": 0.1392,
+      "step": 4780
+    },
+    {
+      "epoch": 0.076496,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.309677419354839e-05,
+      "loss": 0.1191,
+      "step": 4781
+    },
+    {
+      "epoch": 0.076512,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.309516129032259e-05,
+      "loss": 0.2145,
+      "step": 4782
+    },
+    {
+      "epoch": 0.076528,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.309354838709677e-05,
+      "loss": 0.1779,
+      "step": 4783
+    },
+    {
+      "epoch": 0.076544,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.309193548387097e-05,
+      "loss": 0.182,
+      "step": 4784
+    },
+    {
+      "epoch": 0.07656,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.309032258064516e-05,
+      "loss": 0.1931,
+      "step": 4785
+    },
+    {
+      "epoch": 0.076576,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.308870967741936e-05,
+      "loss": 0.2001,
+      "step": 4786
+    },
+    {
+      "epoch": 0.076592,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.308709677419356e-05,
+      "loss": 0.1654,
+      "step": 4787
+    },
+    {
+      "epoch": 0.076608,
+      "grad_norm": 0.478515625,
+      "learning_rate": 9.308548387096776e-05,
+      "loss": 0.1213,
+      "step": 4788
+    },
+    {
+      "epoch": 0.076624,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.308387096774194e-05,
+      "loss": 0.188,
+      "step": 4789
+    },
+    {
+      "epoch": 0.07664,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.308225806451614e-05,
+      "loss": 0.2099,
+      "step": 4790
+    },
+    {
+      "epoch": 0.076656,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.308064516129033e-05,
+      "loss": 0.1688,
+      "step": 4791
+    },
+    {
+      "epoch": 0.076672,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.307903225806453e-05,
+      "loss": 0.184,
+      "step": 4792
+    },
+    {
+      "epoch": 0.076688,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.307741935483871e-05,
+      "loss": 0.1525,
+      "step": 4793
+    },
+    {
+      "epoch": 0.076704,
+      "grad_norm": 1.46875,
+      "learning_rate": 9.307580645161291e-05,
+      "loss": 0.1694,
+      "step": 4794
+    },
+    {
+      "epoch": 0.07672,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.30741935483871e-05,
+      "loss": 0.1758,
+      "step": 4795
+    },
+    {
+      "epoch": 0.076736,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.30725806451613e-05,
+      "loss": 0.1636,
+      "step": 4796
+    },
+    {
+      "epoch": 0.076752,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.307096774193548e-05,
+      "loss": 0.1837,
+      "step": 4797
+    },
+    {
+      "epoch": 0.076768,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.306935483870967e-05,
+      "loss": 0.1653,
+      "step": 4798
+    },
+    {
+      "epoch": 0.076784,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.306774193548387e-05,
+      "loss": 0.1795,
+      "step": 4799
+    },
+    {
+      "epoch": 0.0768,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.306612903225807e-05,
+      "loss": 0.1605,
+      "step": 4800
+    },
+    {
+      "epoch": 0.076816,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.306451612903227e-05,
+      "loss": 0.1923,
+      "step": 4801
+    },
+    {
+      "epoch": 0.076832,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.306290322580646e-05,
+      "loss": 0.1753,
+      "step": 4802
+    },
+    {
+      "epoch": 0.076848,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.306129032258066e-05,
+      "loss": 0.2425,
+      "step": 4803
+    },
+    {
+      "epoch": 0.076864,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.305967741935484e-05,
+      "loss": 0.1841,
+      "step": 4804
+    },
+    {
+      "epoch": 0.07688,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.305806451612904e-05,
+      "loss": 0.1518,
+      "step": 4805
+    },
+    {
+      "epoch": 0.076896,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.305645161290323e-05,
+      "loss": 0.1893,
+      "step": 4806
+    },
+    {
+      "epoch": 0.076912,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.305483870967743e-05,
+      "loss": 0.1375,
+      "step": 4807
+    },
+    {
+      "epoch": 0.076928,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.305322580645161e-05,
+      "loss": 0.1702,
+      "step": 4808
+    },
+    {
+      "epoch": 0.076944,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.305161290322581e-05,
+      "loss": 0.2316,
+      "step": 4809
+    },
+    {
+      "epoch": 0.07696,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.305e-05,
+      "loss": 0.1934,
+      "step": 4810
+    },
+    {
+      "epoch": 0.076976,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.30483870967742e-05,
+      "loss": 0.2079,
+      "step": 4811
+    },
+    {
+      "epoch": 0.076992,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.30467741935484e-05,
+      "loss": 0.1762,
+      "step": 4812
+    },
+    {
+      "epoch": 0.077008,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.304516129032258e-05,
+      "loss": 0.1849,
+      "step": 4813
+    },
+    {
+      "epoch": 0.077024,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.304354838709678e-05,
+      "loss": 0.1728,
+      "step": 4814
+    },
+    {
+      "epoch": 0.07704,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.304193548387097e-05,
+      "loss": 0.1896,
+      "step": 4815
+    },
+    {
+      "epoch": 0.077056,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.304032258064517e-05,
+      "loss": 0.1732,
+      "step": 4816
+    },
+    {
+      "epoch": 0.077072,
+      "grad_norm": 0.75,
+      "learning_rate": 9.303870967741936e-05,
+      "loss": 0.2081,
+      "step": 4817
+    },
+    {
+      "epoch": 0.077088,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.303709677419355e-05,
+      "loss": 0.2132,
+      "step": 4818
+    },
+    {
+      "epoch": 0.077104,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.303548387096774e-05,
+      "loss": 0.1929,
+      "step": 4819
+    },
+    {
+      "epoch": 0.07712,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.303387096774194e-05,
+      "loss": 0.2153,
+      "step": 4820
+    },
+    {
+      "epoch": 0.077136,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.303225806451613e-05,
+      "loss": 0.1912,
+      "step": 4821
+    },
+    {
+      "epoch": 0.077152,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.303064516129033e-05,
+      "loss": 0.1847,
+      "step": 4822
+    },
+    {
+      "epoch": 0.077168,
+      "grad_norm": 0.875,
+      "learning_rate": 9.302903225806451e-05,
+      "loss": 0.1624,
+      "step": 4823
+    },
+    {
+      "epoch": 0.077184,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.302741935483871e-05,
+      "loss": 0.1798,
+      "step": 4824
+    },
+    {
+      "epoch": 0.0772,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.302580645161291e-05,
+      "loss": 0.1658,
+      "step": 4825
+    },
+    {
+      "epoch": 0.077216,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.302419354838711e-05,
+      "loss": 0.207,
+      "step": 4826
+    },
+    {
+      "epoch": 0.077232,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.30225806451613e-05,
+      "loss": 0.1968,
+      "step": 4827
+    },
+    {
+      "epoch": 0.077248,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.302096774193548e-05,
+      "loss": 0.2071,
+      "step": 4828
+    },
+    {
+      "epoch": 0.077264,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.301935483870968e-05,
+      "loss": 0.241,
+      "step": 4829
+    },
+    {
+      "epoch": 0.07728,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.301774193548387e-05,
+      "loss": 0.1746,
+      "step": 4830
+    },
+    {
+      "epoch": 0.077296,
+      "grad_norm": 0.5,
+      "learning_rate": 9.301612903225807e-05,
+      "loss": 0.1737,
+      "step": 4831
+    },
+    {
+      "epoch": 0.077312,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.301451612903225e-05,
+      "loss": 0.188,
+      "step": 4832
+    },
+    {
+      "epoch": 0.077328,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.301290322580645e-05,
+      "loss": 0.1765,
+      "step": 4833
+    },
+    {
+      "epoch": 0.077344,
+      "grad_norm": 1.3828125,
+      "learning_rate": 9.301129032258064e-05,
+      "loss": 0.2012,
+      "step": 4834
+    },
+    {
+      "epoch": 0.07736,
+      "grad_norm": 0.49609375,
+      "learning_rate": 9.300967741935484e-05,
+      "loss": 0.139,
+      "step": 4835
+    },
+    {
+      "epoch": 0.077376,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.300806451612904e-05,
+      "loss": 0.1793,
+      "step": 4836
+    },
+    {
+      "epoch": 0.077392,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.300645161290324e-05,
+      "loss": 0.1951,
+      "step": 4837
+    },
+    {
+      "epoch": 0.077408,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.300483870967743e-05,
+      "loss": 0.1568,
+      "step": 4838
+    },
+    {
+      "epoch": 0.077424,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.300322580645163e-05,
+      "loss": 0.1547,
+      "step": 4839
+    },
+    {
+      "epoch": 0.07744,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.300161290322581e-05,
+      "loss": 0.2235,
+      "step": 4840
+    },
+    {
+      "epoch": 0.077456,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.300000000000001e-05,
+      "loss": 0.1751,
+      "step": 4841
+    },
+    {
+      "epoch": 0.077472,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.29983870967742e-05,
+      "loss": 0.1322,
+      "step": 4842
+    },
+    {
+      "epoch": 0.077488,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.299677419354838e-05,
+      "loss": 0.1538,
+      "step": 4843
+    },
+    {
+      "epoch": 0.077504,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.299516129032258e-05,
+      "loss": 0.199,
+      "step": 4844
+    },
+    {
+      "epoch": 0.07752,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.299354838709677e-05,
+      "loss": 0.1712,
+      "step": 4845
+    },
+    {
+      "epoch": 0.077536,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.299193548387097e-05,
+      "loss": 0.1323,
+      "step": 4846
+    },
+    {
+      "epoch": 0.077552,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.299032258064517e-05,
+      "loss": 0.1667,
+      "step": 4847
+    },
+    {
+      "epoch": 0.077568,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.298870967741937e-05,
+      "loss": 0.1425,
+      "step": 4848
+    },
+    {
+      "epoch": 0.077584,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.298709677419355e-05,
+      "loss": 0.1943,
+      "step": 4849
+    },
+    {
+      "epoch": 0.0776,
+      "grad_norm": 1.125,
+      "learning_rate": 9.298548387096775e-05,
+      "loss": 0.2045,
+      "step": 4850
+    },
+    {
+      "epoch": 0.077616,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.298387096774194e-05,
+      "loss": 0.1493,
+      "step": 4851
+    },
+    {
+      "epoch": 0.077632,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.298225806451614e-05,
+      "loss": 0.2624,
+      "step": 4852
+    },
+    {
+      "epoch": 0.077648,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.298064516129033e-05,
+      "loss": 0.1555,
+      "step": 4853
+    },
+    {
+      "epoch": 0.077664,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.297903225806452e-05,
+      "loss": 0.188,
+      "step": 4854
+    },
+    {
+      "epoch": 0.07768,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.297741935483871e-05,
+      "loss": 0.1675,
+      "step": 4855
+    },
+    {
+      "epoch": 0.077696,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.297580645161291e-05,
+      "loss": 0.1724,
+      "step": 4856
+    },
+    {
+      "epoch": 0.077712,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.29741935483871e-05,
+      "loss": 0.1933,
+      "step": 4857
+    },
+    {
+      "epoch": 0.077728,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.29725806451613e-05,
+      "loss": 0.2189,
+      "step": 4858
+    },
+    {
+      "epoch": 0.077744,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.297096774193548e-05,
+      "loss": 0.1963,
+      "step": 4859
+    },
+    {
+      "epoch": 0.07776,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.296935483870968e-05,
+      "loss": 0.1739,
+      "step": 4860
+    },
+    {
+      "epoch": 0.077776,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.296774193548388e-05,
+      "loss": 0.1595,
+      "step": 4861
+    },
+    {
+      "epoch": 0.077792,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.296612903225807e-05,
+      "loss": 0.1733,
+      "step": 4862
+    },
+    {
+      "epoch": 0.077808,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.296451612903227e-05,
+      "loss": 0.1745,
+      "step": 4863
+    },
+    {
+      "epoch": 0.077824,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.296290322580645e-05,
+      "loss": 0.2057,
+      "step": 4864
+    },
+    {
+      "epoch": 0.07784,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.296129032258065e-05,
+      "loss": 0.1833,
+      "step": 4865
+    },
+    {
+      "epoch": 0.077856,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.295967741935484e-05,
+      "loss": 0.1996,
+      "step": 4866
+    },
+    {
+      "epoch": 0.077872,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.295806451612904e-05,
+      "loss": 0.1804,
+      "step": 4867
+    },
+    {
+      "epoch": 0.077888,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.295645161290322e-05,
+      "loss": 0.211,
+      "step": 4868
+    },
+    {
+      "epoch": 0.077904,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.295483870967742e-05,
+      "loss": 0.2011,
+      "step": 4869
+    },
+    {
+      "epoch": 0.07792,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.295322580645161e-05,
+      "loss": 0.2048,
+      "step": 4870
+    },
+    {
+      "epoch": 0.077936,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.295161290322581e-05,
+      "loss": 0.1392,
+      "step": 4871
+    },
+    {
+      "epoch": 0.077952,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.295000000000001e-05,
+      "loss": 0.1878,
+      "step": 4872
+    },
+    {
+      "epoch": 0.077968,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.294838709677421e-05,
+      "loss": 0.1798,
+      "step": 4873
+    },
+    {
+      "epoch": 0.077984,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.29467741935484e-05,
+      "loss": 0.1845,
+      "step": 4874
+    },
+    {
+      "epoch": 0.078,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.294516129032258e-05,
+      "loss": 0.1717,
+      "step": 4875
+    },
+    {
+      "epoch": 0.078016,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.294354838709678e-05,
+      "loss": 0.1833,
+      "step": 4876
+    },
+    {
+      "epoch": 0.078032,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.294193548387097e-05,
+      "loss": 0.2121,
+      "step": 4877
+    },
+    {
+      "epoch": 0.078048,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.294032258064517e-05,
+      "loss": 0.1426,
+      "step": 4878
+    },
+    {
+      "epoch": 0.078064,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.293870967741935e-05,
+      "loss": 0.2493,
+      "step": 4879
+    },
+    {
+      "epoch": 0.07808,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.293709677419355e-05,
+      "loss": 0.1787,
+      "step": 4880
+    },
+    {
+      "epoch": 0.078096,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.293548387096774e-05,
+      "loss": 0.1819,
+      "step": 4881
+    },
+    {
+      "epoch": 0.078112,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.293387096774194e-05,
+      "loss": 0.1891,
+      "step": 4882
+    },
+    {
+      "epoch": 0.078128,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.293225806451614e-05,
+      "loss": 0.1749,
+      "step": 4883
+    },
+    {
+      "epoch": 0.078144,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.293064516129032e-05,
+      "loss": 0.2306,
+      "step": 4884
+    },
+    {
+      "epoch": 0.07816,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.292903225806452e-05,
+      "loss": 0.1385,
+      "step": 4885
+    },
+    {
+      "epoch": 0.078176,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.292741935483872e-05,
+      "loss": 0.19,
+      "step": 4886
+    },
+    {
+      "epoch": 0.078192,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.292580645161291e-05,
+      "loss": 0.2117,
+      "step": 4887
+    },
+    {
+      "epoch": 0.078208,
+      "grad_norm": 0.75,
+      "learning_rate": 9.292419354838711e-05,
+      "loss": 0.1912,
+      "step": 4888
+    },
+    {
+      "epoch": 0.078224,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.29225806451613e-05,
+      "loss": 0.1897,
+      "step": 4889
+    },
+    {
+      "epoch": 0.07824,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.292096774193548e-05,
+      "loss": 0.1784,
+      "step": 4890
+    },
+    {
+      "epoch": 0.078256,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.291935483870968e-05,
+      "loss": 0.1667,
+      "step": 4891
+    },
+    {
+      "epoch": 0.078272,
+      "grad_norm": 0.875,
+      "learning_rate": 9.291774193548387e-05,
+      "loss": 0.1884,
+      "step": 4892
+    },
+    {
+      "epoch": 0.078288,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.291612903225807e-05,
+      "loss": 0.188,
+      "step": 4893
+    },
+    {
+      "epoch": 0.078304,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.291451612903225e-05,
+      "loss": 0.1814,
+      "step": 4894
+    },
+    {
+      "epoch": 0.07832,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.291290322580645e-05,
+      "loss": 0.2097,
+      "step": 4895
+    },
+    {
+      "epoch": 0.078336,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.291129032258065e-05,
+      "loss": 0.1968,
+      "step": 4896
+    },
+    {
+      "epoch": 0.078352,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.290967741935485e-05,
+      "loss": 0.2286,
+      "step": 4897
+    },
+    {
+      "epoch": 0.078368,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.290806451612904e-05,
+      "loss": 0.1875,
+      "step": 4898
+    },
+    {
+      "epoch": 0.078384,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.290645161290324e-05,
+      "loss": 0.1774,
+      "step": 4899
+    },
+    {
+      "epoch": 0.0784,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.290483870967742e-05,
+      "loss": 0.1919,
+      "step": 4900
+    },
+    {
+      "epoch": 0.078416,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.290322580645162e-05,
+      "loss": 0.1701,
+      "step": 4901
+    },
+    {
+      "epoch": 0.078432,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.290161290322581e-05,
+      "loss": 0.1727,
+      "step": 4902
+    },
+    {
+      "epoch": 0.078448,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.290000000000001e-05,
+      "loss": 0.1885,
+      "step": 4903
+    },
+    {
+      "epoch": 0.078464,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.28983870967742e-05,
+      "loss": 0.1952,
+      "step": 4904
+    },
+    {
+      "epoch": 0.07848,
+      "grad_norm": 0.625,
+      "learning_rate": 9.28967741935484e-05,
+      "loss": 0.1575,
+      "step": 4905
+    },
+    {
+      "epoch": 0.078496,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.289516129032258e-05,
+      "loss": 0.1981,
+      "step": 4906
+    },
+    {
+      "epoch": 0.078512,
+      "grad_norm": 1.359375,
+      "learning_rate": 9.289354838709678e-05,
+      "loss": 0.1339,
+      "step": 4907
+    },
+    {
+      "epoch": 0.078528,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.289193548387098e-05,
+      "loss": 0.1931,
+      "step": 4908
+    },
+    {
+      "epoch": 0.078544,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.289032258064517e-05,
+      "loss": 0.2131,
+      "step": 4909
+    },
+    {
+      "epoch": 0.07856,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.288870967741937e-05,
+      "loss": 0.2078,
+      "step": 4910
+    },
+    {
+      "epoch": 0.078576,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.288709677419355e-05,
+      "loss": 0.2043,
+      "step": 4911
+    },
+    {
+      "epoch": 0.078592,
+      "grad_norm": 0.75,
+      "learning_rate": 9.288548387096775e-05,
+      "loss": 0.1988,
+      "step": 4912
+    },
+    {
+      "epoch": 0.078608,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.288387096774194e-05,
+      "loss": 0.1589,
+      "step": 4913
+    },
+    {
+      "epoch": 0.078624,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.288225806451614e-05,
+      "loss": 0.2493,
+      "step": 4914
+    },
+    {
+      "epoch": 0.07864,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.288064516129032e-05,
+      "loss": 0.1792,
+      "step": 4915
+    },
+    {
+      "epoch": 0.078656,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.287903225806452e-05,
+      "loss": 0.1966,
+      "step": 4916
+    },
+    {
+      "epoch": 0.078672,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.287741935483871e-05,
+      "loss": 0.1539,
+      "step": 4917
+    },
+    {
+      "epoch": 0.078688,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.287580645161291e-05,
+      "loss": 0.1508,
+      "step": 4918
+    },
+    {
+      "epoch": 0.078704,
+      "grad_norm": 1.515625,
+      "learning_rate": 9.28741935483871e-05,
+      "loss": 0.2264,
+      "step": 4919
+    },
+    {
+      "epoch": 0.07872,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.28725806451613e-05,
+      "loss": 0.1953,
+      "step": 4920
+    },
+    {
+      "epoch": 0.078736,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.28709677419355e-05,
+      "loss": 0.1358,
+      "step": 4921
+    },
+    {
+      "epoch": 0.078752,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.286935483870968e-05,
+      "loss": 0.1513,
+      "step": 4922
+    },
+    {
+      "epoch": 0.078768,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.286774193548388e-05,
+      "loss": 0.2006,
+      "step": 4923
+    },
+    {
+      "epoch": 0.078784,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.286612903225807e-05,
+      "loss": 0.1726,
+      "step": 4924
+    },
+    {
+      "epoch": 0.0788,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.286451612903226e-05,
+      "loss": 0.1648,
+      "step": 4925
+    },
+    {
+      "epoch": 0.078816,
+      "grad_norm": 1.359375,
+      "learning_rate": 9.286290322580645e-05,
+      "loss": 0.1952,
+      "step": 4926
+    },
+    {
+      "epoch": 0.078832,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.286129032258065e-05,
+      "loss": 0.1905,
+      "step": 4927
+    },
+    {
+      "epoch": 0.078848,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.285967741935484e-05,
+      "loss": 0.1765,
+      "step": 4928
+    },
+    {
+      "epoch": 0.078864,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.285806451612904e-05,
+      "loss": 0.1894,
+      "step": 4929
+    },
+    {
+      "epoch": 0.07888,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.285645161290322e-05,
+      "loss": 0.1938,
+      "step": 4930
+    },
+    {
+      "epoch": 0.078896,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.285483870967742e-05,
+      "loss": 0.16,
+      "step": 4931
+    },
+    {
+      "epoch": 0.078912,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.285322580645162e-05,
+      "loss": 0.1687,
+      "step": 4932
+    },
+    {
+      "epoch": 0.078928,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.285161290322582e-05,
+      "loss": 0.1668,
+      "step": 4933
+    },
+    {
+      "epoch": 0.078944,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.285000000000001e-05,
+      "loss": 0.2074,
+      "step": 4934
+    },
+    {
+      "epoch": 0.07896,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.284838709677421e-05,
+      "loss": 0.1802,
+      "step": 4935
+    },
+    {
+      "epoch": 0.078976,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.284677419354839e-05,
+      "loss": 0.2066,
+      "step": 4936
+    },
+    {
+      "epoch": 0.078992,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.284516129032258e-05,
+      "loss": 0.1769,
+      "step": 4937
+    },
+    {
+      "epoch": 0.079008,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.284354838709678e-05,
+      "loss": 0.1711,
+      "step": 4938
+    },
+    {
+      "epoch": 0.079024,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.284193548387096e-05,
+      "loss": 0.2191,
+      "step": 4939
+    },
+    {
+      "epoch": 0.07904,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.284032258064516e-05,
+      "loss": 0.1704,
+      "step": 4940
+    },
+    {
+      "epoch": 0.079056,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.283870967741935e-05,
+      "loss": 0.1424,
+      "step": 4941
+    },
+    {
+      "epoch": 0.079072,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.283709677419355e-05,
+      "loss": 0.1847,
+      "step": 4942
+    },
+    {
+      "epoch": 0.079088,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.283548387096775e-05,
+      "loss": 0.1457,
+      "step": 4943
+    },
+    {
+      "epoch": 0.079104,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.283387096774195e-05,
+      "loss": 0.1719,
+      "step": 4944
+    },
+    {
+      "epoch": 0.07912,
+      "grad_norm": 1.0,
+      "learning_rate": 9.283225806451614e-05,
+      "loss": 0.138,
+      "step": 4945
+    },
+    {
+      "epoch": 0.079136,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.283064516129033e-05,
+      "loss": 0.207,
+      "step": 4946
+    },
+    {
+      "epoch": 0.079152,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.282903225806452e-05,
+      "loss": 0.2117,
+      "step": 4947
+    },
+    {
+      "epoch": 0.079168,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.282741935483872e-05,
+      "loss": 0.2119,
+      "step": 4948
+    },
+    {
+      "epoch": 0.079184,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.28258064516129e-05,
+      "loss": 0.1962,
+      "step": 4949
+    },
+    {
+      "epoch": 0.0792,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.28241935483871e-05,
+      "loss": 0.2043,
+      "step": 4950
+    },
+    {
+      "epoch": 0.079216,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.282258064516129e-05,
+      "loss": 0.2297,
+      "step": 4951
+    },
+    {
+      "epoch": 0.079232,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.282096774193548e-05,
+      "loss": 0.2015,
+      "step": 4952
+    },
+    {
+      "epoch": 0.079248,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.281935483870968e-05,
+      "loss": 0.1672,
+      "step": 4953
+    },
+    {
+      "epoch": 0.079264,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.281774193548386e-05,
+      "loss": 0.2104,
+      "step": 4954
+    },
+    {
+      "epoch": 0.07928,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.281612903225806e-05,
+      "loss": 0.2081,
+      "step": 4955
+    },
+    {
+      "epoch": 0.079296,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.281451612903226e-05,
+      "loss": 0.1519,
+      "step": 4956
+    },
+    {
+      "epoch": 0.079312,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.281290322580646e-05,
+      "loss": 0.1567,
+      "step": 4957
+    },
+    {
+      "epoch": 0.079328,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.281129032258065e-05,
+      "loss": 0.2049,
+      "step": 4958
+    },
+    {
+      "epoch": 0.079344,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.280967741935485e-05,
+      "loss": 0.1897,
+      "step": 4959
+    },
+    {
+      "epoch": 0.07936,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.280806451612903e-05,
+      "loss": 0.2075,
+      "step": 4960
+    },
+    {
+      "epoch": 0.079376,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.280645161290323e-05,
+      "loss": 0.1362,
+      "step": 4961
+    },
+    {
+      "epoch": 0.079392,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.280483870967742e-05,
+      "loss": 0.1736,
+      "step": 4962
+    },
+    {
+      "epoch": 0.079408,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.280322580645162e-05,
+      "loss": 0.2285,
+      "step": 4963
+    },
+    {
+      "epoch": 0.079424,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.28016129032258e-05,
+      "loss": 0.2093,
+      "step": 4964
+    },
+    {
+      "epoch": 0.07944,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.28e-05,
+      "loss": 0.1236,
+      "step": 4965
+    },
+    {
+      "epoch": 0.079456,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.279838709677419e-05,
+      "loss": 0.1968,
+      "step": 4966
+    },
+    {
+      "epoch": 0.079472,
+      "grad_norm": 1.3984375,
+      "learning_rate": 9.279677419354839e-05,
+      "loss": 0.1729,
+      "step": 4967
+    },
+    {
+      "epoch": 0.079488,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.279516129032259e-05,
+      "loss": 0.2,
+      "step": 4968
+    },
+    {
+      "epoch": 0.079504,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.279354838709678e-05,
+      "loss": 0.1707,
+      "step": 4969
+    },
+    {
+      "epoch": 0.07952,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.279193548387098e-05,
+      "loss": 0.1909,
+      "step": 4970
+    },
+    {
+      "epoch": 0.079536,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.279032258064516e-05,
+      "loss": 0.1818,
+      "step": 4971
+    },
+    {
+      "epoch": 0.079552,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.278870967741936e-05,
+      "loss": 0.1606,
+      "step": 4972
+    },
+    {
+      "epoch": 0.079568,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.278709677419355e-05,
+      "loss": 0.1796,
+      "step": 4973
+    },
+    {
+      "epoch": 0.079584,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.278548387096775e-05,
+      "loss": 0.1816,
+      "step": 4974
+    },
+    {
+      "epoch": 0.0796,
+      "grad_norm": 0.625,
+      "learning_rate": 9.278387096774193e-05,
+      "loss": 0.1862,
+      "step": 4975
+    },
+    {
+      "epoch": 0.079616,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.278225806451613e-05,
+      "loss": 0.2219,
+      "step": 4976
+    },
+    {
+      "epoch": 0.079632,
+      "grad_norm": 0.875,
+      "learning_rate": 9.278064516129032e-05,
+      "loss": 0.1962,
+      "step": 4977
+    },
+    {
+      "epoch": 0.079648,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.277903225806452e-05,
+      "loss": 0.1554,
+      "step": 4978
+    },
+    {
+      "epoch": 0.079664,
+      "grad_norm": 0.875,
+      "learning_rate": 9.277741935483872e-05,
+      "loss": 0.1393,
+      "step": 4979
+    },
+    {
+      "epoch": 0.07968,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.27758064516129e-05,
+      "loss": 0.1709,
+      "step": 4980
+    },
+    {
+      "epoch": 0.079696,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.27741935483871e-05,
+      "loss": 0.1778,
+      "step": 4981
+    },
+    {
+      "epoch": 0.079712,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.27725806451613e-05,
+      "loss": 0.1542,
+      "step": 4982
+    },
+    {
+      "epoch": 0.079728,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.277096774193549e-05,
+      "loss": 0.2006,
+      "step": 4983
+    },
+    {
+      "epoch": 0.079744,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.276935483870968e-05,
+      "loss": 0.1956,
+      "step": 4984
+    },
+    {
+      "epoch": 0.07976,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.276774193548388e-05,
+      "loss": 0.1905,
+      "step": 4985
+    },
+    {
+      "epoch": 0.079776,
+      "grad_norm": 1.4609375,
+      "learning_rate": 9.276612903225806e-05,
+      "loss": 0.1876,
+      "step": 4986
+    },
+    {
+      "epoch": 0.079792,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.276451612903226e-05,
+      "loss": 0.1888,
+      "step": 4987
+    },
+    {
+      "epoch": 0.079808,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.276290322580645e-05,
+      "loss": 0.1782,
+      "step": 4988
+    },
+    {
+      "epoch": 0.079824,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.276129032258065e-05,
+      "loss": 0.1868,
+      "step": 4989
+    },
+    {
+      "epoch": 0.07984,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.275967741935483e-05,
+      "loss": 0.1765,
+      "step": 4990
+    },
+    {
+      "epoch": 0.079856,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.275806451612903e-05,
+      "loss": 0.2643,
+      "step": 4991
+    },
+    {
+      "epoch": 0.079872,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.275645161290323e-05,
+      "loss": 0.1922,
+      "step": 4992
+    },
+    {
+      "epoch": 0.079888,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.275483870967743e-05,
+      "loss": 0.1725,
+      "step": 4993
+    },
+    {
+      "epoch": 0.079904,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.275322580645162e-05,
+      "loss": 0.1395,
+      "step": 4994
+    },
+    {
+      "epoch": 0.07992,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.275161290322582e-05,
+      "loss": 0.1487,
+      "step": 4995
+    },
+    {
+      "epoch": 0.079936,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.275e-05,
+      "loss": 0.1826,
+      "step": 4996
+    },
+    {
+      "epoch": 0.079952,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.27483870967742e-05,
+      "loss": 0.2011,
+      "step": 4997
+    },
+    {
+      "epoch": 0.079968,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.274677419354839e-05,
+      "loss": 0.1586,
+      "step": 4998
+    },
+    {
+      "epoch": 0.079984,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.274516129032258e-05,
+      "loss": 0.2126,
+      "step": 4999
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.274354838709678e-05,
+      "loss": 0.2455,
+      "step": 5000
+    },
+    {
+      "epoch": 0.080016,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.274193548387096e-05,
+      "loss": 0.1953,
+      "step": 5001
+    },
+    {
+      "epoch": 0.080032,
+      "grad_norm": 0.5078125,
+      "learning_rate": 9.274032258064516e-05,
+      "loss": 0.1513,
+      "step": 5002
+    },
+    {
+      "epoch": 0.080048,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.273870967741936e-05,
+      "loss": 0.2245,
+      "step": 5003
+    },
+    {
+      "epoch": 0.080064,
+      "grad_norm": 0.875,
+      "learning_rate": 9.273709677419356e-05,
+      "loss": 0.2207,
+      "step": 5004
+    },
+    {
+      "epoch": 0.08008,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.273548387096775e-05,
+      "loss": 0.2219,
+      "step": 5005
+    },
+    {
+      "epoch": 0.080096,
+      "grad_norm": 0.44921875,
+      "learning_rate": 9.273387096774195e-05,
+      "loss": 0.1571,
+      "step": 5006
+    },
+    {
+      "epoch": 0.080112,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.273225806451613e-05,
+      "loss": 0.232,
+      "step": 5007
+    },
+    {
+      "epoch": 0.080128,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.273064516129033e-05,
+      "loss": 0.1968,
+      "step": 5008
+    },
+    {
+      "epoch": 0.080144,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.272903225806452e-05,
+      "loss": 0.1817,
+      "step": 5009
+    },
+    {
+      "epoch": 0.08016,
+      "grad_norm": 1.3984375,
+      "learning_rate": 9.272741935483872e-05,
+      "loss": 0.1921,
+      "step": 5010
+    },
+    {
+      "epoch": 0.080176,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.27258064516129e-05,
+      "loss": 0.1716,
+      "step": 5011
+    },
+    {
+      "epoch": 0.080192,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.27241935483871e-05,
+      "loss": 0.1993,
+      "step": 5012
+    },
+    {
+      "epoch": 0.080208,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.272258064516129e-05,
+      "loss": 0.1722,
+      "step": 5013
+    },
+    {
+      "epoch": 0.080224,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.272096774193549e-05,
+      "loss": 0.1871,
+      "step": 5014
+    },
+    {
+      "epoch": 0.08024,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.271935483870968e-05,
+      "loss": 0.1637,
+      "step": 5015
+    },
+    {
+      "epoch": 0.080256,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.271774193548388e-05,
+      "loss": 0.185,
+      "step": 5016
+    },
+    {
+      "epoch": 0.080272,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.271612903225808e-05,
+      "loss": 0.1778,
+      "step": 5017
+    },
+    {
+      "epoch": 0.080288,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.271451612903226e-05,
+      "loss": 0.1649,
+      "step": 5018
+    },
+    {
+      "epoch": 0.080304,
+      "grad_norm": 1.3828125,
+      "learning_rate": 9.271290322580646e-05,
+      "loss": 0.2247,
+      "step": 5019
+    },
+    {
+      "epoch": 0.08032,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.271129032258065e-05,
+      "loss": 0.1288,
+      "step": 5020
+    },
+    {
+      "epoch": 0.080336,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.270967741935485e-05,
+      "loss": 0.1733,
+      "step": 5021
+    },
+    {
+      "epoch": 0.080352,
+      "grad_norm": 1.9375,
+      "learning_rate": 9.270806451612903e-05,
+      "loss": 0.1909,
+      "step": 5022
+    },
+    {
+      "epoch": 0.080368,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.270645161290323e-05,
+      "loss": 0.1649,
+      "step": 5023
+    },
+    {
+      "epoch": 0.080384,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.270483870967742e-05,
+      "loss": 0.1947,
+      "step": 5024
+    },
+    {
+      "epoch": 0.0804,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.270322580645162e-05,
+      "loss": 0.1861,
+      "step": 5025
+    },
+    {
+      "epoch": 0.080416,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.27016129032258e-05,
+      "loss": 0.1759,
+      "step": 5026
+    },
+    {
+      "epoch": 0.080432,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.27e-05,
+      "loss": 0.196,
+      "step": 5027
+    },
+    {
+      "epoch": 0.080448,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.26983870967742e-05,
+      "loss": 0.2159,
+      "step": 5028
+    },
+    {
+      "epoch": 0.080464,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.26967741935484e-05,
+      "loss": 0.2078,
+      "step": 5029
+    },
+    {
+      "epoch": 0.08048,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.269516129032259e-05,
+      "loss": 0.1348,
+      "step": 5030
+    },
+    {
+      "epoch": 0.080496,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.269354838709677e-05,
+      "loss": 0.1094,
+      "step": 5031
+    },
+    {
+      "epoch": 0.080512,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.269193548387097e-05,
+      "loss": 0.1803,
+      "step": 5032
+    },
+    {
+      "epoch": 0.080528,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.269032258064516e-05,
+      "loss": 0.143,
+      "step": 5033
+    },
+    {
+      "epoch": 0.080544,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.268870967741936e-05,
+      "loss": 0.1964,
+      "step": 5034
+    },
+    {
+      "epoch": 0.08056,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.268709677419355e-05,
+      "loss": 0.1955,
+      "step": 5035
+    },
+    {
+      "epoch": 0.080576,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.268548387096775e-05,
+      "loss": 0.1497,
+      "step": 5036
+    },
+    {
+      "epoch": 0.080592,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.268387096774193e-05,
+      "loss": 0.1757,
+      "step": 5037
+    },
+    {
+      "epoch": 0.080608,
+      "grad_norm": 1.359375,
+      "learning_rate": 9.268225806451613e-05,
+      "loss": 0.1689,
+      "step": 5038
+    },
+    {
+      "epoch": 0.080624,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.268064516129033e-05,
+      "loss": 0.1802,
+      "step": 5039
+    },
+    {
+      "epoch": 0.08064,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.267903225806453e-05,
+      "loss": 0.1769,
+      "step": 5040
+    },
+    {
+      "epoch": 0.080656,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.267741935483872e-05,
+      "loss": 0.1075,
+      "step": 5041
+    },
+    {
+      "epoch": 0.080672,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.267580645161292e-05,
+      "loss": 0.1791,
+      "step": 5042
+    },
+    {
+      "epoch": 0.080688,
+      "grad_norm": 1.0,
+      "learning_rate": 9.26741935483871e-05,
+      "loss": 0.1831,
+      "step": 5043
+    },
+    {
+      "epoch": 0.080704,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.26725806451613e-05,
+      "loss": 0.1584,
+      "step": 5044
+    },
+    {
+      "epoch": 0.08072,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.267096774193549e-05,
+      "loss": 0.1762,
+      "step": 5045
+    },
+    {
+      "epoch": 0.080736,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.266935483870967e-05,
+      "loss": 0.1859,
+      "step": 5046
+    },
+    {
+      "epoch": 0.080752,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.266774193548387e-05,
+      "loss": 0.2117,
+      "step": 5047
+    },
+    {
+      "epoch": 0.080768,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.266612903225806e-05,
+      "loss": 0.2253,
+      "step": 5048
+    },
+    {
+      "epoch": 0.080784,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.266451612903226e-05,
+      "loss": 0.2188,
+      "step": 5049
+    },
+    {
+      "epoch": 0.0808,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.266290322580645e-05,
+      "loss": 0.1636,
+      "step": 5050
+    },
+    {
+      "epoch": 0.080816,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.266129032258065e-05,
+      "loss": 0.1901,
+      "step": 5051
+    },
+    {
+      "epoch": 0.080832,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.265967741935485e-05,
+      "loss": 0.2141,
+      "step": 5052
+    },
+    {
+      "epoch": 0.080848,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.265806451612904e-05,
+      "loss": 0.2264,
+      "step": 5053
+    },
+    {
+      "epoch": 0.080864,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.265645161290323e-05,
+      "loss": 0.1748,
+      "step": 5054
+    },
+    {
+      "epoch": 0.08088,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.265483870967743e-05,
+      "loss": 0.1571,
+      "step": 5055
+    },
+    {
+      "epoch": 0.080896,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.265322580645162e-05,
+      "loss": 0.1544,
+      "step": 5056
+    },
+    {
+      "epoch": 0.080912,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.265161290322582e-05,
+      "loss": 0.1942,
+      "step": 5057
+    },
+    {
+      "epoch": 0.080928,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.265e-05,
+      "loss": 0.1919,
+      "step": 5058
+    },
+    {
+      "epoch": 0.080944,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.26483870967742e-05,
+      "loss": 0.1612,
+      "step": 5059
+    },
+    {
+      "epoch": 0.08096,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.264677419354839e-05,
+      "loss": 0.1655,
+      "step": 5060
+    },
+    {
+      "epoch": 0.080976,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.264516129032257e-05,
+      "loss": 0.1944,
+      "step": 5061
+    },
+    {
+      "epoch": 0.080992,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.264354838709677e-05,
+      "loss": 0.2176,
+      "step": 5062
+    },
+    {
+      "epoch": 0.081008,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.264193548387097e-05,
+      "loss": 0.2036,
+      "step": 5063
+    },
+    {
+      "epoch": 0.081024,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.264032258064517e-05,
+      "loss": 0.2047,
+      "step": 5064
+    },
+    {
+      "epoch": 0.08104,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.263870967741936e-05,
+      "loss": 0.1847,
+      "step": 5065
+    },
+    {
+      "epoch": 0.081056,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.263709677419356e-05,
+      "loss": 0.1594,
+      "step": 5066
+    },
+    {
+      "epoch": 0.081072,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.263548387096774e-05,
+      "loss": 0.1714,
+      "step": 5067
+    },
+    {
+      "epoch": 0.081088,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.263387096774194e-05,
+      "loss": 0.1673,
+      "step": 5068
+    },
+    {
+      "epoch": 0.081104,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.263225806451613e-05,
+      "loss": 0.1284,
+      "step": 5069
+    },
+    {
+      "epoch": 0.08112,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.263064516129033e-05,
+      "loss": 0.2079,
+      "step": 5070
+    },
+    {
+      "epoch": 0.081136,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.262903225806452e-05,
+      "loss": 0.1759,
+      "step": 5071
+    },
+    {
+      "epoch": 0.081152,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.262741935483872e-05,
+      "loss": 0.1749,
+      "step": 5072
+    },
+    {
+      "epoch": 0.081168,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.26258064516129e-05,
+      "loss": 0.1716,
+      "step": 5073
+    },
+    {
+      "epoch": 0.081184,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.26241935483871e-05,
+      "loss": 0.1935,
+      "step": 5074
+    },
+    {
+      "epoch": 0.0812,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.262258064516129e-05,
+      "loss": 0.154,
+      "step": 5075
+    },
+    {
+      "epoch": 0.081216,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.262096774193549e-05,
+      "loss": 0.2094,
+      "step": 5076
+    },
+    {
+      "epoch": 0.081232,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.261935483870969e-05,
+      "loss": 0.1728,
+      "step": 5077
+    },
+    {
+      "epoch": 0.081248,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.261774193548387e-05,
+      "loss": 0.1869,
+      "step": 5078
+    },
+    {
+      "epoch": 0.081264,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.261612903225807e-05,
+      "loss": 0.1522,
+      "step": 5079
+    },
+    {
+      "epoch": 0.08128,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.261451612903226e-05,
+      "loss": 0.1775,
+      "step": 5080
+    },
+    {
+      "epoch": 0.081296,
+      "grad_norm": 1.0,
+      "learning_rate": 9.261290322580646e-05,
+      "loss": 0.1954,
+      "step": 5081
+    },
+    {
+      "epoch": 0.081312,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.261129032258064e-05,
+      "loss": 0.1502,
+      "step": 5082
+    },
+    {
+      "epoch": 0.081328,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.260967741935484e-05,
+      "loss": 0.2043,
+      "step": 5083
+    },
+    {
+      "epoch": 0.081344,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.260806451612903e-05,
+      "loss": 0.1633,
+      "step": 5084
+    },
+    {
+      "epoch": 0.08136,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.260645161290323e-05,
+      "loss": 0.1429,
+      "step": 5085
+    },
+    {
+      "epoch": 0.081376,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.260483870967742e-05,
+      "loss": 0.1643,
+      "step": 5086
+    },
+    {
+      "epoch": 0.081392,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.260322580645162e-05,
+      "loss": 0.1873,
+      "step": 5087
+    },
+    {
+      "epoch": 0.081408,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.260161290322582e-05,
+      "loss": 0.1639,
+      "step": 5088
+    },
+    {
+      "epoch": 0.081424,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.260000000000001e-05,
+      "loss": 0.1627,
+      "step": 5089
+    },
+    {
+      "epoch": 0.08144,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.25983870967742e-05,
+      "loss": 0.1887,
+      "step": 5090
+    },
+    {
+      "epoch": 0.081456,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.25967741935484e-05,
+      "loss": 0.169,
+      "step": 5091
+    },
+    {
+      "epoch": 0.081472,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.259516129032259e-05,
+      "loss": 0.1732,
+      "step": 5092
+    },
+    {
+      "epoch": 0.081488,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.259354838709677e-05,
+      "loss": 0.1628,
+      "step": 5093
+    },
+    {
+      "epoch": 0.081504,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.259193548387097e-05,
+      "loss": 0.2083,
+      "step": 5094
+    },
+    {
+      "epoch": 0.08152,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.259032258064516e-05,
+      "loss": 0.1685,
+      "step": 5095
+    },
+    {
+      "epoch": 0.081536,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.258870967741936e-05,
+      "loss": 0.1314,
+      "step": 5096
+    },
+    {
+      "epoch": 0.081552,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.258709677419354e-05,
+      "loss": 0.1606,
+      "step": 5097
+    },
+    {
+      "epoch": 0.081568,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.258548387096774e-05,
+      "loss": 0.1551,
+      "step": 5098
+    },
+    {
+      "epoch": 0.081584,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.258387096774194e-05,
+      "loss": 0.1893,
+      "step": 5099
+    },
+    {
+      "epoch": 0.0816,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.258225806451614e-05,
+      "loss": 0.1856,
+      "step": 5100
+    },
+    {
+      "epoch": 0.081616,
+      "grad_norm": 1.421875,
+      "learning_rate": 9.258064516129033e-05,
+      "loss": 0.1644,
+      "step": 5101
+    },
+    {
+      "epoch": 0.081632,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.257903225806453e-05,
+      "loss": 0.1455,
+      "step": 5102
+    },
+    {
+      "epoch": 0.081648,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.257741935483871e-05,
+      "loss": 0.2002,
+      "step": 5103
+    },
+    {
+      "epoch": 0.081664,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.257580645161291e-05,
+      "loss": 0.1433,
+      "step": 5104
+    },
+    {
+      "epoch": 0.08168,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.25741935483871e-05,
+      "loss": 0.1713,
+      "step": 5105
+    },
+    {
+      "epoch": 0.081696,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.25725806451613e-05,
+      "loss": 0.1758,
+      "step": 5106
+    },
+    {
+      "epoch": 0.081712,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.257096774193549e-05,
+      "loss": 0.1704,
+      "step": 5107
+    },
+    {
+      "epoch": 0.081728,
+      "grad_norm": 1.4453125,
+      "learning_rate": 9.256935483870967e-05,
+      "loss": 0.2269,
+      "step": 5108
+    },
+    {
+      "epoch": 0.081744,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.256774193548387e-05,
+      "loss": 0.1791,
+      "step": 5109
+    },
+    {
+      "epoch": 0.08176,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.256612903225806e-05,
+      "loss": 0.1637,
+      "step": 5110
+    },
+    {
+      "epoch": 0.081776,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.256451612903226e-05,
+      "loss": 0.2018,
+      "step": 5111
+    },
+    {
+      "epoch": 0.081792,
+      "grad_norm": 1.125,
+      "learning_rate": 9.256290322580646e-05,
+      "loss": 0.1954,
+      "step": 5112
+    },
+    {
+      "epoch": 0.081808,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.256129032258066e-05,
+      "loss": 0.1583,
+      "step": 5113
+    },
+    {
+      "epoch": 0.081824,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.255967741935484e-05,
+      "loss": 0.1585,
+      "step": 5114
+    },
+    {
+      "epoch": 0.08184,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.255806451612904e-05,
+      "loss": 0.2226,
+      "step": 5115
+    },
+    {
+      "epoch": 0.081856,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.255645161290323e-05,
+      "loss": 0.1869,
+      "step": 5116
+    },
+    {
+      "epoch": 0.081872,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.255483870967743e-05,
+      "loss": 0.1717,
+      "step": 5117
+    },
+    {
+      "epoch": 0.081888,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.255322580645161e-05,
+      "loss": 0.1446,
+      "step": 5118
+    },
+    {
+      "epoch": 0.081904,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.255161290322581e-05,
+      "loss": 0.2291,
+      "step": 5119
+    },
+    {
+      "epoch": 0.08192,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.255e-05,
+      "loss": 0.156,
+      "step": 5120
+    },
+    {
+      "epoch": 0.081936,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.25483870967742e-05,
+      "loss": 0.1885,
+      "step": 5121
+    },
+    {
+      "epoch": 0.081952,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.254677419354839e-05,
+      "loss": 0.1857,
+      "step": 5122
+    },
+    {
+      "epoch": 0.081968,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.254516129032259e-05,
+      "loss": 0.1997,
+      "step": 5123
+    },
+    {
+      "epoch": 0.081984,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.254354838709678e-05,
+      "loss": 0.1757,
+      "step": 5124
+    },
+    {
+      "epoch": 0.082,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.254193548387097e-05,
+      "loss": 0.147,
+      "step": 5125
+    },
+    {
+      "epoch": 0.082016,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.254032258064517e-05,
+      "loss": 0.1713,
+      "step": 5126
+    },
+    {
+      "epoch": 0.082032,
+      "grad_norm": 1.359375,
+      "learning_rate": 9.253870967741936e-05,
+      "loss": 0.174,
+      "step": 5127
+    },
+    {
+      "epoch": 0.082048,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.253709677419356e-05,
+      "loss": 0.1579,
+      "step": 5128
+    },
+    {
+      "epoch": 0.082064,
+      "grad_norm": 0.75,
+      "learning_rate": 9.253548387096774e-05,
+      "loss": 0.1692,
+      "step": 5129
+    },
+    {
+      "epoch": 0.08208,
+      "grad_norm": 1.125,
+      "learning_rate": 9.253387096774194e-05,
+      "loss": 0.1928,
+      "step": 5130
+    },
+    {
+      "epoch": 0.082096,
+      "grad_norm": 1.4921875,
+      "learning_rate": 9.253225806451613e-05,
+      "loss": 0.2348,
+      "step": 5131
+    },
+    {
+      "epoch": 0.082112,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.253064516129033e-05,
+      "loss": 0.1402,
+      "step": 5132
+    },
+    {
+      "epoch": 0.082128,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.252903225806451e-05,
+      "loss": 0.1456,
+      "step": 5133
+    },
+    {
+      "epoch": 0.082144,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.252741935483871e-05,
+      "loss": 0.1737,
+      "step": 5134
+    },
+    {
+      "epoch": 0.08216,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.252580645161291e-05,
+      "loss": 0.1879,
+      "step": 5135
+    },
+    {
+      "epoch": 0.082176,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.25241935483871e-05,
+      "loss": 0.2016,
+      "step": 5136
+    },
+    {
+      "epoch": 0.082192,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.25225806451613e-05,
+      "loss": 0.2086,
+      "step": 5137
+    },
+    {
+      "epoch": 0.082208,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.25209677419355e-05,
+      "loss": 0.1683,
+      "step": 5138
+    },
+    {
+      "epoch": 0.082224,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.251935483870968e-05,
+      "loss": 0.144,
+      "step": 5139
+    },
+    {
+      "epoch": 0.08224,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.251774193548387e-05,
+      "loss": 0.2025,
+      "step": 5140
+    },
+    {
+      "epoch": 0.082256,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.251612903225807e-05,
+      "loss": 0.17,
+      "step": 5141
+    },
+    {
+      "epoch": 0.082272,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.251451612903226e-05,
+      "loss": 0.1548,
+      "step": 5142
+    },
+    {
+      "epoch": 0.082288,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.251290322580646e-05,
+      "loss": 0.189,
+      "step": 5143
+    },
+    {
+      "epoch": 0.082304,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.251129032258064e-05,
+      "loss": 0.1566,
+      "step": 5144
+    },
+    {
+      "epoch": 0.08232,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.250967741935484e-05,
+      "loss": 0.2086,
+      "step": 5145
+    },
+    {
+      "epoch": 0.082336,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.250806451612903e-05,
+      "loss": 0.1846,
+      "step": 5146
+    },
+    {
+      "epoch": 0.082352,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.250645161290323e-05,
+      "loss": 0.183,
+      "step": 5147
+    },
+    {
+      "epoch": 0.082368,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.250483870967743e-05,
+      "loss": 0.1621,
+      "step": 5148
+    },
+    {
+      "epoch": 0.082384,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.250322580645163e-05,
+      "loss": 0.205,
+      "step": 5149
+    },
+    {
+      "epoch": 0.0824,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.250161290322581e-05,
+      "loss": 0.1818,
+      "step": 5150
+    },
+    {
+      "epoch": 0.082416,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.250000000000001e-05,
+      "loss": 0.1515,
+      "step": 5151
+    },
+    {
+      "epoch": 0.082432,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.24983870967742e-05,
+      "loss": 0.1907,
+      "step": 5152
+    },
+    {
+      "epoch": 0.082448,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.24967741935484e-05,
+      "loss": 0.1606,
+      "step": 5153
+    },
+    {
+      "epoch": 0.082464,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.249516129032258e-05,
+      "loss": 0.1464,
+      "step": 5154
+    },
+    {
+      "epoch": 0.08248,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.249354838709677e-05,
+      "loss": 0.1598,
+      "step": 5155
+    },
+    {
+      "epoch": 0.082496,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.249193548387097e-05,
+      "loss": 0.1703,
+      "step": 5156
+    },
+    {
+      "epoch": 0.082512,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.249032258064516e-05,
+      "loss": 0.2034,
+      "step": 5157
+    },
+    {
+      "epoch": 0.082528,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.248870967741936e-05,
+      "loss": 0.1784,
+      "step": 5158
+    },
+    {
+      "epoch": 0.082544,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.248709677419356e-05,
+      "loss": 0.1623,
+      "step": 5159
+    },
+    {
+      "epoch": 0.08256,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.248548387096775e-05,
+      "loss": 0.2208,
+      "step": 5160
+    },
+    {
+      "epoch": 0.082576,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.248387096774194e-05,
+      "loss": 0.137,
+      "step": 5161
+    },
+    {
+      "epoch": 0.082592,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.248225806451614e-05,
+      "loss": 0.1999,
+      "step": 5162
+    },
+    {
+      "epoch": 0.082608,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.248064516129033e-05,
+      "loss": 0.1482,
+      "step": 5163
+    },
+    {
+      "epoch": 0.082624,
+      "grad_norm": 1.3515625,
+      "learning_rate": 9.247903225806453e-05,
+      "loss": 0.1511,
+      "step": 5164
+    },
+    {
+      "epoch": 0.08264,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.247741935483871e-05,
+      "loss": 0.1774,
+      "step": 5165
+    },
+    {
+      "epoch": 0.082656,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.247580645161291e-05,
+      "loss": 0.1954,
+      "step": 5166
+    },
+    {
+      "epoch": 0.082672,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.24741935483871e-05,
+      "loss": 0.2368,
+      "step": 5167
+    },
+    {
+      "epoch": 0.082688,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.24725806451613e-05,
+      "loss": 0.241,
+      "step": 5168
+    },
+    {
+      "epoch": 0.082704,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.247096774193548e-05,
+      "loss": 0.1566,
+      "step": 5169
+    },
+    {
+      "epoch": 0.08272,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.246935483870967e-05,
+      "loss": 0.1735,
+      "step": 5170
+    },
+    {
+      "epoch": 0.082736,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.246774193548387e-05,
+      "loss": 0.1879,
+      "step": 5171
+    },
+    {
+      "epoch": 0.082752,
+      "grad_norm": 0.51171875,
+      "learning_rate": 9.246612903225807e-05,
+      "loss": 0.1613,
+      "step": 5172
+    },
+    {
+      "epoch": 0.082768,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.246451612903227e-05,
+      "loss": 0.2303,
+      "step": 5173
+    },
+    {
+      "epoch": 0.082784,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.246290322580645e-05,
+      "loss": 0.1718,
+      "step": 5174
+    },
+    {
+      "epoch": 0.0828,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.246129032258065e-05,
+      "loss": 0.1904,
+      "step": 5175
+    },
+    {
+      "epoch": 0.082816,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.245967741935484e-05,
+      "loss": 0.1393,
+      "step": 5176
+    },
+    {
+      "epoch": 0.082832,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.245806451612904e-05,
+      "loss": 0.1485,
+      "step": 5177
+    },
+    {
+      "epoch": 0.082848,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.245645161290323e-05,
+      "loss": 0.2354,
+      "step": 5178
+    },
+    {
+      "epoch": 0.082864,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.245483870967743e-05,
+      "loss": 0.16,
+      "step": 5179
+    },
+    {
+      "epoch": 0.08288,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.245322580645161e-05,
+      "loss": 0.1574,
+      "step": 5180
+    },
+    {
+      "epoch": 0.082896,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.245161290322581e-05,
+      "loss": 0.1637,
+      "step": 5181
+    },
+    {
+      "epoch": 0.082912,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.245e-05,
+      "loss": 0.2131,
+      "step": 5182
+    },
+    {
+      "epoch": 0.082928,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.24483870967742e-05,
+      "loss": 0.143,
+      "step": 5183
+    },
+    {
+      "epoch": 0.082944,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.24467741935484e-05,
+      "loss": 0.2292,
+      "step": 5184
+    },
+    {
+      "epoch": 0.08296,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.24451612903226e-05,
+      "loss": 0.1658,
+      "step": 5185
+    },
+    {
+      "epoch": 0.082976,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.244354838709678e-05,
+      "loss": 0.1727,
+      "step": 5186
+    },
+    {
+      "epoch": 0.082992,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.244193548387097e-05,
+      "loss": 0.1554,
+      "step": 5187
+    },
+    {
+      "epoch": 0.083008,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.244032258064517e-05,
+      "loss": 0.1887,
+      "step": 5188
+    },
+    {
+      "epoch": 0.083024,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.243870967741935e-05,
+      "loss": 0.1615,
+      "step": 5189
+    },
+    {
+      "epoch": 0.08304,
+      "grad_norm": 1.3515625,
+      "learning_rate": 9.243709677419355e-05,
+      "loss": 0.1925,
+      "step": 5190
+    },
+    {
+      "epoch": 0.083056,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.243548387096774e-05,
+      "loss": 0.2101,
+      "step": 5191
+    },
+    {
+      "epoch": 0.083072,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.243387096774194e-05,
+      "loss": 0.1823,
+      "step": 5192
+    },
+    {
+      "epoch": 0.083088,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.243225806451613e-05,
+      "loss": 0.1825,
+      "step": 5193
+    },
+    {
+      "epoch": 0.083104,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.243064516129033e-05,
+      "loss": 0.1625,
+      "step": 5194
+    },
+    {
+      "epoch": 0.08312,
+      "grad_norm": 0.498046875,
+      "learning_rate": 9.242903225806452e-05,
+      "loss": 0.1457,
+      "step": 5195
+    },
+    {
+      "epoch": 0.083136,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.242741935483872e-05,
+      "loss": 0.1514,
+      "step": 5196
+    },
+    {
+      "epoch": 0.083152,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.242580645161291e-05,
+      "loss": 0.1981,
+      "step": 5197
+    },
+    {
+      "epoch": 0.083168,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.242419354838711e-05,
+      "loss": 0.1832,
+      "step": 5198
+    },
+    {
+      "epoch": 0.083184,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.24225806451613e-05,
+      "loss": 0.1969,
+      "step": 5199
+    },
+    {
+      "epoch": 0.0832,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.24209677419355e-05,
+      "loss": 0.1681,
+      "step": 5200
+    },
+    {
+      "epoch": 0.083216,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.241935483870968e-05,
+      "loss": 0.1819,
+      "step": 5201
+    },
+    {
+      "epoch": 0.083232,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.241774193548387e-05,
+      "loss": 0.196,
+      "step": 5202
+    },
+    {
+      "epoch": 0.083248,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.241612903225807e-05,
+      "loss": 0.1875,
+      "step": 5203
+    },
+    {
+      "epoch": 0.083264,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.241451612903225e-05,
+      "loss": 0.1317,
+      "step": 5204
+    },
+    {
+      "epoch": 0.08328,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.241290322580645e-05,
+      "loss": 0.1764,
+      "step": 5205
+    },
+    {
+      "epoch": 0.083296,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.241129032258064e-05,
+      "loss": 0.1736,
+      "step": 5206
+    },
+    {
+      "epoch": 0.083312,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.240967741935484e-05,
+      "loss": 0.2217,
+      "step": 5207
+    },
+    {
+      "epoch": 0.083328,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.240806451612904e-05,
+      "loss": 0.1705,
+      "step": 5208
+    },
+    {
+      "epoch": 0.083344,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.240645161290324e-05,
+      "loss": 0.1325,
+      "step": 5209
+    },
+    {
+      "epoch": 0.08336,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.240483870967742e-05,
+      "loss": 0.2074,
+      "step": 5210
+    },
+    {
+      "epoch": 0.083376,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.240322580645162e-05,
+      "loss": 0.186,
+      "step": 5211
+    },
+    {
+      "epoch": 0.083392,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.240161290322581e-05,
+      "loss": 0.1631,
+      "step": 5212
+    },
+    {
+      "epoch": 0.083408,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.240000000000001e-05,
+      "loss": 0.2267,
+      "step": 5213
+    },
+    {
+      "epoch": 0.083424,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.23983870967742e-05,
+      "loss": 0.1954,
+      "step": 5214
+    },
+    {
+      "epoch": 0.08344,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.23967741935484e-05,
+      "loss": 0.223,
+      "step": 5215
+    },
+    {
+      "epoch": 0.083456,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.239516129032258e-05,
+      "loss": 0.2042,
+      "step": 5216
+    },
+    {
+      "epoch": 0.083472,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.239354838709677e-05,
+      "loss": 0.2199,
+      "step": 5217
+    },
+    {
+      "epoch": 0.083488,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.239193548387097e-05,
+      "loss": 0.1632,
+      "step": 5218
+    },
+    {
+      "epoch": 0.083504,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.239032258064517e-05,
+      "loss": 0.1578,
+      "step": 5219
+    },
+    {
+      "epoch": 0.08352,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.238870967741937e-05,
+      "loss": 0.1931,
+      "step": 5220
+    },
+    {
+      "epoch": 0.083536,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.238709677419355e-05,
+      "loss": 0.1557,
+      "step": 5221
+    },
+    {
+      "epoch": 0.083552,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.238548387096775e-05,
+      "loss": 0.2017,
+      "step": 5222
+    },
+    {
+      "epoch": 0.083568,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.238387096774194e-05,
+      "loss": 0.1503,
+      "step": 5223
+    },
+    {
+      "epoch": 0.083584,
+      "grad_norm": 1.328125,
+      "learning_rate": 9.238225806451614e-05,
+      "loss": 0.18,
+      "step": 5224
+    },
+    {
+      "epoch": 0.0836,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.238064516129032e-05,
+      "loss": 0.1843,
+      "step": 5225
+    },
+    {
+      "epoch": 0.083616,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.237903225806452e-05,
+      "loss": 0.1635,
+      "step": 5226
+    },
+    {
+      "epoch": 0.083632,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.237741935483871e-05,
+      "loss": 0.2143,
+      "step": 5227
+    },
+    {
+      "epoch": 0.083648,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.237580645161291e-05,
+      "loss": 0.232,
+      "step": 5228
+    },
+    {
+      "epoch": 0.083664,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.23741935483871e-05,
+      "loss": 0.2089,
+      "step": 5229
+    },
+    {
+      "epoch": 0.08368,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.23725806451613e-05,
+      "loss": 0.2006,
+      "step": 5230
+    },
+    {
+      "epoch": 0.083696,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.23709677419355e-05,
+      "loss": 0.2017,
+      "step": 5231
+    },
+    {
+      "epoch": 0.083712,
+      "grad_norm": 0.625,
+      "learning_rate": 9.236935483870968e-05,
+      "loss": 0.1769,
+      "step": 5232
+    },
+    {
+      "epoch": 0.083728,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.236774193548388e-05,
+      "loss": 0.175,
+      "step": 5233
+    },
+    {
+      "epoch": 0.083744,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.236612903225807e-05,
+      "loss": 0.1714,
+      "step": 5234
+    },
+    {
+      "epoch": 0.08376,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.236451612903227e-05,
+      "loss": 0.2098,
+      "step": 5235
+    },
+    {
+      "epoch": 0.083776,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.236290322580645e-05,
+      "loss": 0.1962,
+      "step": 5236
+    },
+    {
+      "epoch": 0.083792,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.236129032258065e-05,
+      "loss": 0.1613,
+      "step": 5237
+    },
+    {
+      "epoch": 0.083808,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.235967741935484e-05,
+      "loss": 0.1755,
+      "step": 5238
+    },
+    {
+      "epoch": 0.083824,
+      "grad_norm": 1.0,
+      "learning_rate": 9.235806451612904e-05,
+      "loss": 0.2068,
+      "step": 5239
+    },
+    {
+      "epoch": 0.08384,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.235645161290322e-05,
+      "loss": 0.1891,
+      "step": 5240
+    },
+    {
+      "epoch": 0.083856,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.235483870967742e-05,
+      "loss": 0.146,
+      "step": 5241
+    },
+    {
+      "epoch": 0.083872,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.235322580645161e-05,
+      "loss": 0.1697,
+      "step": 5242
+    },
+    {
+      "epoch": 0.083888,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.235161290322581e-05,
+      "loss": 0.1652,
+      "step": 5243
+    },
+    {
+      "epoch": 0.083904,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.235000000000001e-05,
+      "loss": 0.1712,
+      "step": 5244
+    },
+    {
+      "epoch": 0.08392,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.234838709677421e-05,
+      "loss": 0.1611,
+      "step": 5245
+    },
+    {
+      "epoch": 0.083936,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.23467741935484e-05,
+      "loss": 0.152,
+      "step": 5246
+    },
+    {
+      "epoch": 0.083952,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.23451612903226e-05,
+      "loss": 0.1894,
+      "step": 5247
+    },
+    {
+      "epoch": 0.083968,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.234354838709678e-05,
+      "loss": 0.2488,
+      "step": 5248
+    },
+    {
+      "epoch": 0.083984,
+      "grad_norm": 1.40625,
+      "learning_rate": 9.234193548387097e-05,
+      "loss": 0.1895,
+      "step": 5249
+    },
+    {
+      "epoch": 0.084,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.234032258064517e-05,
+      "loss": 0.1543,
+      "step": 5250
+    },
+    {
+      "epoch": 0.084016,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.233870967741935e-05,
+      "loss": 0.2311,
+      "step": 5251
+    },
+    {
+      "epoch": 0.084032,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.233709677419355e-05,
+      "loss": 0.1969,
+      "step": 5252
+    },
+    {
+      "epoch": 0.084048,
+      "grad_norm": 0.75,
+      "learning_rate": 9.233548387096774e-05,
+      "loss": 0.2016,
+      "step": 5253
+    },
+    {
+      "epoch": 0.084064,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.233387096774194e-05,
+      "loss": 0.1595,
+      "step": 5254
+    },
+    {
+      "epoch": 0.08408,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.233225806451614e-05,
+      "loss": 0.1909,
+      "step": 5255
+    },
+    {
+      "epoch": 0.084096,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.233064516129034e-05,
+      "loss": 0.1363,
+      "step": 5256
+    },
+    {
+      "epoch": 0.084112,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.232903225806452e-05,
+      "loss": 0.187,
+      "step": 5257
+    },
+    {
+      "epoch": 0.084128,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.232741935483872e-05,
+      "loss": 0.1848,
+      "step": 5258
+    },
+    {
+      "epoch": 0.084144,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.232580645161291e-05,
+      "loss": 0.1438,
+      "step": 5259
+    },
+    {
+      "epoch": 0.08416,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.232419354838711e-05,
+      "loss": 0.1548,
+      "step": 5260
+    },
+    {
+      "epoch": 0.084176,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.23225806451613e-05,
+      "loss": 0.1514,
+      "step": 5261
+    },
+    {
+      "epoch": 0.084192,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.23209677419355e-05,
+      "loss": 0.2122,
+      "step": 5262
+    },
+    {
+      "epoch": 0.084208,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.231935483870968e-05,
+      "loss": 0.1326,
+      "step": 5263
+    },
+    {
+      "epoch": 0.084224,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.231774193548387e-05,
+      "loss": 0.2082,
+      "step": 5264
+    },
+    {
+      "epoch": 0.08424,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.231612903225807e-05,
+      "loss": 0.1698,
+      "step": 5265
+    },
+    {
+      "epoch": 0.084256,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.231451612903225e-05,
+      "loss": 0.169,
+      "step": 5266
+    },
+    {
+      "epoch": 0.084272,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.231290322580645e-05,
+      "loss": 0.1732,
+      "step": 5267
+    },
+    {
+      "epoch": 0.084288,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.231129032258065e-05,
+      "loss": 0.2071,
+      "step": 5268
+    },
+    {
+      "epoch": 0.084304,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.230967741935485e-05,
+      "loss": 0.1432,
+      "step": 5269
+    },
+    {
+      "epoch": 0.08432,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.230806451612904e-05,
+      "loss": 0.2179,
+      "step": 5270
+    },
+    {
+      "epoch": 0.084336,
+      "grad_norm": 1.390625,
+      "learning_rate": 9.230645161290324e-05,
+      "loss": 0.181,
+      "step": 5271
+    },
+    {
+      "epoch": 0.084352,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.230483870967742e-05,
+      "loss": 0.1814,
+      "step": 5272
+    },
+    {
+      "epoch": 0.084368,
+      "grad_norm": 0.474609375,
+      "learning_rate": 9.230322580645162e-05,
+      "loss": 0.1324,
+      "step": 5273
+    },
+    {
+      "epoch": 0.084384,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.230161290322581e-05,
+      "loss": 0.2218,
+      "step": 5274
+    },
+    {
+      "epoch": 0.0844,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.230000000000001e-05,
+      "loss": 0.1516,
+      "step": 5275
+    },
+    {
+      "epoch": 0.084416,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.22983870967742e-05,
+      "loss": 0.2019,
+      "step": 5276
+    },
+    {
+      "epoch": 0.084432,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.229677419354839e-05,
+      "loss": 0.1901,
+      "step": 5277
+    },
+    {
+      "epoch": 0.084448,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.229516129032258e-05,
+      "loss": 0.1772,
+      "step": 5278
+    },
+    {
+      "epoch": 0.084464,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.229354838709678e-05,
+      "loss": 0.1652,
+      "step": 5279
+    },
+    {
+      "epoch": 0.08448,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.229193548387098e-05,
+      "loss": 0.1937,
+      "step": 5280
+    },
+    {
+      "epoch": 0.084496,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.229032258064516e-05,
+      "loss": 0.1587,
+      "step": 5281
+    },
+    {
+      "epoch": 0.084512,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.228870967741936e-05,
+      "loss": 0.2084,
+      "step": 5282
+    },
+    {
+      "epoch": 0.084528,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.228709677419355e-05,
+      "loss": 0.1445,
+      "step": 5283
+    },
+    {
+      "epoch": 0.084544,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.228548387096775e-05,
+      "loss": 0.1874,
+      "step": 5284
+    },
+    {
+      "epoch": 0.08456,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.228387096774194e-05,
+      "loss": 0.1554,
+      "step": 5285
+    },
+    {
+      "epoch": 0.084576,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.228225806451614e-05,
+      "loss": 0.1481,
+      "step": 5286
+    },
+    {
+      "epoch": 0.084592,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.228064516129032e-05,
+      "loss": 0.1555,
+      "step": 5287
+    },
+    {
+      "epoch": 0.084608,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.227903225806452e-05,
+      "loss": 0.2065,
+      "step": 5288
+    },
+    {
+      "epoch": 0.084624,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.227741935483871e-05,
+      "loss": 0.1268,
+      "step": 5289
+    },
+    {
+      "epoch": 0.08464,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.227580645161291e-05,
+      "loss": 0.1992,
+      "step": 5290
+    },
+    {
+      "epoch": 0.084656,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.22741935483871e-05,
+      "loss": 0.1451,
+      "step": 5291
+    },
+    {
+      "epoch": 0.084672,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.22725806451613e-05,
+      "loss": 0.1681,
+      "step": 5292
+    },
+    {
+      "epoch": 0.084688,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.227096774193549e-05,
+      "loss": 0.1304,
+      "step": 5293
+    },
+    {
+      "epoch": 0.084704,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.226935483870969e-05,
+      "loss": 0.1576,
+      "step": 5294
+    },
+    {
+      "epoch": 0.08472,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.226774193548388e-05,
+      "loss": 0.226,
+      "step": 5295
+    },
+    {
+      "epoch": 0.084736,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.226612903225806e-05,
+      "loss": 0.1563,
+      "step": 5296
+    },
+    {
+      "epoch": 0.084752,
+      "grad_norm": 0.875,
+      "learning_rate": 9.226451612903226e-05,
+      "loss": 0.1814,
+      "step": 5297
+    },
+    {
+      "epoch": 0.084768,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.226290322580645e-05,
+      "loss": 0.1878,
+      "step": 5298
+    },
+    {
+      "epoch": 0.084784,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.226129032258065e-05,
+      "loss": 0.2141,
+      "step": 5299
+    },
+    {
+      "epoch": 0.0848,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.225967741935484e-05,
+      "loss": 0.2118,
+      "step": 5300
+    },
+    {
+      "epoch": 0.084816,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.225806451612904e-05,
+      "loss": 0.1981,
+      "step": 5301
+    },
+    {
+      "epoch": 0.084832,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.225645161290322e-05,
+      "loss": 0.1811,
+      "step": 5302
+    },
+    {
+      "epoch": 0.084848,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.225483870967742e-05,
+      "loss": 0.1754,
+      "step": 5303
+    },
+    {
+      "epoch": 0.084864,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.225322580645162e-05,
+      "loss": 0.146,
+      "step": 5304
+    },
+    {
+      "epoch": 0.08488,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.225161290322582e-05,
+      "loss": 0.1887,
+      "step": 5305
+    },
+    {
+      "epoch": 0.084896,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.225e-05,
+      "loss": 0.1676,
+      "step": 5306
+    },
+    {
+      "epoch": 0.084912,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.22483870967742e-05,
+      "loss": 0.1843,
+      "step": 5307
+    },
+    {
+      "epoch": 0.084928,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.224677419354839e-05,
+      "loss": 0.1967,
+      "step": 5308
+    },
+    {
+      "epoch": 0.084944,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.224516129032259e-05,
+      "loss": 0.1881,
+      "step": 5309
+    },
+    {
+      "epoch": 0.08496,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.224354838709678e-05,
+      "loss": 0.1845,
+      "step": 5310
+    },
+    {
+      "epoch": 0.084976,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.224193548387096e-05,
+      "loss": 0.1462,
+      "step": 5311
+    },
+    {
+      "epoch": 0.084992,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.224032258064516e-05,
+      "loss": 0.1977,
+      "step": 5312
+    },
+    {
+      "epoch": 0.085008,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.223870967741935e-05,
+      "loss": 0.194,
+      "step": 5313
+    },
+    {
+      "epoch": 0.085024,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.223709677419355e-05,
+      "loss": 0.1573,
+      "step": 5314
+    },
+    {
+      "epoch": 0.08504,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.223548387096775e-05,
+      "loss": 0.1838,
+      "step": 5315
+    },
+    {
+      "epoch": 0.085056,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.223387096774195e-05,
+      "loss": 0.1599,
+      "step": 5316
+    },
+    {
+      "epoch": 0.085072,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.223225806451613e-05,
+      "loss": 0.2132,
+      "step": 5317
+    },
+    {
+      "epoch": 0.085088,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.223064516129033e-05,
+      "loss": 0.1709,
+      "step": 5318
+    },
+    {
+      "epoch": 0.085104,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.222903225806452e-05,
+      "loss": 0.1871,
+      "step": 5319
+    },
+    {
+      "epoch": 0.08512,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.222741935483872e-05,
+      "loss": 0.18,
+      "step": 5320
+    },
+    {
+      "epoch": 0.085136,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.22258064516129e-05,
+      "loss": 0.2193,
+      "step": 5321
+    },
+    {
+      "epoch": 0.085152,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.22241935483871e-05,
+      "loss": 0.2004,
+      "step": 5322
+    },
+    {
+      "epoch": 0.085168,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.222258064516129e-05,
+      "loss": 0.2091,
+      "step": 5323
+    },
+    {
+      "epoch": 0.085184,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.222096774193549e-05,
+      "loss": 0.1724,
+      "step": 5324
+    },
+    {
+      "epoch": 0.0852,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.221935483870968e-05,
+      "loss": 0.1822,
+      "step": 5325
+    },
+    {
+      "epoch": 0.085216,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.221774193548388e-05,
+      "loss": 0.1791,
+      "step": 5326
+    },
+    {
+      "epoch": 0.085232,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.221612903225806e-05,
+      "loss": 0.1761,
+      "step": 5327
+    },
+    {
+      "epoch": 0.085248,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.221451612903226e-05,
+      "loss": 0.178,
+      "step": 5328
+    },
+    {
+      "epoch": 0.085264,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.221290322580646e-05,
+      "loss": 0.1825,
+      "step": 5329
+    },
+    {
+      "epoch": 0.08528,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.221129032258065e-05,
+      "loss": 0.1492,
+      "step": 5330
+    },
+    {
+      "epoch": 0.085296,
+      "grad_norm": 0.625,
+      "learning_rate": 9.220967741935485e-05,
+      "loss": 0.1705,
+      "step": 5331
+    },
+    {
+      "epoch": 0.085312,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.220806451612903e-05,
+      "loss": 0.1488,
+      "step": 5332
+    },
+    {
+      "epoch": 0.085328,
+      "grad_norm": 0.50390625,
+      "learning_rate": 9.220645161290323e-05,
+      "loss": 0.2016,
+      "step": 5333
+    },
+    {
+      "epoch": 0.085344,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.220483870967742e-05,
+      "loss": 0.2205,
+      "step": 5334
+    },
+    {
+      "epoch": 0.08536,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.220322580645162e-05,
+      "loss": 0.1718,
+      "step": 5335
+    },
+    {
+      "epoch": 0.085376,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.22016129032258e-05,
+      "loss": 0.1552,
+      "step": 5336
+    },
+    {
+      "epoch": 0.085392,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.22e-05,
+      "loss": 0.2011,
+      "step": 5337
+    },
+    {
+      "epoch": 0.085408,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.219838709677419e-05,
+      "loss": 0.1867,
+      "step": 5338
+    },
+    {
+      "epoch": 0.085424,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.219677419354839e-05,
+      "loss": 0.1818,
+      "step": 5339
+    },
+    {
+      "epoch": 0.08544,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.219516129032259e-05,
+      "loss": 0.173,
+      "step": 5340
+    },
+    {
+      "epoch": 0.085456,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.219354838709678e-05,
+      "loss": 0.2133,
+      "step": 5341
+    },
+    {
+      "epoch": 0.085472,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.219193548387098e-05,
+      "loss": 0.1597,
+      "step": 5342
+    },
+    {
+      "epoch": 0.085488,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.219032258064516e-05,
+      "loss": 0.1432,
+      "step": 5343
+    },
+    {
+      "epoch": 0.085504,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.218870967741936e-05,
+      "loss": 0.2004,
+      "step": 5344
+    },
+    {
+      "epoch": 0.08552,
+      "grad_norm": 1.0,
+      "learning_rate": 9.218709677419355e-05,
+      "loss": 0.1699,
+      "step": 5345
+    },
+    {
+      "epoch": 0.085536,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.218548387096775e-05,
+      "loss": 0.1825,
+      "step": 5346
+    },
+    {
+      "epoch": 0.085552,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.218387096774193e-05,
+      "loss": 0.2017,
+      "step": 5347
+    },
+    {
+      "epoch": 0.085568,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.218225806451613e-05,
+      "loss": 0.1743,
+      "step": 5348
+    },
+    {
+      "epoch": 0.085584,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.218064516129032e-05,
+      "loss": 0.1523,
+      "step": 5349
+    },
+    {
+      "epoch": 0.0856,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.217903225806452e-05,
+      "loss": 0.1971,
+      "step": 5350
+    },
+    {
+      "epoch": 0.085616,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.217741935483872e-05,
+      "loss": 0.1637,
+      "step": 5351
+    },
+    {
+      "epoch": 0.085632,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.217580645161292e-05,
+      "loss": 0.1641,
+      "step": 5352
+    },
+    {
+      "epoch": 0.085648,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.21741935483871e-05,
+      "loss": 0.197,
+      "step": 5353
+    },
+    {
+      "epoch": 0.085664,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.21725806451613e-05,
+      "loss": 0.1674,
+      "step": 5354
+    },
+    {
+      "epoch": 0.08568,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.217096774193549e-05,
+      "loss": 0.172,
+      "step": 5355
+    },
+    {
+      "epoch": 0.085696,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.216935483870969e-05,
+      "loss": 0.1769,
+      "step": 5356
+    },
+    {
+      "epoch": 0.085712,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.216774193548388e-05,
+      "loss": 0.1659,
+      "step": 5357
+    },
+    {
+      "epoch": 0.085728,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.216612903225806e-05,
+      "loss": 0.2006,
+      "step": 5358
+    },
+    {
+      "epoch": 0.085744,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.216451612903226e-05,
+      "loss": 0.1699,
+      "step": 5359
+    },
+    {
+      "epoch": 0.08576,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.216290322580645e-05,
+      "loss": 0.1497,
+      "step": 5360
+    },
+    {
+      "epoch": 0.085776,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.216129032258065e-05,
+      "loss": 0.1897,
+      "step": 5361
+    },
+    {
+      "epoch": 0.085792,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.215967741935483e-05,
+      "loss": 0.1911,
+      "step": 5362
+    },
+    {
+      "epoch": 0.085808,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.215806451612903e-05,
+      "loss": 0.1534,
+      "step": 5363
+    },
+    {
+      "epoch": 0.085824,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.215645161290323e-05,
+      "loss": 0.1727,
+      "step": 5364
+    },
+    {
+      "epoch": 0.08584,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.215483870967743e-05,
+      "loss": 0.1404,
+      "step": 5365
+    },
+    {
+      "epoch": 0.085856,
+      "grad_norm": 0.447265625,
+      "learning_rate": 9.215322580645162e-05,
+      "loss": 0.1529,
+      "step": 5366
+    },
+    {
+      "epoch": 0.085872,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.215161290322582e-05,
+      "loss": 0.1901,
+      "step": 5367
+    },
+    {
+      "epoch": 0.085888,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.215e-05,
+      "loss": 0.1942,
+      "step": 5368
+    },
+    {
+      "epoch": 0.085904,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.21483870967742e-05,
+      "loss": 0.1801,
+      "step": 5369
+    },
+    {
+      "epoch": 0.08592,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.214677419354839e-05,
+      "loss": 0.17,
+      "step": 5370
+    },
+    {
+      "epoch": 0.085936,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.214516129032259e-05,
+      "loss": 0.2352,
+      "step": 5371
+    },
+    {
+      "epoch": 0.085952,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.214354838709678e-05,
+      "loss": 0.1614,
+      "step": 5372
+    },
+    {
+      "epoch": 0.085968,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.214193548387096e-05,
+      "loss": 0.1519,
+      "step": 5373
+    },
+    {
+      "epoch": 0.085984,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.214032258064516e-05,
+      "loss": 0.1498,
+      "step": 5374
+    },
+    {
+      "epoch": 0.086,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.213870967741936e-05,
+      "loss": 0.1683,
+      "step": 5375
+    },
+    {
+      "epoch": 0.086016,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.213709677419356e-05,
+      "loss": 0.1942,
+      "step": 5376
+    },
+    {
+      "epoch": 0.086032,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.213548387096775e-05,
+      "loss": 0.1661,
+      "step": 5377
+    },
+    {
+      "epoch": 0.086048,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.213387096774195e-05,
+      "loss": 0.2177,
+      "step": 5378
+    },
+    {
+      "epoch": 0.086064,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.213225806451613e-05,
+      "loss": 0.2055,
+      "step": 5379
+    },
+    {
+      "epoch": 0.08608,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.213064516129033e-05,
+      "loss": 0.1702,
+      "step": 5380
+    },
+    {
+      "epoch": 0.086096,
+      "grad_norm": 1.3359375,
+      "learning_rate": 9.212903225806452e-05,
+      "loss": 0.1861,
+      "step": 5381
+    },
+    {
+      "epoch": 0.086112,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.212741935483872e-05,
+      "loss": 0.1516,
+      "step": 5382
+    },
+    {
+      "epoch": 0.086128,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.21258064516129e-05,
+      "loss": 0.128,
+      "step": 5383
+    },
+    {
+      "epoch": 0.086144,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.21241935483871e-05,
+      "loss": 0.189,
+      "step": 5384
+    },
+    {
+      "epoch": 0.08616,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.212258064516129e-05,
+      "loss": 0.1557,
+      "step": 5385
+    },
+    {
+      "epoch": 0.086176,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.212096774193549e-05,
+      "loss": 0.1593,
+      "step": 5386
+    },
+    {
+      "epoch": 0.086192,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.211935483870969e-05,
+      "loss": 0.2173,
+      "step": 5387
+    },
+    {
+      "epoch": 0.086208,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.211774193548387e-05,
+      "loss": 0.1803,
+      "step": 5388
+    },
+    {
+      "epoch": 0.086224,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.211612903225807e-05,
+      "loss": 0.197,
+      "step": 5389
+    },
+    {
+      "epoch": 0.08624,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.211451612903226e-05,
+      "loss": 0.1747,
+      "step": 5390
+    },
+    {
+      "epoch": 0.086256,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.211290322580646e-05,
+      "loss": 0.1719,
+      "step": 5391
+    },
+    {
+      "epoch": 0.086272,
+      "grad_norm": 0.75,
+      "learning_rate": 9.211129032258065e-05,
+      "loss": 0.1852,
+      "step": 5392
+    },
+    {
+      "epoch": 0.086288,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.210967741935485e-05,
+      "loss": 0.1991,
+      "step": 5393
+    },
+    {
+      "epoch": 0.086304,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.210806451612903e-05,
+      "loss": 0.1968,
+      "step": 5394
+    },
+    {
+      "epoch": 0.08632,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.210645161290323e-05,
+      "loss": 0.171,
+      "step": 5395
+    },
+    {
+      "epoch": 0.086336,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.210483870967742e-05,
+      "loss": 0.1804,
+      "step": 5396
+    },
+    {
+      "epoch": 0.086352,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.210322580645162e-05,
+      "loss": 0.1655,
+      "step": 5397
+    },
+    {
+      "epoch": 0.086368,
+      "grad_norm": 0.75,
+      "learning_rate": 9.21016129032258e-05,
+      "loss": 0.1651,
+      "step": 5398
+    },
+    {
+      "epoch": 0.086384,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.21e-05,
+      "loss": 0.1425,
+      "step": 5399
+    },
+    {
+      "epoch": 0.0864,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.20983870967742e-05,
+      "loss": 0.1864,
+      "step": 5400
+    },
+    {
+      "epoch": 0.086416,
+      "grad_norm": 0.875,
+      "learning_rate": 9.20967741935484e-05,
+      "loss": 0.2062,
+      "step": 5401
+    },
+    {
+      "epoch": 0.086432,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.209516129032259e-05,
+      "loss": 0.1621,
+      "step": 5402
+    },
+    {
+      "epoch": 0.086448,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.209354838709679e-05,
+      "loss": 0.257,
+      "step": 5403
+    },
+    {
+      "epoch": 0.086464,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.209193548387097e-05,
+      "loss": 0.1538,
+      "step": 5404
+    },
+    {
+      "epoch": 0.08648,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.209032258064516e-05,
+      "loss": 0.1675,
+      "step": 5405
+    },
+    {
+      "epoch": 0.086496,
+      "grad_norm": 1.34375,
+      "learning_rate": 9.208870967741936e-05,
+      "loss": 0.1894,
+      "step": 5406
+    },
+    {
+      "epoch": 0.086512,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.208709677419355e-05,
+      "loss": 0.1682,
+      "step": 5407
+    },
+    {
+      "epoch": 0.086528,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.208548387096774e-05,
+      "loss": 0.1681,
+      "step": 5408
+    },
+    {
+      "epoch": 0.086544,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.208387096774193e-05,
+      "loss": 0.1533,
+      "step": 5409
+    },
+    {
+      "epoch": 0.08656,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.208225806451613e-05,
+      "loss": 0.1938,
+      "step": 5410
+    },
+    {
+      "epoch": 0.086576,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.208064516129033e-05,
+      "loss": 0.1894,
+      "step": 5411
+    },
+    {
+      "epoch": 0.086592,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.207903225806453e-05,
+      "loss": 0.215,
+      "step": 5412
+    },
+    {
+      "epoch": 0.086608,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.207741935483872e-05,
+      "loss": 0.182,
+      "step": 5413
+    },
+    {
+      "epoch": 0.086624,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.207580645161292e-05,
+      "loss": 0.1534,
+      "step": 5414
+    },
+    {
+      "epoch": 0.08664,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.20741935483871e-05,
+      "loss": 0.1824,
+      "step": 5415
+    },
+    {
+      "epoch": 0.086656,
+      "grad_norm": 0.490234375,
+      "learning_rate": 9.20725806451613e-05,
+      "loss": 0.1415,
+      "step": 5416
+    },
+    {
+      "epoch": 0.086672,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.207096774193549e-05,
+      "loss": 0.1666,
+      "step": 5417
+    },
+    {
+      "epoch": 0.086688,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.206935483870969e-05,
+      "loss": 0.1912,
+      "step": 5418
+    },
+    {
+      "epoch": 0.086704,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.206774193548387e-05,
+      "loss": 0.1772,
+      "step": 5419
+    },
+    {
+      "epoch": 0.08672,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.206612903225806e-05,
+      "loss": 0.1786,
+      "step": 5420
+    },
+    {
+      "epoch": 0.086736,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.206451612903226e-05,
+      "loss": 0.1508,
+      "step": 5421
+    },
+    {
+      "epoch": 0.086752,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.206290322580644e-05,
+      "loss": 0.1939,
+      "step": 5422
+    },
+    {
+      "epoch": 0.086768,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.206129032258064e-05,
+      "loss": 0.2117,
+      "step": 5423
+    },
+    {
+      "epoch": 0.086784,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.205967741935484e-05,
+      "loss": 0.1943,
+      "step": 5424
+    },
+    {
+      "epoch": 0.0868,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.205806451612904e-05,
+      "loss": 0.1745,
+      "step": 5425
+    },
+    {
+      "epoch": 0.086816,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.205645161290323e-05,
+      "loss": 0.1716,
+      "step": 5426
+    },
+    {
+      "epoch": 0.086832,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.205483870967743e-05,
+      "loss": 0.2341,
+      "step": 5427
+    },
+    {
+      "epoch": 0.086848,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.205322580645162e-05,
+      "loss": 0.1883,
+      "step": 5428
+    },
+    {
+      "epoch": 0.086864,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.205161290322582e-05,
+      "loss": 0.1831,
+      "step": 5429
+    },
+    {
+      "epoch": 0.08688,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.205e-05,
+      "loss": 0.1985,
+      "step": 5430
+    },
+    {
+      "epoch": 0.086896,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.20483870967742e-05,
+      "loss": 0.201,
+      "step": 5431
+    },
+    {
+      "epoch": 0.086912,
+      "grad_norm": 1.125,
+      "learning_rate": 9.204677419354839e-05,
+      "loss": 0.2274,
+      "step": 5432
+    },
+    {
+      "epoch": 0.086928,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.204516129032259e-05,
+      "loss": 0.1698,
+      "step": 5433
+    },
+    {
+      "epoch": 0.086944,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.204354838709677e-05,
+      "loss": 0.1743,
+      "step": 5434
+    },
+    {
+      "epoch": 0.08696,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.204193548387097e-05,
+      "loss": 0.1926,
+      "step": 5435
+    },
+    {
+      "epoch": 0.086976,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.204032258064517e-05,
+      "loss": 0.1525,
+      "step": 5436
+    },
+    {
+      "epoch": 0.086992,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.203870967741936e-05,
+      "loss": 0.1519,
+      "step": 5437
+    },
+    {
+      "epoch": 0.087008,
+      "grad_norm": 1.515625,
+      "learning_rate": 9.203709677419356e-05,
+      "loss": 0.2205,
+      "step": 5438
+    },
+    {
+      "epoch": 0.087024,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.203548387096774e-05,
+      "loss": 0.164,
+      "step": 5439
+    },
+    {
+      "epoch": 0.08704,
+      "grad_norm": 1.359375,
+      "learning_rate": 9.203387096774194e-05,
+      "loss": 0.164,
+      "step": 5440
+    },
+    {
+      "epoch": 0.087056,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.203225806451613e-05,
+      "loss": 0.1325,
+      "step": 5441
+    },
+    {
+      "epoch": 0.087072,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.203064516129033e-05,
+      "loss": 0.2436,
+      "step": 5442
+    },
+    {
+      "epoch": 0.087088,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.202903225806452e-05,
+      "loss": 0.164,
+      "step": 5443
+    },
+    {
+      "epoch": 0.087104,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.202741935483871e-05,
+      "loss": 0.1961,
+      "step": 5444
+    },
+    {
+      "epoch": 0.08712,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.20258064516129e-05,
+      "loss": 0.1712,
+      "step": 5445
+    },
+    {
+      "epoch": 0.087136,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.20241935483871e-05,
+      "loss": 0.1724,
+      "step": 5446
+    },
+    {
+      "epoch": 0.087152,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.20225806451613e-05,
+      "loss": 0.1675,
+      "step": 5447
+    },
+    {
+      "epoch": 0.087168,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.20209677419355e-05,
+      "loss": 0.2119,
+      "step": 5448
+    },
+    {
+      "epoch": 0.087184,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.201935483870969e-05,
+      "loss": 0.2271,
+      "step": 5449
+    },
+    {
+      "epoch": 0.0872,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.201774193548387e-05,
+      "loss": 0.1942,
+      "step": 5450
+    },
+    {
+      "epoch": 0.087216,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.201612903225807e-05,
+      "loss": 0.1447,
+      "step": 5451
+    },
+    {
+      "epoch": 0.087232,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.201451612903226e-05,
+      "loss": 0.2061,
+      "step": 5452
+    },
+    {
+      "epoch": 0.087248,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.201290322580646e-05,
+      "loss": 0.2052,
+      "step": 5453
+    },
+    {
+      "epoch": 0.087264,
+      "grad_norm": 1.953125,
+      "learning_rate": 9.201129032258064e-05,
+      "loss": 0.1864,
+      "step": 5454
+    },
+    {
+      "epoch": 0.08728,
+      "grad_norm": 1.296875,
+      "learning_rate": 9.200967741935484e-05,
+      "loss": 0.1821,
+      "step": 5455
+    },
+    {
+      "epoch": 0.087296,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.200806451612903e-05,
+      "loss": 0.1847,
+      "step": 5456
+    },
+    {
+      "epoch": 0.087312,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.200645161290323e-05,
+      "loss": 0.2177,
+      "step": 5457
+    },
+    {
+      "epoch": 0.087328,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.200483870967741e-05,
+      "loss": 0.1543,
+      "step": 5458
+    },
+    {
+      "epoch": 0.087344,
+      "grad_norm": 1.8125,
+      "learning_rate": 9.200322580645161e-05,
+      "loss": 0.1767,
+      "step": 5459
+    },
+    {
+      "epoch": 0.08736,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.200161290322581e-05,
+      "loss": 0.167,
+      "step": 5460
+    },
+    {
+      "epoch": 0.087376,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.200000000000001e-05,
+      "loss": 0.1518,
+      "step": 5461
+    },
+    {
+      "epoch": 0.087392,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.19983870967742e-05,
+      "loss": 0.1833,
+      "step": 5462
+    },
+    {
+      "epoch": 0.087408,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.19967741935484e-05,
+      "loss": 0.1763,
+      "step": 5463
+    },
+    {
+      "epoch": 0.087424,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.199516129032259e-05,
+      "loss": 0.2047,
+      "step": 5464
+    },
+    {
+      "epoch": 0.08744,
+      "grad_norm": 1.296875,
+      "learning_rate": 9.199354838709679e-05,
+      "loss": 0.1798,
+      "step": 5465
+    },
+    {
+      "epoch": 0.087456,
+      "grad_norm": 1.3515625,
+      "learning_rate": 9.199193548387097e-05,
+      "loss": 0.1853,
+      "step": 5466
+    },
+    {
+      "epoch": 0.087472,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.199032258064516e-05,
+      "loss": 0.1859,
+      "step": 5467
+    },
+    {
+      "epoch": 0.087488,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.198870967741936e-05,
+      "loss": 0.1721,
+      "step": 5468
+    },
+    {
+      "epoch": 0.087504,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.198709677419354e-05,
+      "loss": 0.1772,
+      "step": 5469
+    },
+    {
+      "epoch": 0.08752,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.198548387096774e-05,
+      "loss": 0.2085,
+      "step": 5470
+    },
+    {
+      "epoch": 0.087536,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.198387096774194e-05,
+      "loss": 0.2085,
+      "step": 5471
+    },
+    {
+      "epoch": 0.087552,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.198225806451614e-05,
+      "loss": 0.1641,
+      "step": 5472
+    },
+    {
+      "epoch": 0.087568,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.198064516129033e-05,
+      "loss": 0.1914,
+      "step": 5473
+    },
+    {
+      "epoch": 0.087584,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.197903225806453e-05,
+      "loss": 0.1784,
+      "step": 5474
+    },
+    {
+      "epoch": 0.0876,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.197741935483871e-05,
+      "loss": 0.1437,
+      "step": 5475
+    },
+    {
+      "epoch": 0.087616,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.197580645161291e-05,
+      "loss": 0.1672,
+      "step": 5476
+    },
+    {
+      "epoch": 0.087632,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.19741935483871e-05,
+      "loss": 0.2004,
+      "step": 5477
+    },
+    {
+      "epoch": 0.087648,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.19725806451613e-05,
+      "loss": 0.1784,
+      "step": 5478
+    },
+    {
+      "epoch": 0.087664,
+      "grad_norm": 1.453125,
+      "learning_rate": 9.197096774193548e-05,
+      "loss": 0.2439,
+      "step": 5479
+    },
+    {
+      "epoch": 0.08768,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.196935483870968e-05,
+      "loss": 0.1695,
+      "step": 5480
+    },
+    {
+      "epoch": 0.087696,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.196774193548387e-05,
+      "loss": 0.1824,
+      "step": 5481
+    },
+    {
+      "epoch": 0.087712,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.196612903225807e-05,
+      "loss": 0.1792,
+      "step": 5482
+    },
+    {
+      "epoch": 0.087728,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.196451612903227e-05,
+      "loss": 0.1668,
+      "step": 5483
+    },
+    {
+      "epoch": 0.087744,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.196290322580646e-05,
+      "loss": 0.1783,
+      "step": 5484
+    },
+    {
+      "epoch": 0.08776,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.196129032258066e-05,
+      "loss": 0.1497,
+      "step": 5485
+    },
+    {
+      "epoch": 0.087776,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.195967741935484e-05,
+      "loss": 0.1562,
+      "step": 5486
+    },
+    {
+      "epoch": 0.087792,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.195806451612904e-05,
+      "loss": 0.1832,
+      "step": 5487
+    },
+    {
+      "epoch": 0.087808,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.195645161290323e-05,
+      "loss": 0.1749,
+      "step": 5488
+    },
+    {
+      "epoch": 0.087824,
+      "grad_norm": 1.4375,
+      "learning_rate": 9.195483870967743e-05,
+      "loss": 0.2118,
+      "step": 5489
+    },
+    {
+      "epoch": 0.08784,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.195322580645161e-05,
+      "loss": 0.2265,
+      "step": 5490
+    },
+    {
+      "epoch": 0.087856,
+      "grad_norm": 1.9609375,
+      "learning_rate": 9.195161290322581e-05,
+      "loss": 0.2289,
+      "step": 5491
+    },
+    {
+      "epoch": 0.087872,
+      "grad_norm": 1.3359375,
+      "learning_rate": 9.195e-05,
+      "loss": 0.1681,
+      "step": 5492
+    },
+    {
+      "epoch": 0.087888,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.19483870967742e-05,
+      "loss": 0.1788,
+      "step": 5493
+    },
+    {
+      "epoch": 0.087904,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.194677419354838e-05,
+      "loss": 0.1783,
+      "step": 5494
+    },
+    {
+      "epoch": 0.08792,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.194516129032258e-05,
+      "loss": 0.1701,
+      "step": 5495
+    },
+    {
+      "epoch": 0.087936,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.194354838709678e-05,
+      "loss": 0.1779,
+      "step": 5496
+    },
+    {
+      "epoch": 0.087952,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.194193548387097e-05,
+      "loss": 0.2077,
+      "step": 5497
+    },
+    {
+      "epoch": 0.087968,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.194032258064517e-05,
+      "loss": 0.1677,
+      "step": 5498
+    },
+    {
+      "epoch": 0.087984,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.193870967741936e-05,
+      "loss": 0.2093,
+      "step": 5499
+    },
+    {
+      "epoch": 0.088,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.193709677419356e-05,
+      "loss": 0.2172,
+      "step": 5500
+    },
+    {
+      "epoch": 0.088016,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.193548387096774e-05,
+      "loss": 0.1705,
+      "step": 5501
+    },
+    {
+      "epoch": 0.088032,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.193387096774194e-05,
+      "loss": 0.1689,
+      "step": 5502
+    },
+    {
+      "epoch": 0.088048,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.193225806451613e-05,
+      "loss": 0.1822,
+      "step": 5503
+    },
+    {
+      "epoch": 0.088064,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.193064516129033e-05,
+      "loss": 0.1544,
+      "step": 5504
+    },
+    {
+      "epoch": 0.08808,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.192903225806451e-05,
+      "loss": 0.1542,
+      "step": 5505
+    },
+    {
+      "epoch": 0.088096,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.192741935483871e-05,
+      "loss": 0.1945,
+      "step": 5506
+    },
+    {
+      "epoch": 0.088112,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.192580645161291e-05,
+      "loss": 0.175,
+      "step": 5507
+    },
+    {
+      "epoch": 0.088128,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.192419354838711e-05,
+      "loss": 0.1637,
+      "step": 5508
+    },
+    {
+      "epoch": 0.088144,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.19225806451613e-05,
+      "loss": 0.2008,
+      "step": 5509
+    },
+    {
+      "epoch": 0.08816,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.19209677419355e-05,
+      "loss": 0.2039,
+      "step": 5510
+    },
+    {
+      "epoch": 0.088176,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.191935483870968e-05,
+      "loss": 0.1775,
+      "step": 5511
+    },
+    {
+      "epoch": 0.088192,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.191774193548388e-05,
+      "loss": 0.1784,
+      "step": 5512
+    },
+    {
+      "epoch": 0.088208,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.191612903225807e-05,
+      "loss": 0.1964,
+      "step": 5513
+    },
+    {
+      "epoch": 0.088224,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.191451612903226e-05,
+      "loss": 0.1908,
+      "step": 5514
+    },
+    {
+      "epoch": 0.08824,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.191290322580645e-05,
+      "loss": 0.155,
+      "step": 5515
+    },
+    {
+      "epoch": 0.088256,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.191129032258064e-05,
+      "loss": 0.1487,
+      "step": 5516
+    },
+    {
+      "epoch": 0.088272,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.190967741935484e-05,
+      "loss": 0.2028,
+      "step": 5517
+    },
+    {
+      "epoch": 0.088288,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.190806451612903e-05,
+      "loss": 0.1601,
+      "step": 5518
+    },
+    {
+      "epoch": 0.088304,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.190645161290323e-05,
+      "loss": 0.205,
+      "step": 5519
+    },
+    {
+      "epoch": 0.08832,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.190483870967743e-05,
+      "loss": 0.1774,
+      "step": 5520
+    },
+    {
+      "epoch": 0.088336,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.190322580645163e-05,
+      "loss": 0.199,
+      "step": 5521
+    },
+    {
+      "epoch": 0.088352,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.190161290322581e-05,
+      "loss": 0.1751,
+      "step": 5522
+    },
+    {
+      "epoch": 0.088368,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.190000000000001e-05,
+      "loss": 0.1612,
+      "step": 5523
+    },
+    {
+      "epoch": 0.088384,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.18983870967742e-05,
+      "loss": 0.1652,
+      "step": 5524
+    },
+    {
+      "epoch": 0.0884,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.18967741935484e-05,
+      "loss": 0.1821,
+      "step": 5525
+    },
+    {
+      "epoch": 0.088416,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.189516129032258e-05,
+      "loss": 0.163,
+      "step": 5526
+    },
+    {
+      "epoch": 0.088432,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.189354838709678e-05,
+      "loss": 0.2083,
+      "step": 5527
+    },
+    {
+      "epoch": 0.088448,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.189193548387097e-05,
+      "loss": 0.1565,
+      "step": 5528
+    },
+    {
+      "epoch": 0.088464,
+      "grad_norm": 1.3359375,
+      "learning_rate": 9.189032258064515e-05,
+      "loss": 0.1732,
+      "step": 5529
+    },
+    {
+      "epoch": 0.08848,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.188870967741935e-05,
+      "loss": 0.2179,
+      "step": 5530
+    },
+    {
+      "epoch": 0.088496,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.188709677419355e-05,
+      "loss": 0.127,
+      "step": 5531
+    },
+    {
+      "epoch": 0.088512,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.188548387096775e-05,
+      "loss": 0.1393,
+      "step": 5532
+    },
+    {
+      "epoch": 0.088528,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.188387096774194e-05,
+      "loss": 0.1745,
+      "step": 5533
+    },
+    {
+      "epoch": 0.088544,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.188225806451614e-05,
+      "loss": 0.2382,
+      "step": 5534
+    },
+    {
+      "epoch": 0.08856,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.188064516129033e-05,
+      "loss": 0.1816,
+      "step": 5535
+    },
+    {
+      "epoch": 0.088576,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.187903225806453e-05,
+      "loss": 0.1913,
+      "step": 5536
+    },
+    {
+      "epoch": 0.088592,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.187741935483871e-05,
+      "loss": 0.1317,
+      "step": 5537
+    },
+    {
+      "epoch": 0.088608,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.187580645161291e-05,
+      "loss": 0.1889,
+      "step": 5538
+    },
+    {
+      "epoch": 0.088624,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.18741935483871e-05,
+      "loss": 0.1611,
+      "step": 5539
+    },
+    {
+      "epoch": 0.08864,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.18725806451613e-05,
+      "loss": 0.1836,
+      "step": 5540
+    },
+    {
+      "epoch": 0.088656,
+      "grad_norm": 1.421875,
+      "learning_rate": 9.187096774193548e-05,
+      "loss": 0.1664,
+      "step": 5541
+    },
+    {
+      "epoch": 0.088672,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.186935483870968e-05,
+      "loss": 0.1867,
+      "step": 5542
+    },
+    {
+      "epoch": 0.088688,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.186774193548388e-05,
+      "loss": 0.1738,
+      "step": 5543
+    },
+    {
+      "epoch": 0.088704,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.186612903225807e-05,
+      "loss": 0.1423,
+      "step": 5544
+    },
+    {
+      "epoch": 0.08872,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.186451612903227e-05,
+      "loss": 0.166,
+      "step": 5545
+    },
+    {
+      "epoch": 0.088736,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.186290322580645e-05,
+      "loss": 0.1492,
+      "step": 5546
+    },
+    {
+      "epoch": 0.088752,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.186129032258065e-05,
+      "loss": 0.2107,
+      "step": 5547
+    },
+    {
+      "epoch": 0.088768,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.185967741935484e-05,
+      "loss": 0.1721,
+      "step": 5548
+    },
+    {
+      "epoch": 0.088784,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.185806451612904e-05,
+      "loss": 0.1469,
+      "step": 5549
+    },
+    {
+      "epoch": 0.0888,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.185645161290323e-05,
+      "loss": 0.1852,
+      "step": 5550
+    },
+    {
+      "epoch": 0.088816,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.185483870967742e-05,
+      "loss": 0.1711,
+      "step": 5551
+    },
+    {
+      "epoch": 0.088832,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.185322580645161e-05,
+      "loss": 0.1848,
+      "step": 5552
+    },
+    {
+      "epoch": 0.088848,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.185161290322581e-05,
+      "loss": 0.1837,
+      "step": 5553
+    },
+    {
+      "epoch": 0.088864,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.185e-05,
+      "loss": 0.1396,
+      "step": 5554
+    },
+    {
+      "epoch": 0.08888,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.18483870967742e-05,
+      "loss": 0.1381,
+      "step": 5555
+    },
+    {
+      "epoch": 0.088896,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.18467741935484e-05,
+      "loss": 0.1294,
+      "step": 5556
+    },
+    {
+      "epoch": 0.088912,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.18451612903226e-05,
+      "loss": 0.1864,
+      "step": 5557
+    },
+    {
+      "epoch": 0.088928,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.184354838709678e-05,
+      "loss": 0.1644,
+      "step": 5558
+    },
+    {
+      "epoch": 0.088944,
+      "grad_norm": 1.4453125,
+      "learning_rate": 9.184193548387097e-05,
+      "loss": 0.1741,
+      "step": 5559
+    },
+    {
+      "epoch": 0.08896,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.184032258064517e-05,
+      "loss": 0.1998,
+      "step": 5560
+    },
+    {
+      "epoch": 0.088976,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.183870967741935e-05,
+      "loss": 0.1794,
+      "step": 5561
+    },
+    {
+      "epoch": 0.088992,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.183709677419355e-05,
+      "loss": 0.1699,
+      "step": 5562
+    },
+    {
+      "epoch": 0.089008,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.183548387096774e-05,
+      "loss": 0.1456,
+      "step": 5563
+    },
+    {
+      "epoch": 0.089024,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.183387096774194e-05,
+      "loss": 0.1812,
+      "step": 5564
+    },
+    {
+      "epoch": 0.08904,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.183225806451612e-05,
+      "loss": 0.2052,
+      "step": 5565
+    },
+    {
+      "epoch": 0.089056,
+      "grad_norm": 0.75,
+      "learning_rate": 9.183064516129032e-05,
+      "loss": 0.1681,
+      "step": 5566
+    },
+    {
+      "epoch": 0.089072,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.182903225806452e-05,
+      "loss": 0.1797,
+      "step": 5567
+    },
+    {
+      "epoch": 0.089088,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.182741935483872e-05,
+      "loss": 0.1841,
+      "step": 5568
+    },
+    {
+      "epoch": 0.089104,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.182580645161291e-05,
+      "loss": 0.2327,
+      "step": 5569
+    },
+    {
+      "epoch": 0.08912,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.182419354838711e-05,
+      "loss": 0.1689,
+      "step": 5570
+    },
+    {
+      "epoch": 0.089136,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.18225806451613e-05,
+      "loss": 0.1553,
+      "step": 5571
+    },
+    {
+      "epoch": 0.089152,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.18209677419355e-05,
+      "loss": 0.1457,
+      "step": 5572
+    },
+    {
+      "epoch": 0.089168,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.181935483870968e-05,
+      "loss": 0.1993,
+      "step": 5573
+    },
+    {
+      "epoch": 0.089184,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.181774193548388e-05,
+      "loss": 0.1358,
+      "step": 5574
+    },
+    {
+      "epoch": 0.0892,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.181612903225807e-05,
+      "loss": 0.1509,
+      "step": 5575
+    },
+    {
+      "epoch": 0.089216,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.181451612903225e-05,
+      "loss": 0.1652,
+      "step": 5576
+    },
+    {
+      "epoch": 0.089232,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.181290322580645e-05,
+      "loss": 0.196,
+      "step": 5577
+    },
+    {
+      "epoch": 0.089248,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.181129032258065e-05,
+      "loss": 0.2005,
+      "step": 5578
+    },
+    {
+      "epoch": 0.089264,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.180967741935484e-05,
+      "loss": 0.2054,
+      "step": 5579
+    },
+    {
+      "epoch": 0.08928,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.180806451612904e-05,
+      "loss": 0.1801,
+      "step": 5580
+    },
+    {
+      "epoch": 0.089296,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.180645161290324e-05,
+      "loss": 0.1665,
+      "step": 5581
+    },
+    {
+      "epoch": 0.089312,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.180483870967742e-05,
+      "loss": 0.17,
+      "step": 5582
+    },
+    {
+      "epoch": 0.089328,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.180322580645162e-05,
+      "loss": 0.1754,
+      "step": 5583
+    },
+    {
+      "epoch": 0.089344,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.180161290322581e-05,
+      "loss": 0.1652,
+      "step": 5584
+    },
+    {
+      "epoch": 0.08936,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.180000000000001e-05,
+      "loss": 0.1917,
+      "step": 5585
+    },
+    {
+      "epoch": 0.089376,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.17983870967742e-05,
+      "loss": 0.1667,
+      "step": 5586
+    },
+    {
+      "epoch": 0.089392,
+      "grad_norm": 0.625,
+      "learning_rate": 9.17967741935484e-05,
+      "loss": 0.1762,
+      "step": 5587
+    },
+    {
+      "epoch": 0.089408,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.179516129032258e-05,
+      "loss": 0.137,
+      "step": 5588
+    },
+    {
+      "epoch": 0.089424,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.179354838709678e-05,
+      "loss": 0.2214,
+      "step": 5589
+    },
+    {
+      "epoch": 0.08944,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.179193548387097e-05,
+      "loss": 0.1833,
+      "step": 5590
+    },
+    {
+      "epoch": 0.089456,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.179032258064517e-05,
+      "loss": 0.2061,
+      "step": 5591
+    },
+    {
+      "epoch": 0.089472,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.178870967741937e-05,
+      "loss": 0.1532,
+      "step": 5592
+    },
+    {
+      "epoch": 0.089488,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.178709677419355e-05,
+      "loss": 0.2267,
+      "step": 5593
+    },
+    {
+      "epoch": 0.089504,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.178548387096775e-05,
+      "loss": 0.1859,
+      "step": 5594
+    },
+    {
+      "epoch": 0.08952,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.178387096774194e-05,
+      "loss": 0.1858,
+      "step": 5595
+    },
+    {
+      "epoch": 0.089536,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.178225806451614e-05,
+      "loss": 0.2036,
+      "step": 5596
+    },
+    {
+      "epoch": 0.089552,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.178064516129032e-05,
+      "loss": 0.154,
+      "step": 5597
+    },
+    {
+      "epoch": 0.089568,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.177903225806452e-05,
+      "loss": 0.1731,
+      "step": 5598
+    },
+    {
+      "epoch": 0.089584,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.177741935483871e-05,
+      "loss": 0.1599,
+      "step": 5599
+    },
+    {
+      "epoch": 0.0896,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.177580645161291e-05,
+      "loss": 0.1846,
+      "step": 5600
+    },
+    {
+      "epoch": 0.089616,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.17741935483871e-05,
+      "loss": 0.1493,
+      "step": 5601
+    },
+    {
+      "epoch": 0.089632,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.17725806451613e-05,
+      "loss": 0.1766,
+      "step": 5602
+    },
+    {
+      "epoch": 0.089648,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.17709677419355e-05,
+      "loss": 0.1736,
+      "step": 5603
+    },
+    {
+      "epoch": 0.089664,
+      "grad_norm": 1.3515625,
+      "learning_rate": 9.17693548387097e-05,
+      "loss": 0.1539,
+      "step": 5604
+    },
+    {
+      "epoch": 0.08968,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.176774193548388e-05,
+      "loss": 0.1692,
+      "step": 5605
+    },
+    {
+      "epoch": 0.089696,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.176612903225807e-05,
+      "loss": 0.1986,
+      "step": 5606
+    },
+    {
+      "epoch": 0.089712,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.176451612903227e-05,
+      "loss": 0.1617,
+      "step": 5607
+    },
+    {
+      "epoch": 0.089728,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.176290322580645e-05,
+      "loss": 0.1822,
+      "step": 5608
+    },
+    {
+      "epoch": 0.089744,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.176129032258065e-05,
+      "loss": 0.1809,
+      "step": 5609
+    },
+    {
+      "epoch": 0.08976,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.175967741935484e-05,
+      "loss": 0.1268,
+      "step": 5610
+    },
+    {
+      "epoch": 0.089776,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.175806451612904e-05,
+      "loss": 0.1816,
+      "step": 5611
+    },
+    {
+      "epoch": 0.089792,
+      "grad_norm": 0.498046875,
+      "learning_rate": 9.175645161290322e-05,
+      "loss": 0.1628,
+      "step": 5612
+    },
+    {
+      "epoch": 0.089808,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.175483870967742e-05,
+      "loss": 0.1796,
+      "step": 5613
+    },
+    {
+      "epoch": 0.089824,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.175322580645161e-05,
+      "loss": 0.1943,
+      "step": 5614
+    },
+    {
+      "epoch": 0.08984,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.175161290322581e-05,
+      "loss": 0.1284,
+      "step": 5615
+    },
+    {
+      "epoch": 0.089856,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.175000000000001e-05,
+      "loss": 0.1461,
+      "step": 5616
+    },
+    {
+      "epoch": 0.089872,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.174838709677421e-05,
+      "loss": 0.1865,
+      "step": 5617
+    },
+    {
+      "epoch": 0.089888,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.17467741935484e-05,
+      "loss": 0.1988,
+      "step": 5618
+    },
+    {
+      "epoch": 0.089904,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.174516129032259e-05,
+      "loss": 0.1727,
+      "step": 5619
+    },
+    {
+      "epoch": 0.08992,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.174354838709678e-05,
+      "loss": 0.1909,
+      "step": 5620
+    },
+    {
+      "epoch": 0.089936,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.174193548387097e-05,
+      "loss": 0.1488,
+      "step": 5621
+    },
+    {
+      "epoch": 0.089952,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.174032258064516e-05,
+      "loss": 0.1819,
+      "step": 5622
+    },
+    {
+      "epoch": 0.089968,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.173870967741935e-05,
+      "loss": 0.2102,
+      "step": 5623
+    },
+    {
+      "epoch": 0.089984,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.173709677419355e-05,
+      "loss": 0.1659,
+      "step": 5624
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.173548387096774e-05,
+      "loss": 0.2047,
+      "step": 5625
+    },
+    {
+      "epoch": 0.090016,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.173387096774194e-05,
+      "loss": 0.1807,
+      "step": 5626
+    },
+    {
+      "epoch": 0.090032,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.173225806451614e-05,
+      "loss": 0.1733,
+      "step": 5627
+    },
+    {
+      "epoch": 0.090048,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.173064516129034e-05,
+      "loss": 0.1809,
+      "step": 5628
+    },
+    {
+      "epoch": 0.090064,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.172903225806452e-05,
+      "loss": 0.1378,
+      "step": 5629
+    },
+    {
+      "epoch": 0.09008,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.172741935483872e-05,
+      "loss": 0.1387,
+      "step": 5630
+    },
+    {
+      "epoch": 0.090096,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.172580645161291e-05,
+      "loss": 0.1305,
+      "step": 5631
+    },
+    {
+      "epoch": 0.090112,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.17241935483871e-05,
+      "loss": 0.1879,
+      "step": 5632
+    },
+    {
+      "epoch": 0.090128,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.172258064516129e-05,
+      "loss": 0.1332,
+      "step": 5633
+    },
+    {
+      "epoch": 0.090144,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.172096774193549e-05,
+      "loss": 0.1525,
+      "step": 5634
+    },
+    {
+      "epoch": 0.09016,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.171935483870968e-05,
+      "loss": 0.1696,
+      "step": 5635
+    },
+    {
+      "epoch": 0.090176,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.171774193548388e-05,
+      "loss": 0.1907,
+      "step": 5636
+    },
+    {
+      "epoch": 0.090192,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.171612903225806e-05,
+      "loss": 0.2091,
+      "step": 5637
+    },
+    {
+      "epoch": 0.090208,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.171451612903226e-05,
+      "loss": 0.1912,
+      "step": 5638
+    },
+    {
+      "epoch": 0.090224,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.171290322580646e-05,
+      "loss": 0.1728,
+      "step": 5639
+    },
+    {
+      "epoch": 0.09024,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.171129032258065e-05,
+      "loss": 0.1871,
+      "step": 5640
+    },
+    {
+      "epoch": 0.090256,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.170967741935485e-05,
+      "loss": 0.1458,
+      "step": 5641
+    },
+    {
+      "epoch": 0.090272,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.170806451612904e-05,
+      "loss": 0.1594,
+      "step": 5642
+    },
+    {
+      "epoch": 0.090288,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.170645161290323e-05,
+      "loss": 0.1588,
+      "step": 5643
+    },
+    {
+      "epoch": 0.090304,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.170483870967742e-05,
+      "loss": 0.1491,
+      "step": 5644
+    },
+    {
+      "epoch": 0.09032,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.170322580645162e-05,
+      "loss": 0.1616,
+      "step": 5645
+    },
+    {
+      "epoch": 0.090336,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.17016129032258e-05,
+      "loss": 0.1765,
+      "step": 5646
+    },
+    {
+      "epoch": 0.090352,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.17e-05,
+      "loss": 0.2104,
+      "step": 5647
+    },
+    {
+      "epoch": 0.090368,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.169838709677419e-05,
+      "loss": 0.1945,
+      "step": 5648
+    },
+    {
+      "epoch": 0.090384,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.169677419354839e-05,
+      "loss": 0.162,
+      "step": 5649
+    },
+    {
+      "epoch": 0.0904,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.169516129032258e-05,
+      "loss": 0.1703,
+      "step": 5650
+    },
+    {
+      "epoch": 0.090416,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.169354838709678e-05,
+      "loss": 0.1719,
+      "step": 5651
+    },
+    {
+      "epoch": 0.090432,
+      "grad_norm": 1.296875,
+      "learning_rate": 9.169193548387098e-05,
+      "loss": 0.1807,
+      "step": 5652
+    },
+    {
+      "epoch": 0.090448,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.169032258064516e-05,
+      "loss": 0.1936,
+      "step": 5653
+    },
+    {
+      "epoch": 0.090464,
+      "grad_norm": 0.75,
+      "learning_rate": 9.168870967741936e-05,
+      "loss": 0.1697,
+      "step": 5654
+    },
+    {
+      "epoch": 0.09048,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.168709677419355e-05,
+      "loss": 0.1737,
+      "step": 5655
+    },
+    {
+      "epoch": 0.090496,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.168548387096775e-05,
+      "loss": 0.1703,
+      "step": 5656
+    },
+    {
+      "epoch": 0.090512,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.168387096774193e-05,
+      "loss": 0.1705,
+      "step": 5657
+    },
+    {
+      "epoch": 0.090528,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.168225806451613e-05,
+      "loss": 0.2018,
+      "step": 5658
+    },
+    {
+      "epoch": 0.090544,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.168064516129032e-05,
+      "loss": 0.1513,
+      "step": 5659
+    },
+    {
+      "epoch": 0.09056,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.167903225806452e-05,
+      "loss": 0.1966,
+      "step": 5660
+    },
+    {
+      "epoch": 0.090576,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.16774193548387e-05,
+      "loss": 0.1698,
+      "step": 5661
+    },
+    {
+      "epoch": 0.090592,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.16758064516129e-05,
+      "loss": 0.189,
+      "step": 5662
+    },
+    {
+      "epoch": 0.090608,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.16741935483871e-05,
+      "loss": 0.1953,
+      "step": 5663
+    },
+    {
+      "epoch": 0.090624,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.16725806451613e-05,
+      "loss": 0.2101,
+      "step": 5664
+    },
+    {
+      "epoch": 0.09064,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.167096774193549e-05,
+      "loss": 0.1614,
+      "step": 5665
+    },
+    {
+      "epoch": 0.090656,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.166935483870969e-05,
+      "loss": 0.1606,
+      "step": 5666
+    },
+    {
+      "epoch": 0.090672,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.166774193548388e-05,
+      "loss": 0.1835,
+      "step": 5667
+    },
+    {
+      "epoch": 0.090688,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.166612903225806e-05,
+      "loss": 0.1825,
+      "step": 5668
+    },
+    {
+      "epoch": 0.090704,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.166451612903226e-05,
+      "loss": 0.1784,
+      "step": 5669
+    },
+    {
+      "epoch": 0.09072,
+      "grad_norm": 0.625,
+      "learning_rate": 9.166290322580645e-05,
+      "loss": 0.1547,
+      "step": 5670
+    },
+    {
+      "epoch": 0.090736,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.166129032258065e-05,
+      "loss": 0.2351,
+      "step": 5671
+    },
+    {
+      "epoch": 0.090752,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.165967741935483e-05,
+      "loss": 0.1657,
+      "step": 5672
+    },
+    {
+      "epoch": 0.090768,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.165806451612903e-05,
+      "loss": 0.1757,
+      "step": 5673
+    },
+    {
+      "epoch": 0.090784,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.165645161290322e-05,
+      "loss": 0.201,
+      "step": 5674
+    },
+    {
+      "epoch": 0.0908,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.165483870967742e-05,
+      "loss": 0.1975,
+      "step": 5675
+    },
+    {
+      "epoch": 0.090816,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.165322580645162e-05,
+      "loss": 0.1607,
+      "step": 5676
+    },
+    {
+      "epoch": 0.090832,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.165161290322582e-05,
+      "loss": 0.1566,
+      "step": 5677
+    },
+    {
+      "epoch": 0.090848,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.165e-05,
+      "loss": 0.1696,
+      "step": 5678
+    },
+    {
+      "epoch": 0.090864,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.16483870967742e-05,
+      "loss": 0.2062,
+      "step": 5679
+    },
+    {
+      "epoch": 0.09088,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.164677419354839e-05,
+      "loss": 0.1883,
+      "step": 5680
+    },
+    {
+      "epoch": 0.090896,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.164516129032259e-05,
+      "loss": 0.1936,
+      "step": 5681
+    },
+    {
+      "epoch": 0.090912,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.164354838709678e-05,
+      "loss": 0.1774,
+      "step": 5682
+    },
+    {
+      "epoch": 0.090928,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.164193548387098e-05,
+      "loss": 0.1538,
+      "step": 5683
+    },
+    {
+      "epoch": 0.090944,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.164032258064516e-05,
+      "loss": 0.1996,
+      "step": 5684
+    },
+    {
+      "epoch": 0.09096,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.163870967741935e-05,
+      "loss": 0.1768,
+      "step": 5685
+    },
+    {
+      "epoch": 0.090976,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.163709677419355e-05,
+      "loss": 0.1757,
+      "step": 5686
+    },
+    {
+      "epoch": 0.090992,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.163548387096775e-05,
+      "loss": 0.1649,
+      "step": 5687
+    },
+    {
+      "epoch": 0.091008,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.163387096774195e-05,
+      "loss": 0.1442,
+      "step": 5688
+    },
+    {
+      "epoch": 0.091024,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.163225806451613e-05,
+      "loss": 0.2031,
+      "step": 5689
+    },
+    {
+      "epoch": 0.09104,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.163064516129033e-05,
+      "loss": 0.19,
+      "step": 5690
+    },
+    {
+      "epoch": 0.091056,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.162903225806452e-05,
+      "loss": 0.17,
+      "step": 5691
+    },
+    {
+      "epoch": 0.091072,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.162741935483872e-05,
+      "loss": 0.1365,
+      "step": 5692
+    },
+    {
+      "epoch": 0.091088,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.16258064516129e-05,
+      "loss": 0.2053,
+      "step": 5693
+    },
+    {
+      "epoch": 0.091104,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.16241935483871e-05,
+      "loss": 0.1969,
+      "step": 5694
+    },
+    {
+      "epoch": 0.09112,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.162258064516129e-05,
+      "loss": 0.2045,
+      "step": 5695
+    },
+    {
+      "epoch": 0.091136,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.162096774193549e-05,
+      "loss": 0.1749,
+      "step": 5696
+    },
+    {
+      "epoch": 0.091152,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.161935483870968e-05,
+      "loss": 0.2097,
+      "step": 5697
+    },
+    {
+      "epoch": 0.091168,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.161774193548388e-05,
+      "loss": 0.2156,
+      "step": 5698
+    },
+    {
+      "epoch": 0.091184,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.161612903225808e-05,
+      "loss": 0.1665,
+      "step": 5699
+    },
+    {
+      "epoch": 0.0912,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.161451612903226e-05,
+      "loss": 0.1375,
+      "step": 5700
+    },
+    {
+      "epoch": 0.091216,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.161290322580646e-05,
+      "loss": 0.2084,
+      "step": 5701
+    },
+    {
+      "epoch": 0.091232,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.161129032258065e-05,
+      "loss": 0.215,
+      "step": 5702
+    },
+    {
+      "epoch": 0.091248,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.160967741935485e-05,
+      "loss": 0.1985,
+      "step": 5703
+    },
+    {
+      "epoch": 0.091264,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.160806451612903e-05,
+      "loss": 0.1664,
+      "step": 5704
+    },
+    {
+      "epoch": 0.09128,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.160645161290323e-05,
+      "loss": 0.2183,
+      "step": 5705
+    },
+    {
+      "epoch": 0.091296,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.160483870967742e-05,
+      "loss": 0.2135,
+      "step": 5706
+    },
+    {
+      "epoch": 0.091312,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.160322580645162e-05,
+      "loss": 0.1599,
+      "step": 5707
+    },
+    {
+      "epoch": 0.091328,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.16016129032258e-05,
+      "loss": 0.1796,
+      "step": 5708
+    },
+    {
+      "epoch": 0.091344,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.16e-05,
+      "loss": 0.1652,
+      "step": 5709
+    },
+    {
+      "epoch": 0.09136,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.159838709677419e-05,
+      "loss": 0.2137,
+      "step": 5710
+    },
+    {
+      "epoch": 0.091376,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.159677419354839e-05,
+      "loss": 0.15,
+      "step": 5711
+    },
+    {
+      "epoch": 0.091392,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.159516129032259e-05,
+      "loss": 0.1768,
+      "step": 5712
+    },
+    {
+      "epoch": 0.091408,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.159354838709679e-05,
+      "loss": 0.1604,
+      "step": 5713
+    },
+    {
+      "epoch": 0.091424,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.159193548387097e-05,
+      "loss": 0.1436,
+      "step": 5714
+    },
+    {
+      "epoch": 0.09144,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.159032258064516e-05,
+      "loss": 0.1735,
+      "step": 5715
+    },
+    {
+      "epoch": 0.091456,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.158870967741936e-05,
+      "loss": 0.1726,
+      "step": 5716
+    },
+    {
+      "epoch": 0.091472,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.158709677419355e-05,
+      "loss": 0.1998,
+      "step": 5717
+    },
+    {
+      "epoch": 0.091488,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.158548387096775e-05,
+      "loss": 0.1839,
+      "step": 5718
+    },
+    {
+      "epoch": 0.091504,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.158387096774193e-05,
+      "loss": 0.1737,
+      "step": 5719
+    },
+    {
+      "epoch": 0.09152,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.158225806451613e-05,
+      "loss": 0.1844,
+      "step": 5720
+    },
+    {
+      "epoch": 0.091536,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.158064516129032e-05,
+      "loss": 0.1425,
+      "step": 5721
+    },
+    {
+      "epoch": 0.091552,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.157903225806452e-05,
+      "loss": 0.1473,
+      "step": 5722
+    },
+    {
+      "epoch": 0.091568,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.157741935483872e-05,
+      "loss": 0.1935,
+      "step": 5723
+    },
+    {
+      "epoch": 0.091584,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.157580645161292e-05,
+      "loss": 0.1546,
+      "step": 5724
+    },
+    {
+      "epoch": 0.0916,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.15741935483871e-05,
+      "loss": 0.1701,
+      "step": 5725
+    },
+    {
+      "epoch": 0.091616,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.15725806451613e-05,
+      "loss": 0.1685,
+      "step": 5726
+    },
+    {
+      "epoch": 0.091632,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.157096774193549e-05,
+      "loss": 0.1271,
+      "step": 5727
+    },
+    {
+      "epoch": 0.091648,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.156935483870969e-05,
+      "loss": 0.1506,
+      "step": 5728
+    },
+    {
+      "epoch": 0.091664,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.156774193548387e-05,
+      "loss": 0.1783,
+      "step": 5729
+    },
+    {
+      "epoch": 0.09168,
+      "grad_norm": 0.875,
+      "learning_rate": 9.156612903225806e-05,
+      "loss": 0.1977,
+      "step": 5730
+    },
+    {
+      "epoch": 0.091696,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.156451612903226e-05,
+      "loss": 0.1775,
+      "step": 5731
+    },
+    {
+      "epoch": 0.091712,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.156290322580645e-05,
+      "loss": 0.1947,
+      "step": 5732
+    },
+    {
+      "epoch": 0.091728,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.156129032258065e-05,
+      "loss": 0.1758,
+      "step": 5733
+    },
+    {
+      "epoch": 0.091744,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.155967741935485e-05,
+      "loss": 0.1566,
+      "step": 5734
+    },
+    {
+      "epoch": 0.09176,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.155806451612905e-05,
+      "loss": 0.1346,
+      "step": 5735
+    },
+    {
+      "epoch": 0.091776,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.155645161290323e-05,
+      "loss": 0.206,
+      "step": 5736
+    },
+    {
+      "epoch": 0.091792,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.155483870967743e-05,
+      "loss": 0.1412,
+      "step": 5737
+    },
+    {
+      "epoch": 0.091808,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.155322580645162e-05,
+      "loss": 0.1305,
+      "step": 5738
+    },
+    {
+      "epoch": 0.091824,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.155161290322582e-05,
+      "loss": 0.1422,
+      "step": 5739
+    },
+    {
+      "epoch": 0.09184,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.155e-05,
+      "loss": 0.1442,
+      "step": 5740
+    },
+    {
+      "epoch": 0.091856,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.15483870967742e-05,
+      "loss": 0.2071,
+      "step": 5741
+    },
+    {
+      "epoch": 0.091872,
+      "grad_norm": 1.4609375,
+      "learning_rate": 9.154677419354839e-05,
+      "loss": 0.2004,
+      "step": 5742
+    },
+    {
+      "epoch": 0.091888,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.154516129032259e-05,
+      "loss": 0.1553,
+      "step": 5743
+    },
+    {
+      "epoch": 0.091904,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.154354838709677e-05,
+      "loss": 0.1553,
+      "step": 5744
+    },
+    {
+      "epoch": 0.09192,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.154193548387097e-05,
+      "loss": 0.1753,
+      "step": 5745
+    },
+    {
+      "epoch": 0.091936,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.154032258064516e-05,
+      "loss": 0.1735,
+      "step": 5746
+    },
+    {
+      "epoch": 0.091952,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.153870967741936e-05,
+      "loss": 0.1672,
+      "step": 5747
+    },
+    {
+      "epoch": 0.091968,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.153709677419356e-05,
+      "loss": 0.2162,
+      "step": 5748
+    },
+    {
+      "epoch": 0.091984,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.153548387096775e-05,
+      "loss": 0.1803,
+      "step": 5749
+    },
+    {
+      "epoch": 0.092,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.153387096774194e-05,
+      "loss": 0.1984,
+      "step": 5750
+    },
+    {
+      "epoch": 0.092016,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.153225806451613e-05,
+      "loss": 0.1311,
+      "step": 5751
+    },
+    {
+      "epoch": 0.092032,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.153064516129033e-05,
+      "loss": 0.196,
+      "step": 5752
+    },
+    {
+      "epoch": 0.092048,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.152903225806452e-05,
+      "loss": 0.154,
+      "step": 5753
+    },
+    {
+      "epoch": 0.092064,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.152741935483872e-05,
+      "loss": 0.1583,
+      "step": 5754
+    },
+    {
+      "epoch": 0.09208,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.15258064516129e-05,
+      "loss": 0.1939,
+      "step": 5755
+    },
+    {
+      "epoch": 0.092096,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.15241935483871e-05,
+      "loss": 0.1692,
+      "step": 5756
+    },
+    {
+      "epoch": 0.092112,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.152258064516129e-05,
+      "loss": 0.1773,
+      "step": 5757
+    },
+    {
+      "epoch": 0.092128,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.152096774193549e-05,
+      "loss": 0.1225,
+      "step": 5758
+    },
+    {
+      "epoch": 0.092144,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.151935483870969e-05,
+      "loss": 0.1958,
+      "step": 5759
+    },
+    {
+      "epoch": 0.09216,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.151774193548389e-05,
+      "loss": 0.1951,
+      "step": 5760
+    },
+    {
+      "epoch": 0.092176,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.151612903225807e-05,
+      "loss": 0.1883,
+      "step": 5761
+    },
+    {
+      "epoch": 0.092192,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.151451612903226e-05,
+      "loss": 0.2053,
+      "step": 5762
+    },
+    {
+      "epoch": 0.092208,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.151290322580646e-05,
+      "loss": 0.2196,
+      "step": 5763
+    },
+    {
+      "epoch": 0.092224,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.151129032258064e-05,
+      "loss": 0.1895,
+      "step": 5764
+    },
+    {
+      "epoch": 0.09224,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.150967741935484e-05,
+      "loss": 0.1779,
+      "step": 5765
+    },
+    {
+      "epoch": 0.092256,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.150806451612903e-05,
+      "loss": 0.1775,
+      "step": 5766
+    },
+    {
+      "epoch": 0.092272,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.150645161290323e-05,
+      "loss": 0.1665,
+      "step": 5767
+    },
+    {
+      "epoch": 0.092288,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.150483870967742e-05,
+      "loss": 0.1697,
+      "step": 5768
+    },
+    {
+      "epoch": 0.092304,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.150322580645162e-05,
+      "loss": 0.1683,
+      "step": 5769
+    },
+    {
+      "epoch": 0.09232,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.15016129032258e-05,
+      "loss": 0.163,
+      "step": 5770
+    },
+    {
+      "epoch": 0.092336,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.15e-05,
+      "loss": 0.1901,
+      "step": 5771
+    },
+    {
+      "epoch": 0.092352,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.14983870967742e-05,
+      "loss": 0.1754,
+      "step": 5772
+    },
+    {
+      "epoch": 0.092368,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.14967741935484e-05,
+      "loss": 0.1404,
+      "step": 5773
+    },
+    {
+      "epoch": 0.092384,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.149516129032259e-05,
+      "loss": 0.1787,
+      "step": 5774
+    },
+    {
+      "epoch": 0.0924,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.149354838709679e-05,
+      "loss": 0.1834,
+      "step": 5775
+    },
+    {
+      "epoch": 0.092416,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.149193548387097e-05,
+      "loss": 0.1689,
+      "step": 5776
+    },
+    {
+      "epoch": 0.092432,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.149032258064516e-05,
+      "loss": 0.1832,
+      "step": 5777
+    },
+    {
+      "epoch": 0.092448,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.148870967741936e-05,
+      "loss": 0.2085,
+      "step": 5778
+    },
+    {
+      "epoch": 0.092464,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.148709677419354e-05,
+      "loss": 0.1702,
+      "step": 5779
+    },
+    {
+      "epoch": 0.09248,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.148548387096774e-05,
+      "loss": 0.1924,
+      "step": 5780
+    },
+    {
+      "epoch": 0.092496,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.148387096774193e-05,
+      "loss": 0.1572,
+      "step": 5781
+    },
+    {
+      "epoch": 0.092512,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.148225806451613e-05,
+      "loss": 0.1598,
+      "step": 5782
+    },
+    {
+      "epoch": 0.092528,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.148064516129033e-05,
+      "loss": 0.2084,
+      "step": 5783
+    },
+    {
+      "epoch": 0.092544,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.147903225806453e-05,
+      "loss": 0.1887,
+      "step": 5784
+    },
+    {
+      "epoch": 0.09256,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.147741935483871e-05,
+      "loss": 0.1904,
+      "step": 5785
+    },
+    {
+      "epoch": 0.092576,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.147580645161291e-05,
+      "loss": 0.1895,
+      "step": 5786
+    },
+    {
+      "epoch": 0.092592,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.14741935483871e-05,
+      "loss": 0.1906,
+      "step": 5787
+    },
+    {
+      "epoch": 0.092608,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.14725806451613e-05,
+      "loss": 0.192,
+      "step": 5788
+    },
+    {
+      "epoch": 0.092624,
+      "grad_norm": 1.0,
+      "learning_rate": 9.147096774193549e-05,
+      "loss": 0.2351,
+      "step": 5789
+    },
+    {
+      "epoch": 0.09264,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.146935483870969e-05,
+      "loss": 0.1616,
+      "step": 5790
+    },
+    {
+      "epoch": 0.092656,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.146774193548387e-05,
+      "loss": 0.1896,
+      "step": 5791
+    },
+    {
+      "epoch": 0.092672,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.146612903225807e-05,
+      "loss": 0.1989,
+      "step": 5792
+    },
+    {
+      "epoch": 0.092688,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.146451612903226e-05,
+      "loss": 0.1767,
+      "step": 5793
+    },
+    {
+      "epoch": 0.092704,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.146290322580646e-05,
+      "loss": 0.1498,
+      "step": 5794
+    },
+    {
+      "epoch": 0.09272,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.146129032258066e-05,
+      "loss": 0.1912,
+      "step": 5795
+    },
+    {
+      "epoch": 0.092736,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.145967741935484e-05,
+      "loss": 0.1939,
+      "step": 5796
+    },
+    {
+      "epoch": 0.092752,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.145806451612904e-05,
+      "loss": 0.1506,
+      "step": 5797
+    },
+    {
+      "epoch": 0.092768,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.145645161290323e-05,
+      "loss": 0.1608,
+      "step": 5798
+    },
+    {
+      "epoch": 0.092784,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.145483870967743e-05,
+      "loss": 0.1524,
+      "step": 5799
+    },
+    {
+      "epoch": 0.0928,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.145322580645161e-05,
+      "loss": 0.2551,
+      "step": 5800
+    },
+    {
+      "epoch": 0.092816,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.145161290322581e-05,
+      "loss": 0.2027,
+      "step": 5801
+    },
+    {
+      "epoch": 0.092832,
+      "grad_norm": 1.4609375,
+      "learning_rate": 9.145e-05,
+      "loss": 0.1792,
+      "step": 5802
+    },
+    {
+      "epoch": 0.092848,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.14483870967742e-05,
+      "loss": 0.1511,
+      "step": 5803
+    },
+    {
+      "epoch": 0.092864,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.144677419354839e-05,
+      "loss": 0.1474,
+      "step": 5804
+    },
+    {
+      "epoch": 0.09288,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.144516129032259e-05,
+      "loss": 0.1943,
+      "step": 5805
+    },
+    {
+      "epoch": 0.092896,
+      "grad_norm": 0.75,
+      "learning_rate": 9.144354838709677e-05,
+      "loss": 0.1792,
+      "step": 5806
+    },
+    {
+      "epoch": 0.092912,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.144193548387097e-05,
+      "loss": 0.166,
+      "step": 5807
+    },
+    {
+      "epoch": 0.092928,
+      "grad_norm": 0.47265625,
+      "learning_rate": 9.144032258064517e-05,
+      "loss": 0.1475,
+      "step": 5808
+    },
+    {
+      "epoch": 0.092944,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.143870967741936e-05,
+      "loss": 0.195,
+      "step": 5809
+    },
+    {
+      "epoch": 0.09296,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.143709677419356e-05,
+      "loss": 0.1424,
+      "step": 5810
+    },
+    {
+      "epoch": 0.092976,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.143548387096774e-05,
+      "loss": 0.1964,
+      "step": 5811
+    },
+    {
+      "epoch": 0.092992,
+      "grad_norm": 1.296875,
+      "learning_rate": 9.143387096774194e-05,
+      "loss": 0.2049,
+      "step": 5812
+    },
+    {
+      "epoch": 0.093008,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.143225806451613e-05,
+      "loss": 0.1622,
+      "step": 5813
+    },
+    {
+      "epoch": 0.093024,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.143064516129033e-05,
+      "loss": 0.1983,
+      "step": 5814
+    },
+    {
+      "epoch": 0.09304,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.142903225806451e-05,
+      "loss": 0.1932,
+      "step": 5815
+    },
+    {
+      "epoch": 0.093056,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.142741935483871e-05,
+      "loss": 0.1938,
+      "step": 5816
+    },
+    {
+      "epoch": 0.093072,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.14258064516129e-05,
+      "loss": 0.1825,
+      "step": 5817
+    },
+    {
+      "epoch": 0.093088,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.14241935483871e-05,
+      "loss": 0.2051,
+      "step": 5818
+    },
+    {
+      "epoch": 0.093104,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.14225806451613e-05,
+      "loss": 0.2102,
+      "step": 5819
+    },
+    {
+      "epoch": 0.09312,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.14209677419355e-05,
+      "loss": 0.1582,
+      "step": 5820
+    },
+    {
+      "epoch": 0.093136,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.141935483870968e-05,
+      "loss": 0.187,
+      "step": 5821
+    },
+    {
+      "epoch": 0.093152,
+      "grad_norm": 0.625,
+      "learning_rate": 9.141774193548388e-05,
+      "loss": 0.1689,
+      "step": 5822
+    },
+    {
+      "epoch": 0.093168,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.141612903225807e-05,
+      "loss": 0.1232,
+      "step": 5823
+    },
+    {
+      "epoch": 0.093184,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.141451612903226e-05,
+      "loss": 0.2064,
+      "step": 5824
+    },
+    {
+      "epoch": 0.0932,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.141290322580646e-05,
+      "loss": 0.166,
+      "step": 5825
+    },
+    {
+      "epoch": 0.093216,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.141129032258064e-05,
+      "loss": 0.1734,
+      "step": 5826
+    },
+    {
+      "epoch": 0.093232,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.140967741935484e-05,
+      "loss": 0.1753,
+      "step": 5827
+    },
+    {
+      "epoch": 0.093248,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.140806451612903e-05,
+      "loss": 0.1705,
+      "step": 5828
+    },
+    {
+      "epoch": 0.093264,
+      "grad_norm": 1.359375,
+      "learning_rate": 9.140645161290323e-05,
+      "loss": 0.1985,
+      "step": 5829
+    },
+    {
+      "epoch": 0.09328,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.140483870967743e-05,
+      "loss": 0.1703,
+      "step": 5830
+    },
+    {
+      "epoch": 0.093296,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.140322580645161e-05,
+      "loss": 0.1549,
+      "step": 5831
+    },
+    {
+      "epoch": 0.093312,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.140161290322581e-05,
+      "loss": 0.1629,
+      "step": 5832
+    },
+    {
+      "epoch": 0.093328,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.140000000000001e-05,
+      "loss": 0.1946,
+      "step": 5833
+    },
+    {
+      "epoch": 0.093344,
+      "grad_norm": 1.4375,
+      "learning_rate": 9.13983870967742e-05,
+      "loss": 0.1786,
+      "step": 5834
+    },
+    {
+      "epoch": 0.09336,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.13967741935484e-05,
+      "loss": 0.1773,
+      "step": 5835
+    },
+    {
+      "epoch": 0.093376,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.139516129032258e-05,
+      "loss": 0.1875,
+      "step": 5836
+    },
+    {
+      "epoch": 0.093392,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.139354838709678e-05,
+      "loss": 0.1802,
+      "step": 5837
+    },
+    {
+      "epoch": 0.093408,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.139193548387097e-05,
+      "loss": 0.1592,
+      "step": 5838
+    },
+    {
+      "epoch": 0.093424,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.139032258064516e-05,
+      "loss": 0.166,
+      "step": 5839
+    },
+    {
+      "epoch": 0.09344,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.138870967741936e-05,
+      "loss": 0.1729,
+      "step": 5840
+    },
+    {
+      "epoch": 0.093456,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.138709677419354e-05,
+      "loss": 0.1734,
+      "step": 5841
+    },
+    {
+      "epoch": 0.093472,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.138548387096774e-05,
+      "loss": 0.2272,
+      "step": 5842
+    },
+    {
+      "epoch": 0.093488,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.138387096774194e-05,
+      "loss": 0.1796,
+      "step": 5843
+    },
+    {
+      "epoch": 0.093504,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.138225806451614e-05,
+      "loss": 0.1701,
+      "step": 5844
+    },
+    {
+      "epoch": 0.09352,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.138064516129033e-05,
+      "loss": 0.1901,
+      "step": 5845
+    },
+    {
+      "epoch": 0.093536,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.137903225806453e-05,
+      "loss": 0.1932,
+      "step": 5846
+    },
+    {
+      "epoch": 0.093552,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.137741935483871e-05,
+      "loss": 0.1905,
+      "step": 5847
+    },
+    {
+      "epoch": 0.093568,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.137580645161291e-05,
+      "loss": 0.172,
+      "step": 5848
+    },
+    {
+      "epoch": 0.093584,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.13741935483871e-05,
+      "loss": 0.1632,
+      "step": 5849
+    },
+    {
+      "epoch": 0.0936,
+      "grad_norm": 0.75,
+      "learning_rate": 9.13725806451613e-05,
+      "loss": 0.1636,
+      "step": 5850
+    },
+    {
+      "epoch": 0.093616,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.137096774193548e-05,
+      "loss": 0.157,
+      "step": 5851
+    },
+    {
+      "epoch": 0.093632,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.136935483870968e-05,
+      "loss": 0.1808,
+      "step": 5852
+    },
+    {
+      "epoch": 0.093648,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.136774193548387e-05,
+      "loss": 0.137,
+      "step": 5853
+    },
+    {
+      "epoch": 0.093664,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.136612903225807e-05,
+      "loss": 0.2035,
+      "step": 5854
+    },
+    {
+      "epoch": 0.09368,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.136451612903227e-05,
+      "loss": 0.1339,
+      "step": 5855
+    },
+    {
+      "epoch": 0.093696,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.136290322580645e-05,
+      "loss": 0.1525,
+      "step": 5856
+    },
+    {
+      "epoch": 0.093712,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.136129032258065e-05,
+      "loss": 0.151,
+      "step": 5857
+    },
+    {
+      "epoch": 0.093728,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.135967741935484e-05,
+      "loss": 0.1938,
+      "step": 5858
+    },
+    {
+      "epoch": 0.093744,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.135806451612904e-05,
+      "loss": 0.1756,
+      "step": 5859
+    },
+    {
+      "epoch": 0.09376,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.135645161290323e-05,
+      "loss": 0.2047,
+      "step": 5860
+    },
+    {
+      "epoch": 0.093776,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.135483870967743e-05,
+      "loss": 0.17,
+      "step": 5861
+    },
+    {
+      "epoch": 0.093792,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.135322580645161e-05,
+      "loss": 0.2126,
+      "step": 5862
+    },
+    {
+      "epoch": 0.093808,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.135161290322581e-05,
+      "loss": 0.1792,
+      "step": 5863
+    },
+    {
+      "epoch": 0.093824,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.135e-05,
+      "loss": 0.2191,
+      "step": 5864
+    },
+    {
+      "epoch": 0.09384,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.13483870967742e-05,
+      "loss": 0.2126,
+      "step": 5865
+    },
+    {
+      "epoch": 0.093856,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.134677419354838e-05,
+      "loss": 0.1626,
+      "step": 5866
+    },
+    {
+      "epoch": 0.093872,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.134516129032258e-05,
+      "loss": 0.1956,
+      "step": 5867
+    },
+    {
+      "epoch": 0.093888,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.134354838709678e-05,
+      "loss": 0.1798,
+      "step": 5868
+    },
+    {
+      "epoch": 0.093904,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.134193548387098e-05,
+      "loss": 0.1732,
+      "step": 5869
+    },
+    {
+      "epoch": 0.09392,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.134032258064517e-05,
+      "loss": 0.1797,
+      "step": 5870
+    },
+    {
+      "epoch": 0.093936,
+      "grad_norm": 1.453125,
+      "learning_rate": 9.133870967741935e-05,
+      "loss": 0.1924,
+      "step": 5871
+    },
+    {
+      "epoch": 0.093952,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.133709677419355e-05,
+      "loss": 0.2293,
+      "step": 5872
+    },
+    {
+      "epoch": 0.093968,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.133548387096774e-05,
+      "loss": 0.1805,
+      "step": 5873
+    },
+    {
+      "epoch": 0.093984,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.133387096774194e-05,
+      "loss": 0.1551,
+      "step": 5874
+    },
+    {
+      "epoch": 0.094,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.133225806451613e-05,
+      "loss": 0.1952,
+      "step": 5875
+    },
+    {
+      "epoch": 0.094016,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.133064516129033e-05,
+      "loss": 0.1714,
+      "step": 5876
+    },
+    {
+      "epoch": 0.094032,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.132903225806451e-05,
+      "loss": 0.1911,
+      "step": 5877
+    },
+    {
+      "epoch": 0.094048,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.132741935483871e-05,
+      "loss": 0.1773,
+      "step": 5878
+    },
+    {
+      "epoch": 0.094064,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.132580645161291e-05,
+      "loss": 0.2192,
+      "step": 5879
+    },
+    {
+      "epoch": 0.09408,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.132419354838711e-05,
+      "loss": 0.1683,
+      "step": 5880
+    },
+    {
+      "epoch": 0.094096,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.13225806451613e-05,
+      "loss": 0.1776,
+      "step": 5881
+    },
+    {
+      "epoch": 0.094112,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.13209677419355e-05,
+      "loss": 0.1878,
+      "step": 5882
+    },
+    {
+      "epoch": 0.094128,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.131935483870968e-05,
+      "loss": 0.1975,
+      "step": 5883
+    },
+    {
+      "epoch": 0.094144,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.131774193548388e-05,
+      "loss": 0.1769,
+      "step": 5884
+    },
+    {
+      "epoch": 0.09416,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.131612903225807e-05,
+      "loss": 0.1798,
+      "step": 5885
+    },
+    {
+      "epoch": 0.094176,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.131451612903225e-05,
+      "loss": 0.1307,
+      "step": 5886
+    },
+    {
+      "epoch": 0.094192,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.131290322580645e-05,
+      "loss": 0.2111,
+      "step": 5887
+    },
+    {
+      "epoch": 0.094208,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.131129032258064e-05,
+      "loss": 0.1932,
+      "step": 5888
+    },
+    {
+      "epoch": 0.094224,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.130967741935484e-05,
+      "loss": 0.1941,
+      "step": 5889
+    },
+    {
+      "epoch": 0.09424,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.130806451612904e-05,
+      "loss": 0.2201,
+      "step": 5890
+    },
+    {
+      "epoch": 0.094256,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.130645161290324e-05,
+      "loss": 0.1447,
+      "step": 5891
+    },
+    {
+      "epoch": 0.094272,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.130483870967742e-05,
+      "loss": 0.2032,
+      "step": 5892
+    },
+    {
+      "epoch": 0.094288,
+      "grad_norm": 1.390625,
+      "learning_rate": 9.130322580645162e-05,
+      "loss": 0.2179,
+      "step": 5893
+    },
+    {
+      "epoch": 0.094304,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.130161290322581e-05,
+      "loss": 0.1786,
+      "step": 5894
+    },
+    {
+      "epoch": 0.09432,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.130000000000001e-05,
+      "loss": 0.1744,
+      "step": 5895
+    },
+    {
+      "epoch": 0.094336,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.12983870967742e-05,
+      "loss": 0.1663,
+      "step": 5896
+    },
+    {
+      "epoch": 0.094352,
+      "grad_norm": 1.40625,
+      "learning_rate": 9.12967741935484e-05,
+      "loss": 0.1432,
+      "step": 5897
+    },
+    {
+      "epoch": 0.094368,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.129516129032258e-05,
+      "loss": 0.1337,
+      "step": 5898
+    },
+    {
+      "epoch": 0.094384,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.129354838709678e-05,
+      "loss": 0.158,
+      "step": 5899
+    },
+    {
+      "epoch": 0.0944,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.129193548387097e-05,
+      "loss": 0.1771,
+      "step": 5900
+    },
+    {
+      "epoch": 0.094416,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.129032258064517e-05,
+      "loss": 0.1943,
+      "step": 5901
+    },
+    {
+      "epoch": 0.094432,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.128870967741935e-05,
+      "loss": 0.1743,
+      "step": 5902
+    },
+    {
+      "epoch": 0.094448,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.128709677419355e-05,
+      "loss": 0.2073,
+      "step": 5903
+    },
+    {
+      "epoch": 0.094464,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.128548387096775e-05,
+      "loss": 0.1819,
+      "step": 5904
+    },
+    {
+      "epoch": 0.09448,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.128387096774194e-05,
+      "loss": 0.2371,
+      "step": 5905
+    },
+    {
+      "epoch": 0.094496,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.128225806451614e-05,
+      "loss": 0.2188,
+      "step": 5906
+    },
+    {
+      "epoch": 0.094512,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.128064516129032e-05,
+      "loss": 0.1668,
+      "step": 5907
+    },
+    {
+      "epoch": 0.094528,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.127903225806452e-05,
+      "loss": 0.2214,
+      "step": 5908
+    },
+    {
+      "epoch": 0.094544,
+      "grad_norm": 0.51171875,
+      "learning_rate": 9.127741935483871e-05,
+      "loss": 0.1372,
+      "step": 5909
+    },
+    {
+      "epoch": 0.09456,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.127580645161291e-05,
+      "loss": 0.1736,
+      "step": 5910
+    },
+    {
+      "epoch": 0.094576,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.12741935483871e-05,
+      "loss": 0.1607,
+      "step": 5911
+    },
+    {
+      "epoch": 0.094592,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.12725806451613e-05,
+      "loss": 0.1699,
+      "step": 5912
+    },
+    {
+      "epoch": 0.094608,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.127096774193548e-05,
+      "loss": 0.171,
+      "step": 5913
+    },
+    {
+      "epoch": 0.094624,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.126935483870968e-05,
+      "loss": 0.1725,
+      "step": 5914
+    },
+    {
+      "epoch": 0.09464,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.126774193548388e-05,
+      "loss": 0.1719,
+      "step": 5915
+    },
+    {
+      "epoch": 0.094656,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.126612903225808e-05,
+      "loss": 0.2169,
+      "step": 5916
+    },
+    {
+      "epoch": 0.094672,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.126451612903227e-05,
+      "loss": 0.1609,
+      "step": 5917
+    },
+    {
+      "epoch": 0.094688,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.126290322580645e-05,
+      "loss": 0.2024,
+      "step": 5918
+    },
+    {
+      "epoch": 0.094704,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.126129032258065e-05,
+      "loss": 0.1458,
+      "step": 5919
+    },
+    {
+      "epoch": 0.09472,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.125967741935484e-05,
+      "loss": 0.1611,
+      "step": 5920
+    },
+    {
+      "epoch": 0.094736,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.125806451612904e-05,
+      "loss": 0.1306,
+      "step": 5921
+    },
+    {
+      "epoch": 0.094752,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.125645161290322e-05,
+      "loss": 0.1588,
+      "step": 5922
+    },
+    {
+      "epoch": 0.094768,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.125483870967742e-05,
+      "loss": 0.2075,
+      "step": 5923
+    },
+    {
+      "epoch": 0.094784,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.125322580645161e-05,
+      "loss": 0.1838,
+      "step": 5924
+    },
+    {
+      "epoch": 0.0948,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.125161290322581e-05,
+      "loss": 0.1789,
+      "step": 5925
+    },
+    {
+      "epoch": 0.094816,
+      "grad_norm": 1.9296875,
+      "learning_rate": 9.125e-05,
+      "loss": 0.1564,
+      "step": 5926
+    },
+    {
+      "epoch": 0.094832,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.12483870967742e-05,
+      "loss": 0.1842,
+      "step": 5927
+    },
+    {
+      "epoch": 0.094848,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.12467741935484e-05,
+      "loss": 0.1695,
+      "step": 5928
+    },
+    {
+      "epoch": 0.094864,
+      "grad_norm": 1.421875,
+      "learning_rate": 9.12451612903226e-05,
+      "loss": 0.1796,
+      "step": 5929
+    },
+    {
+      "epoch": 0.09488,
+      "grad_norm": 1.3125,
+      "learning_rate": 9.124354838709678e-05,
+      "loss": 0.1781,
+      "step": 5930
+    },
+    {
+      "epoch": 0.094896,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.124193548387098e-05,
+      "loss": 0.1913,
+      "step": 5931
+    },
+    {
+      "epoch": 0.094912,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.124032258064517e-05,
+      "loss": 0.1324,
+      "step": 5932
+    },
+    {
+      "epoch": 0.094928,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.123870967741935e-05,
+      "loss": 0.1607,
+      "step": 5933
+    },
+    {
+      "epoch": 0.094944,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.123709677419355e-05,
+      "loss": 0.2044,
+      "step": 5934
+    },
+    {
+      "epoch": 0.09496,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.123548387096774e-05,
+      "loss": 0.1731,
+      "step": 5935
+    },
+    {
+      "epoch": 0.094976,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.123387096774194e-05,
+      "loss": 0.1831,
+      "step": 5936
+    },
+    {
+      "epoch": 0.094992,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.123225806451612e-05,
+      "loss": 0.1614,
+      "step": 5937
+    },
+    {
+      "epoch": 0.095008,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.123064516129032e-05,
+      "loss": 0.1792,
+      "step": 5938
+    },
+    {
+      "epoch": 0.095024,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.122903225806452e-05,
+      "loss": 0.1973,
+      "step": 5939
+    },
+    {
+      "epoch": 0.09504,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.122741935483872e-05,
+      "loss": 0.1576,
+      "step": 5940
+    },
+    {
+      "epoch": 0.095056,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.122580645161291e-05,
+      "loss": 0.1701,
+      "step": 5941
+    },
+    {
+      "epoch": 0.095072,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.122419354838711e-05,
+      "loss": 0.1383,
+      "step": 5942
+    },
+    {
+      "epoch": 0.095088,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.12225806451613e-05,
+      "loss": 0.1808,
+      "step": 5943
+    },
+    {
+      "epoch": 0.095104,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.12209677419355e-05,
+      "loss": 0.1895,
+      "step": 5944
+    },
+    {
+      "epoch": 0.09512,
+      "grad_norm": 0.875,
+      "learning_rate": 9.121935483870968e-05,
+      "loss": 0.1863,
+      "step": 5945
+    },
+    {
+      "epoch": 0.095136,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.121774193548388e-05,
+      "loss": 0.2083,
+      "step": 5946
+    },
+    {
+      "epoch": 0.095152,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.121612903225807e-05,
+      "loss": 0.1909,
+      "step": 5947
+    },
+    {
+      "epoch": 0.095168,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.121451612903225e-05,
+      "loss": 0.1691,
+      "step": 5948
+    },
+    {
+      "epoch": 0.095184,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.121290322580645e-05,
+      "loss": 0.1883,
+      "step": 5949
+    },
+    {
+      "epoch": 0.0952,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.121129032258065e-05,
+      "loss": 0.1614,
+      "step": 5950
+    },
+    {
+      "epoch": 0.095216,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.120967741935485e-05,
+      "loss": 0.1968,
+      "step": 5951
+    },
+    {
+      "epoch": 0.095232,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.120806451612904e-05,
+      "loss": 0.1776,
+      "step": 5952
+    },
+    {
+      "epoch": 0.095248,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.120645161290324e-05,
+      "loss": 0.1824,
+      "step": 5953
+    },
+    {
+      "epoch": 0.095264,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.120483870967742e-05,
+      "loss": 0.2348,
+      "step": 5954
+    },
+    {
+      "epoch": 0.09528,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.120322580645162e-05,
+      "loss": 0.1319,
+      "step": 5955
+    },
+    {
+      "epoch": 0.095296,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.120161290322581e-05,
+      "loss": 0.1926,
+      "step": 5956
+    },
+    {
+      "epoch": 0.095312,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.120000000000001e-05,
+      "loss": 0.1808,
+      "step": 5957
+    },
+    {
+      "epoch": 0.095328,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.11983870967742e-05,
+      "loss": 0.1971,
+      "step": 5958
+    },
+    {
+      "epoch": 0.095344,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.11967741935484e-05,
+      "loss": 0.1895,
+      "step": 5959
+    },
+    {
+      "epoch": 0.09536,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.119516129032258e-05,
+      "loss": 0.1942,
+      "step": 5960
+    },
+    {
+      "epoch": 0.095376,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.119354838709678e-05,
+      "loss": 0.2248,
+      "step": 5961
+    },
+    {
+      "epoch": 0.095392,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.119193548387097e-05,
+      "loss": 0.1644,
+      "step": 5962
+    },
+    {
+      "epoch": 0.095408,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.119032258064516e-05,
+      "loss": 0.1922,
+      "step": 5963
+    },
+    {
+      "epoch": 0.095424,
+      "grad_norm": 1.3515625,
+      "learning_rate": 9.118870967741936e-05,
+      "loss": 0.2122,
+      "step": 5964
+    },
+    {
+      "epoch": 0.09544,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.118709677419355e-05,
+      "loss": 0.1476,
+      "step": 5965
+    },
+    {
+      "epoch": 0.095456,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.118548387096775e-05,
+      "loss": 0.1991,
+      "step": 5966
+    },
+    {
+      "epoch": 0.095472,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.118387096774194e-05,
+      "loss": 0.139,
+      "step": 5967
+    },
+    {
+      "epoch": 0.095488,
+      "grad_norm": 1.5234375,
+      "learning_rate": 9.118225806451614e-05,
+      "loss": 0.1764,
+      "step": 5968
+    },
+    {
+      "epoch": 0.095504,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.118064516129032e-05,
+      "loss": 0.2027,
+      "step": 5969
+    },
+    {
+      "epoch": 0.09552,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.117903225806452e-05,
+      "loss": 0.1622,
+      "step": 5970
+    },
+    {
+      "epoch": 0.095536,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.117741935483871e-05,
+      "loss": 0.1828,
+      "step": 5971
+    },
+    {
+      "epoch": 0.095552,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.117580645161291e-05,
+      "loss": 0.2173,
+      "step": 5972
+    },
+    {
+      "epoch": 0.095568,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.11741935483871e-05,
+      "loss": 0.1431,
+      "step": 5973
+    },
+    {
+      "epoch": 0.095584,
+      "grad_norm": 0.75,
+      "learning_rate": 9.117258064516129e-05,
+      "loss": 0.1713,
+      "step": 5974
+    },
+    {
+      "epoch": 0.0956,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.117096774193549e-05,
+      "loss": 0.1704,
+      "step": 5975
+    },
+    {
+      "epoch": 0.095616,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.116935483870969e-05,
+      "loss": 0.1544,
+      "step": 5976
+    },
+    {
+      "epoch": 0.095632,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.116774193548388e-05,
+      "loss": 0.2166,
+      "step": 5977
+    },
+    {
+      "epoch": 0.095648,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.116612903225808e-05,
+      "loss": 0.1913,
+      "step": 5978
+    },
+    {
+      "epoch": 0.095664,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.116451612903226e-05,
+      "loss": 0.1979,
+      "step": 5979
+    },
+    {
+      "epoch": 0.09568,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.116290322580645e-05,
+      "loss": 0.1918,
+      "step": 5980
+    },
+    {
+      "epoch": 0.095696,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.116129032258065e-05,
+      "loss": 0.143,
+      "step": 5981
+    },
+    {
+      "epoch": 0.095712,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.115967741935484e-05,
+      "loss": 0.214,
+      "step": 5982
+    },
+    {
+      "epoch": 0.095728,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.115806451612904e-05,
+      "loss": 0.2083,
+      "step": 5983
+    },
+    {
+      "epoch": 0.095744,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.115645161290322e-05,
+      "loss": 0.16,
+      "step": 5984
+    },
+    {
+      "epoch": 0.09576,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.115483870967742e-05,
+      "loss": 0.1742,
+      "step": 5985
+    },
+    {
+      "epoch": 0.095776,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.115322580645162e-05,
+      "loss": 0.1614,
+      "step": 5986
+    },
+    {
+      "epoch": 0.095792,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.115161290322582e-05,
+      "loss": 0.17,
+      "step": 5987
+    },
+    {
+      "epoch": 0.095808,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.115e-05,
+      "loss": 0.1548,
+      "step": 5988
+    },
+    {
+      "epoch": 0.095824,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.11483870967742e-05,
+      "loss": 0.1787,
+      "step": 5989
+    },
+    {
+      "epoch": 0.09584,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.114677419354839e-05,
+      "loss": 0.1604,
+      "step": 5990
+    },
+    {
+      "epoch": 0.095856,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.114516129032259e-05,
+      "loss": 0.1393,
+      "step": 5991
+    },
+    {
+      "epoch": 0.095872,
+      "grad_norm": 0.5,
+      "learning_rate": 9.114354838709678e-05,
+      "loss": 0.1708,
+      "step": 5992
+    },
+    {
+      "epoch": 0.095888,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.114193548387098e-05,
+      "loss": 0.1768,
+      "step": 5993
+    },
+    {
+      "epoch": 0.095904,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.114032258064516e-05,
+      "loss": 0.1449,
+      "step": 5994
+    },
+    {
+      "epoch": 0.09592,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.113870967741935e-05,
+      "loss": 0.182,
+      "step": 5995
+    },
+    {
+      "epoch": 0.095936,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.113709677419355e-05,
+      "loss": 0.1778,
+      "step": 5996
+    },
+    {
+      "epoch": 0.095952,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.113548387096774e-05,
+      "loss": 0.1851,
+      "step": 5997
+    },
+    {
+      "epoch": 0.095968,
+      "grad_norm": 1.4296875,
+      "learning_rate": 9.113387096774194e-05,
+      "loss": 0.1592,
+      "step": 5998
+    },
+    {
+      "epoch": 0.095984,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.113225806451613e-05,
+      "loss": 0.1888,
+      "step": 5999
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.113064516129033e-05,
+      "loss": 0.1546,
+      "step": 6000
+    },
+    {
+      "epoch": 0.096016,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.112903225806452e-05,
+      "loss": 0.1908,
+      "step": 6001
+    },
+    {
+      "epoch": 0.096032,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.112741935483872e-05,
+      "loss": 0.2097,
+      "step": 6002
+    },
+    {
+      "epoch": 0.096048,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.11258064516129e-05,
+      "loss": 0.176,
+      "step": 6003
+    },
+    {
+      "epoch": 0.096064,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.11241935483871e-05,
+      "loss": 0.1585,
+      "step": 6004
+    },
+    {
+      "epoch": 0.09608,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.112258064516129e-05,
+      "loss": 0.138,
+      "step": 6005
+    },
+    {
+      "epoch": 0.096096,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.112096774193549e-05,
+      "loss": 0.2027,
+      "step": 6006
+    },
+    {
+      "epoch": 0.096112,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.111935483870968e-05,
+      "loss": 0.1561,
+      "step": 6007
+    },
+    {
+      "epoch": 0.096128,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.111774193548388e-05,
+      "loss": 0.1582,
+      "step": 6008
+    },
+    {
+      "epoch": 0.096144,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.111612903225806e-05,
+      "loss": 0.1837,
+      "step": 6009
+    },
+    {
+      "epoch": 0.09616,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.111451612903226e-05,
+      "loss": 0.1422,
+      "step": 6010
+    },
+    {
+      "epoch": 0.096176,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.111290322580646e-05,
+      "loss": 0.1781,
+      "step": 6011
+    },
+    {
+      "epoch": 0.096192,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.111129032258065e-05,
+      "loss": 0.1318,
+      "step": 6012
+    },
+    {
+      "epoch": 0.096208,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.110967741935485e-05,
+      "loss": 0.1785,
+      "step": 6013
+    },
+    {
+      "epoch": 0.096224,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.110806451612903e-05,
+      "loss": 0.1822,
+      "step": 6014
+    },
+    {
+      "epoch": 0.09624,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.110645161290323e-05,
+      "loss": 0.1661,
+      "step": 6015
+    },
+    {
+      "epoch": 0.096256,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.110483870967742e-05,
+      "loss": 0.1586,
+      "step": 6016
+    },
+    {
+      "epoch": 0.096272,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.110322580645162e-05,
+      "loss": 0.2054,
+      "step": 6017
+    },
+    {
+      "epoch": 0.096288,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.11016129032258e-05,
+      "loss": 0.1677,
+      "step": 6018
+    },
+    {
+      "epoch": 0.096304,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.11e-05,
+      "loss": 0.207,
+      "step": 6019
+    },
+    {
+      "epoch": 0.09632,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.109838709677419e-05,
+      "loss": 0.1856,
+      "step": 6020
+    },
+    {
+      "epoch": 0.096336,
+      "grad_norm": 1.40625,
+      "learning_rate": 9.109677419354839e-05,
+      "loss": 0.1964,
+      "step": 6021
+    },
+    {
+      "epoch": 0.096352,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.109516129032258e-05,
+      "loss": 0.1806,
+      "step": 6022
+    },
+    {
+      "epoch": 0.096368,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.109354838709678e-05,
+      "loss": 0.1248,
+      "step": 6023
+    },
+    {
+      "epoch": 0.096384,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.109193548387098e-05,
+      "loss": 0.1624,
+      "step": 6024
+    },
+    {
+      "epoch": 0.0964,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.109032258064518e-05,
+      "loss": 0.2008,
+      "step": 6025
+    },
+    {
+      "epoch": 0.096416,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.108870967741936e-05,
+      "loss": 0.1594,
+      "step": 6026
+    },
+    {
+      "epoch": 0.096432,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.108709677419355e-05,
+      "loss": 0.2201,
+      "step": 6027
+    },
+    {
+      "epoch": 0.096448,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.108548387096775e-05,
+      "loss": 0.1806,
+      "step": 6028
+    },
+    {
+      "epoch": 0.096464,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.108387096774193e-05,
+      "loss": 0.2048,
+      "step": 6029
+    },
+    {
+      "epoch": 0.09648,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.108225806451613e-05,
+      "loss": 0.1476,
+      "step": 6030
+    },
+    {
+      "epoch": 0.096496,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.108064516129032e-05,
+      "loss": 0.1416,
+      "step": 6031
+    },
+    {
+      "epoch": 0.096512,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.107903225806452e-05,
+      "loss": 0.1563,
+      "step": 6032
+    },
+    {
+      "epoch": 0.096528,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.10774193548387e-05,
+      "loss": 0.2091,
+      "step": 6033
+    },
+    {
+      "epoch": 0.096544,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.10758064516129e-05,
+      "loss": 0.1774,
+      "step": 6034
+    },
+    {
+      "epoch": 0.09656,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.10741935483871e-05,
+      "loss": 0.1886,
+      "step": 6035
+    },
+    {
+      "epoch": 0.096576,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.10725806451613e-05,
+      "loss": 0.1514,
+      "step": 6036
+    },
+    {
+      "epoch": 0.096592,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.107096774193549e-05,
+      "loss": 0.2002,
+      "step": 6037
+    },
+    {
+      "epoch": 0.096608,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.106935483870969e-05,
+      "loss": 0.1953,
+      "step": 6038
+    },
+    {
+      "epoch": 0.096624,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.106774193548388e-05,
+      "loss": 0.17,
+      "step": 6039
+    },
+    {
+      "epoch": 0.09664,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.106612903225808e-05,
+      "loss": 0.1861,
+      "step": 6040
+    },
+    {
+      "epoch": 0.096656,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.106451612903226e-05,
+      "loss": 0.1628,
+      "step": 6041
+    },
+    {
+      "epoch": 0.096672,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.106290322580645e-05,
+      "loss": 0.1833,
+      "step": 6042
+    },
+    {
+      "epoch": 0.096688,
+      "grad_norm": 1.5,
+      "learning_rate": 9.106129032258065e-05,
+      "loss": 0.1459,
+      "step": 6043
+    },
+    {
+      "epoch": 0.096704,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.105967741935483e-05,
+      "loss": 0.1665,
+      "step": 6044
+    },
+    {
+      "epoch": 0.09672,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.105806451612903e-05,
+      "loss": 0.1751,
+      "step": 6045
+    },
+    {
+      "epoch": 0.096736,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.105645161290323e-05,
+      "loss": 0.1961,
+      "step": 6046
+    },
+    {
+      "epoch": 0.096752,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.105483870967743e-05,
+      "loss": 0.1687,
+      "step": 6047
+    },
+    {
+      "epoch": 0.096768,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.105322580645162e-05,
+      "loss": 0.1824,
+      "step": 6048
+    },
+    {
+      "epoch": 0.096784,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.105161290322582e-05,
+      "loss": 0.1641,
+      "step": 6049
+    },
+    {
+      "epoch": 0.0968,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.105e-05,
+      "loss": 0.13,
+      "step": 6050
+    },
+    {
+      "epoch": 0.096816,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.10483870967742e-05,
+      "loss": 0.1575,
+      "step": 6051
+    },
+    {
+      "epoch": 0.096832,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.104677419354839e-05,
+      "loss": 0.151,
+      "step": 6052
+    },
+    {
+      "epoch": 0.096848,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.104516129032259e-05,
+      "loss": 0.172,
+      "step": 6053
+    },
+    {
+      "epoch": 0.096864,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.104354838709678e-05,
+      "loss": 0.1709,
+      "step": 6054
+    },
+    {
+      "epoch": 0.09688,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.104193548387098e-05,
+      "loss": 0.1448,
+      "step": 6055
+    },
+    {
+      "epoch": 0.096896,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.104032258064516e-05,
+      "loss": 0.1905,
+      "step": 6056
+    },
+    {
+      "epoch": 0.096912,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.103870967741935e-05,
+      "loss": 0.1814,
+      "step": 6057
+    },
+    {
+      "epoch": 0.096928,
+      "grad_norm": 1.40625,
+      "learning_rate": 9.103709677419355e-05,
+      "loss": 0.1994,
+      "step": 6058
+    },
+    {
+      "epoch": 0.096944,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.103548387096775e-05,
+      "loss": 0.2193,
+      "step": 6059
+    },
+    {
+      "epoch": 0.09696,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.103387096774195e-05,
+      "loss": 0.1589,
+      "step": 6060
+    },
+    {
+      "epoch": 0.096976,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.103225806451613e-05,
+      "loss": 0.2071,
+      "step": 6061
+    },
+    {
+      "epoch": 0.096992,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.103064516129033e-05,
+      "loss": 0.1666,
+      "step": 6062
+    },
+    {
+      "epoch": 0.097008,
+      "grad_norm": 1.34375,
+      "learning_rate": 9.102903225806452e-05,
+      "loss": 0.1884,
+      "step": 6063
+    },
+    {
+      "epoch": 0.097024,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.102741935483872e-05,
+      "loss": 0.1477,
+      "step": 6064
+    },
+    {
+      "epoch": 0.09704,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.10258064516129e-05,
+      "loss": 0.1542,
+      "step": 6065
+    },
+    {
+      "epoch": 0.097056,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.10241935483871e-05,
+      "loss": 0.1838,
+      "step": 6066
+    },
+    {
+      "epoch": 0.097072,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.102258064516129e-05,
+      "loss": 0.1652,
+      "step": 6067
+    },
+    {
+      "epoch": 0.097088,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.102096774193549e-05,
+      "loss": 0.151,
+      "step": 6068
+    },
+    {
+      "epoch": 0.097104,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.101935483870968e-05,
+      "loss": 0.1528,
+      "step": 6069
+    },
+    {
+      "epoch": 0.09712,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.101774193548387e-05,
+      "loss": 0.1663,
+      "step": 6070
+    },
+    {
+      "epoch": 0.097136,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.101612903225807e-05,
+      "loss": 0.1854,
+      "step": 6071
+    },
+    {
+      "epoch": 0.097152,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.101451612903227e-05,
+      "loss": 0.1905,
+      "step": 6072
+    },
+    {
+      "epoch": 0.097168,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.101290322580646e-05,
+      "loss": 0.1837,
+      "step": 6073
+    },
+    {
+      "epoch": 0.097184,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.101129032258065e-05,
+      "loss": 0.2155,
+      "step": 6074
+    },
+    {
+      "epoch": 0.0972,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.100967741935485e-05,
+      "loss": 0.1645,
+      "step": 6075
+    },
+    {
+      "epoch": 0.097216,
+      "grad_norm": 1.328125,
+      "learning_rate": 9.100806451612903e-05,
+      "loss": 0.1665,
+      "step": 6076
+    },
+    {
+      "epoch": 0.097232,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.100645161290323e-05,
+      "loss": 0.1368,
+      "step": 6077
+    },
+    {
+      "epoch": 0.097248,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.100483870967742e-05,
+      "loss": 0.1803,
+      "step": 6078
+    },
+    {
+      "epoch": 0.097264,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.100322580645162e-05,
+      "loss": 0.1768,
+      "step": 6079
+    },
+    {
+      "epoch": 0.09728,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.10016129032258e-05,
+      "loss": 0.2028,
+      "step": 6080
+    },
+    {
+      "epoch": 0.097296,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.1e-05,
+      "loss": 0.1718,
+      "step": 6081
+    },
+    {
+      "epoch": 0.097312,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.09983870967742e-05,
+      "loss": 0.1899,
+      "step": 6082
+    },
+    {
+      "epoch": 0.097328,
+      "grad_norm": 1.359375,
+      "learning_rate": 9.099677419354839e-05,
+      "loss": 0.1907,
+      "step": 6083
+    },
+    {
+      "epoch": 0.097344,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.099516129032259e-05,
+      "loss": 0.1854,
+      "step": 6084
+    },
+    {
+      "epoch": 0.09736,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.099354838709679e-05,
+      "loss": 0.2252,
+      "step": 6085
+    },
+    {
+      "epoch": 0.097376,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.099193548387097e-05,
+      "loss": 0.1813,
+      "step": 6086
+    },
+    {
+      "epoch": 0.097392,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.099032258064517e-05,
+      "loss": 0.1982,
+      "step": 6087
+    },
+    {
+      "epoch": 0.097408,
+      "grad_norm": 1.3515625,
+      "learning_rate": 9.098870967741936e-05,
+      "loss": 0.1706,
+      "step": 6088
+    },
+    {
+      "epoch": 0.097424,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.098709677419355e-05,
+      "loss": 0.2105,
+      "step": 6089
+    },
+    {
+      "epoch": 0.09744,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.098548387096775e-05,
+      "loss": 0.1701,
+      "step": 6090
+    },
+    {
+      "epoch": 0.097456,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.098387096774193e-05,
+      "loss": 0.1724,
+      "step": 6091
+    },
+    {
+      "epoch": 0.097472,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.098225806451613e-05,
+      "loss": 0.1735,
+      "step": 6092
+    },
+    {
+      "epoch": 0.097488,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.098064516129032e-05,
+      "loss": 0.1915,
+      "step": 6093
+    },
+    {
+      "epoch": 0.097504,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.097903225806452e-05,
+      "loss": 0.1756,
+      "step": 6094
+    },
+    {
+      "epoch": 0.09752,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.097741935483872e-05,
+      "loss": 0.1527,
+      "step": 6095
+    },
+    {
+      "epoch": 0.097536,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.097580645161292e-05,
+      "loss": 0.1544,
+      "step": 6096
+    },
+    {
+      "epoch": 0.097552,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.09741935483871e-05,
+      "loss": 0.1576,
+      "step": 6097
+    },
+    {
+      "epoch": 0.097568,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.09725806451613e-05,
+      "loss": 0.1691,
+      "step": 6098
+    },
+    {
+      "epoch": 0.097584,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.097096774193549e-05,
+      "loss": 0.1673,
+      "step": 6099
+    },
+    {
+      "epoch": 0.0976,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.096935483870969e-05,
+      "loss": 0.1769,
+      "step": 6100
+    },
+    {
+      "epoch": 0.097616,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.096774193548387e-05,
+      "loss": 0.1708,
+      "step": 6101
+    },
+    {
+      "epoch": 0.097632,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.096612903225807e-05,
+      "loss": 0.181,
+      "step": 6102
+    },
+    {
+      "epoch": 0.097648,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.096451612903226e-05,
+      "loss": 0.1872,
+      "step": 6103
+    },
+    {
+      "epoch": 0.097664,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.096290322580645e-05,
+      "loss": 0.1883,
+      "step": 6104
+    },
+    {
+      "epoch": 0.09768,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.096129032258064e-05,
+      "loss": 0.1936,
+      "step": 6105
+    },
+    {
+      "epoch": 0.097696,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.095967741935484e-05,
+      "loss": 0.1944,
+      "step": 6106
+    },
+    {
+      "epoch": 0.097712,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.095806451612904e-05,
+      "loss": 0.1852,
+      "step": 6107
+    },
+    {
+      "epoch": 0.097728,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.095645161290323e-05,
+      "loss": 0.1909,
+      "step": 6108
+    },
+    {
+      "epoch": 0.097744,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.095483870967743e-05,
+      "loss": 0.2392,
+      "step": 6109
+    },
+    {
+      "epoch": 0.09776,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.095322580645162e-05,
+      "loss": 0.1701,
+      "step": 6110
+    },
+    {
+      "epoch": 0.097776,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.095161290322582e-05,
+      "loss": 0.1633,
+      "step": 6111
+    },
+    {
+      "epoch": 0.097792,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.095e-05,
+      "loss": 0.187,
+      "step": 6112
+    },
+    {
+      "epoch": 0.097808,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.09483870967742e-05,
+      "loss": 0.1613,
+      "step": 6113
+    },
+    {
+      "epoch": 0.097824,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.094677419354839e-05,
+      "loss": 0.1684,
+      "step": 6114
+    },
+    {
+      "epoch": 0.09784,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.094516129032259e-05,
+      "loss": 0.1681,
+      "step": 6115
+    },
+    {
+      "epoch": 0.097856,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.094354838709677e-05,
+      "loss": 0.1939,
+      "step": 6116
+    },
+    {
+      "epoch": 0.097872,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.094193548387097e-05,
+      "loss": 0.1848,
+      "step": 6117
+    },
+    {
+      "epoch": 0.097888,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.094032258064516e-05,
+      "loss": 0.2167,
+      "step": 6118
+    },
+    {
+      "epoch": 0.097904,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.093870967741936e-05,
+      "loss": 0.1619,
+      "step": 6119
+    },
+    {
+      "epoch": 0.09792,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.093709677419356e-05,
+      "loss": 0.2358,
+      "step": 6120
+    },
+    {
+      "epoch": 0.097936,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.093548387096774e-05,
+      "loss": 0.1653,
+      "step": 6121
+    },
+    {
+      "epoch": 0.097952,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.093387096774194e-05,
+      "loss": 0.1716,
+      "step": 6122
+    },
+    {
+      "epoch": 0.097968,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.093225806451613e-05,
+      "loss": 0.1551,
+      "step": 6123
+    },
+    {
+      "epoch": 0.097984,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.093064516129033e-05,
+      "loss": 0.1872,
+      "step": 6124
+    },
+    {
+      "epoch": 0.098,
+      "grad_norm": 1.0,
+      "learning_rate": 9.092903225806452e-05,
+      "loss": 0.2084,
+      "step": 6125
+    },
+    {
+      "epoch": 0.098016,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.092741935483872e-05,
+      "loss": 0.2106,
+      "step": 6126
+    },
+    {
+      "epoch": 0.098032,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.09258064516129e-05,
+      "loss": 0.1936,
+      "step": 6127
+    },
+    {
+      "epoch": 0.098048,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.09241935483871e-05,
+      "loss": 0.1528,
+      "step": 6128
+    },
+    {
+      "epoch": 0.098064,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.092258064516129e-05,
+      "loss": 0.1583,
+      "step": 6129
+    },
+    {
+      "epoch": 0.09808,
+      "grad_norm": 1.125,
+      "learning_rate": 9.092096774193549e-05,
+      "loss": 0.2273,
+      "step": 6130
+    },
+    {
+      "epoch": 0.098096,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.091935483870969e-05,
+      "loss": 0.1959,
+      "step": 6131
+    },
+    {
+      "epoch": 0.098112,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.091774193548389e-05,
+      "loss": 0.2293,
+      "step": 6132
+    },
+    {
+      "epoch": 0.098128,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.091612903225807e-05,
+      "loss": 0.1472,
+      "step": 6133
+    },
+    {
+      "epoch": 0.098144,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.091451612903227e-05,
+      "loss": 0.1838,
+      "step": 6134
+    },
+    {
+      "epoch": 0.09816,
+      "grad_norm": 1.0,
+      "learning_rate": 9.091290322580646e-05,
+      "loss": 0.2142,
+      "step": 6135
+    },
+    {
+      "epoch": 0.098176,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.091129032258064e-05,
+      "loss": 0.1524,
+      "step": 6136
+    },
+    {
+      "epoch": 0.098192,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.090967741935484e-05,
+      "loss": 0.2179,
+      "step": 6137
+    },
+    {
+      "epoch": 0.098208,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.090806451612903e-05,
+      "loss": 0.1867,
+      "step": 6138
+    },
+    {
+      "epoch": 0.098224,
+      "grad_norm": 1.0,
+      "learning_rate": 9.090645161290323e-05,
+      "loss": 0.211,
+      "step": 6139
+    },
+    {
+      "epoch": 0.09824,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.090483870967742e-05,
+      "loss": 0.1682,
+      "step": 6140
+    },
+    {
+      "epoch": 0.098256,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.090322580645161e-05,
+      "loss": 0.1805,
+      "step": 6141
+    },
+    {
+      "epoch": 0.098272,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.090161290322581e-05,
+      "loss": 0.183,
+      "step": 6142
+    },
+    {
+      "epoch": 0.098288,
+      "grad_norm": 0.875,
+      "learning_rate": 9.090000000000001e-05,
+      "loss": 0.1627,
+      "step": 6143
+    },
+    {
+      "epoch": 0.098304,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.08983870967742e-05,
+      "loss": 0.2007,
+      "step": 6144
+    },
+    {
+      "epoch": 0.09832,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.08967741935484e-05,
+      "loss": 0.1535,
+      "step": 6145
+    },
+    {
+      "epoch": 0.098336,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.089516129032259e-05,
+      "loss": 0.2155,
+      "step": 6146
+    },
+    {
+      "epoch": 0.098352,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.089354838709679e-05,
+      "loss": 0.1835,
+      "step": 6147
+    },
+    {
+      "epoch": 0.098368,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.089193548387097e-05,
+      "loss": 0.1514,
+      "step": 6148
+    },
+    {
+      "epoch": 0.098384,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.089032258064517e-05,
+      "loss": 0.1958,
+      "step": 6149
+    },
+    {
+      "epoch": 0.0984,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.088870967741936e-05,
+      "loss": 0.1609,
+      "step": 6150
+    },
+    {
+      "epoch": 0.098416,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.088709677419354e-05,
+      "loss": 0.19,
+      "step": 6151
+    },
+    {
+      "epoch": 0.098432,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.088548387096774e-05,
+      "loss": 0.188,
+      "step": 6152
+    },
+    {
+      "epoch": 0.098448,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.088387096774193e-05,
+      "loss": 0.1734,
+      "step": 6153
+    },
+    {
+      "epoch": 0.098464,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.088225806451613e-05,
+      "loss": 0.1695,
+      "step": 6154
+    },
+    {
+      "epoch": 0.09848,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.088064516129033e-05,
+      "loss": 0.1464,
+      "step": 6155
+    },
+    {
+      "epoch": 0.098496,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.087903225806453e-05,
+      "loss": 0.2051,
+      "step": 6156
+    },
+    {
+      "epoch": 0.098512,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.087741935483871e-05,
+      "loss": 0.2007,
+      "step": 6157
+    },
+    {
+      "epoch": 0.098528,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.087580645161291e-05,
+      "loss": 0.1726,
+      "step": 6158
+    },
+    {
+      "epoch": 0.098544,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.08741935483871e-05,
+      "loss": 0.2113,
+      "step": 6159
+    },
+    {
+      "epoch": 0.09856,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.08725806451613e-05,
+      "loss": 0.2117,
+      "step": 6160
+    },
+    {
+      "epoch": 0.098576,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.087096774193549e-05,
+      "loss": 0.1934,
+      "step": 6161
+    },
+    {
+      "epoch": 0.098592,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.086935483870968e-05,
+      "loss": 0.1648,
+      "step": 6162
+    },
+    {
+      "epoch": 0.098608,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.086774193548387e-05,
+      "loss": 0.1384,
+      "step": 6163
+    },
+    {
+      "epoch": 0.098624,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.086612903225807e-05,
+      "loss": 0.174,
+      "step": 6164
+    },
+    {
+      "epoch": 0.09864,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.086451612903226e-05,
+      "loss": 0.1661,
+      "step": 6165
+    },
+    {
+      "epoch": 0.098656,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.086290322580646e-05,
+      "loss": 0.1556,
+      "step": 6166
+    },
+    {
+      "epoch": 0.098672,
+      "grad_norm": 0.875,
+      "learning_rate": 9.086129032258066e-05,
+      "loss": 0.1618,
+      "step": 6167
+    },
+    {
+      "epoch": 0.098688,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.085967741935484e-05,
+      "loss": 0.1963,
+      "step": 6168
+    },
+    {
+      "epoch": 0.098704,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.085806451612904e-05,
+      "loss": 0.1923,
+      "step": 6169
+    },
+    {
+      "epoch": 0.09872,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.085645161290323e-05,
+      "loss": 0.1844,
+      "step": 6170
+    },
+    {
+      "epoch": 0.098736,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.085483870967743e-05,
+      "loss": 0.1889,
+      "step": 6171
+    },
+    {
+      "epoch": 0.098752,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.085322580645161e-05,
+      "loss": 0.1612,
+      "step": 6172
+    },
+    {
+      "epoch": 0.098768,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.085161290322581e-05,
+      "loss": 0.1759,
+      "step": 6173
+    },
+    {
+      "epoch": 0.098784,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.085e-05,
+      "loss": 0.1689,
+      "step": 6174
+    },
+    {
+      "epoch": 0.0988,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.08483870967742e-05,
+      "loss": 0.1803,
+      "step": 6175
+    },
+    {
+      "epoch": 0.098816,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.084677419354838e-05,
+      "loss": 0.171,
+      "step": 6176
+    },
+    {
+      "epoch": 0.098832,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.084516129032258e-05,
+      "loss": 0.2228,
+      "step": 6177
+    },
+    {
+      "epoch": 0.098848,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.084354838709677e-05,
+      "loss": 0.2094,
+      "step": 6178
+    },
+    {
+      "epoch": 0.098864,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.084193548387097e-05,
+      "loss": 0.1753,
+      "step": 6179
+    },
+    {
+      "epoch": 0.09888,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.084032258064517e-05,
+      "loss": 0.1564,
+      "step": 6180
+    },
+    {
+      "epoch": 0.098896,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.083870967741937e-05,
+      "loss": 0.2001,
+      "step": 6181
+    },
+    {
+      "epoch": 0.098912,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.083709677419356e-05,
+      "loss": 0.1699,
+      "step": 6182
+    },
+    {
+      "epoch": 0.098928,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.083548387096774e-05,
+      "loss": 0.1869,
+      "step": 6183
+    },
+    {
+      "epoch": 0.098944,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.083387096774194e-05,
+      "loss": 0.1683,
+      "step": 6184
+    },
+    {
+      "epoch": 0.09896,
+      "grad_norm": 0.75,
+      "learning_rate": 9.083225806451613e-05,
+      "loss": 0.1948,
+      "step": 6185
+    },
+    {
+      "epoch": 0.098976,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.083064516129033e-05,
+      "loss": 0.16,
+      "step": 6186
+    },
+    {
+      "epoch": 0.098992,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.082903225806451e-05,
+      "loss": 0.2032,
+      "step": 6187
+    },
+    {
+      "epoch": 0.099008,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.082741935483871e-05,
+      "loss": 0.1297,
+      "step": 6188
+    },
+    {
+      "epoch": 0.099024,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.08258064516129e-05,
+      "loss": 0.1733,
+      "step": 6189
+    },
+    {
+      "epoch": 0.09904,
+      "grad_norm": 1.515625,
+      "learning_rate": 9.08241935483871e-05,
+      "loss": 0.2269,
+      "step": 6190
+    },
+    {
+      "epoch": 0.099056,
+      "grad_norm": 1.4921875,
+      "learning_rate": 9.08225806451613e-05,
+      "loss": 0.1703,
+      "step": 6191
+    },
+    {
+      "epoch": 0.099072,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.08209677419355e-05,
+      "loss": 0.1964,
+      "step": 6192
+    },
+    {
+      "epoch": 0.099088,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.081935483870968e-05,
+      "loss": 0.2063,
+      "step": 6193
+    },
+    {
+      "epoch": 0.099104,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.081774193548388e-05,
+      "loss": 0.1812,
+      "step": 6194
+    },
+    {
+      "epoch": 0.09912,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.081612903225807e-05,
+      "loss": 0.1559,
+      "step": 6195
+    },
+    {
+      "epoch": 0.099136,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.081451612903227e-05,
+      "loss": 0.1654,
+      "step": 6196
+    },
+    {
+      "epoch": 0.099152,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.081290322580646e-05,
+      "loss": 0.1909,
+      "step": 6197
+    },
+    {
+      "epoch": 0.099168,
+      "grad_norm": 1.3515625,
+      "learning_rate": 9.081129032258064e-05,
+      "loss": 0.2072,
+      "step": 6198
+    },
+    {
+      "epoch": 0.099184,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.080967741935484e-05,
+      "loss": 0.2134,
+      "step": 6199
+    },
+    {
+      "epoch": 0.0992,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.080806451612903e-05,
+      "loss": 0.1732,
+      "step": 6200
+    },
+    {
+      "epoch": 0.099216,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.080645161290323e-05,
+      "loss": 0.2026,
+      "step": 6201
+    },
+    {
+      "epoch": 0.099232,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.080483870967743e-05,
+      "loss": 0.1964,
+      "step": 6202
+    },
+    {
+      "epoch": 0.099248,
+      "grad_norm": 1.4609375,
+      "learning_rate": 9.080322580645163e-05,
+      "loss": 0.1556,
+      "step": 6203
+    },
+    {
+      "epoch": 0.099264,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.080161290322581e-05,
+      "loss": 0.2284,
+      "step": 6204
+    },
+    {
+      "epoch": 0.09928,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.080000000000001e-05,
+      "loss": 0.1626,
+      "step": 6205
+    },
+    {
+      "epoch": 0.099296,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.07983870967742e-05,
+      "loss": 0.1455,
+      "step": 6206
+    },
+    {
+      "epoch": 0.099312,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.07967741935484e-05,
+      "loss": 0.1974,
+      "step": 6207
+    },
+    {
+      "epoch": 0.099328,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.079516129032258e-05,
+      "loss": 0.23,
+      "step": 6208
+    },
+    {
+      "epoch": 0.099344,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.079354838709678e-05,
+      "loss": 0.1695,
+      "step": 6209
+    },
+    {
+      "epoch": 0.09936,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.079193548387097e-05,
+      "loss": 0.1409,
+      "step": 6210
+    },
+    {
+      "epoch": 0.099376,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.079032258064517e-05,
+      "loss": 0.1986,
+      "step": 6211
+    },
+    {
+      "epoch": 0.099392,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.078870967741935e-05,
+      "loss": 0.2219,
+      "step": 6212
+    },
+    {
+      "epoch": 0.099408,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.078709677419354e-05,
+      "loss": 0.1603,
+      "step": 6213
+    },
+    {
+      "epoch": 0.099424,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.078548387096774e-05,
+      "loss": 0.1777,
+      "step": 6214
+    },
+    {
+      "epoch": 0.09944,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.078387096774194e-05,
+      "loss": 0.2025,
+      "step": 6215
+    },
+    {
+      "epoch": 0.099456,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.078225806451614e-05,
+      "loss": 0.1834,
+      "step": 6216
+    },
+    {
+      "epoch": 0.099472,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.078064516129033e-05,
+      "loss": 0.1881,
+      "step": 6217
+    },
+    {
+      "epoch": 0.099488,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.077903225806453e-05,
+      "loss": 0.2313,
+      "step": 6218
+    },
+    {
+      "epoch": 0.099504,
+      "grad_norm": 0.625,
+      "learning_rate": 9.077741935483871e-05,
+      "loss": 0.1715,
+      "step": 6219
+    },
+    {
+      "epoch": 0.09952,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.077580645161291e-05,
+      "loss": 0.1843,
+      "step": 6220
+    },
+    {
+      "epoch": 0.099536,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.07741935483871e-05,
+      "loss": 0.1756,
+      "step": 6221
+    },
+    {
+      "epoch": 0.099552,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.07725806451613e-05,
+      "loss": 0.2074,
+      "step": 6222
+    },
+    {
+      "epoch": 0.099568,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.077096774193548e-05,
+      "loss": 0.1632,
+      "step": 6223
+    },
+    {
+      "epoch": 0.099584,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.076935483870968e-05,
+      "loss": 0.1843,
+      "step": 6224
+    },
+    {
+      "epoch": 0.0996,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.076774193548387e-05,
+      "loss": 0.1902,
+      "step": 6225
+    },
+    {
+      "epoch": 0.099616,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.076612903225807e-05,
+      "loss": 0.1623,
+      "step": 6226
+    },
+    {
+      "epoch": 0.099632,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.076451612903227e-05,
+      "loss": 0.1925,
+      "step": 6227
+    },
+    {
+      "epoch": 0.099648,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.076290322580645e-05,
+      "loss": 0.2036,
+      "step": 6228
+    },
+    {
+      "epoch": 0.099664,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.076129032258065e-05,
+      "loss": 0.1802,
+      "step": 6229
+    },
+    {
+      "epoch": 0.09968,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.075967741935484e-05,
+      "loss": 0.1944,
+      "step": 6230
+    },
+    {
+      "epoch": 0.099696,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.075806451612904e-05,
+      "loss": 0.1598,
+      "step": 6231
+    },
+    {
+      "epoch": 0.099712,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.075645161290323e-05,
+      "loss": 0.1872,
+      "step": 6232
+    },
+    {
+      "epoch": 0.099728,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.075483870967743e-05,
+      "loss": 0.1816,
+      "step": 6233
+    },
+    {
+      "epoch": 0.099744,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.075322580645161e-05,
+      "loss": 0.1819,
+      "step": 6234
+    },
+    {
+      "epoch": 0.09976,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.075161290322581e-05,
+      "loss": 0.1913,
+      "step": 6235
+    },
+    {
+      "epoch": 0.099776,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.075e-05,
+      "loss": 0.1747,
+      "step": 6236
+    },
+    {
+      "epoch": 0.099792,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.07483870967742e-05,
+      "loss": 0.1854,
+      "step": 6237
+    },
+    {
+      "epoch": 0.099808,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.07467741935484e-05,
+      "loss": 0.139,
+      "step": 6238
+    },
+    {
+      "epoch": 0.099824,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.07451612903226e-05,
+      "loss": 0.159,
+      "step": 6239
+    },
+    {
+      "epoch": 0.09984,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.074354838709678e-05,
+      "loss": 0.1879,
+      "step": 6240
+    },
+    {
+      "epoch": 0.099856,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.074193548387098e-05,
+      "loss": 0.1742,
+      "step": 6241
+    },
+    {
+      "epoch": 0.099872,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.074032258064517e-05,
+      "loss": 0.1423,
+      "step": 6242
+    },
+    {
+      "epoch": 0.099888,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.073870967741937e-05,
+      "loss": 0.1764,
+      "step": 6243
+    },
+    {
+      "epoch": 0.099904,
+      "grad_norm": 1.515625,
+      "learning_rate": 9.073709677419355e-05,
+      "loss": 0.1934,
+      "step": 6244
+    },
+    {
+      "epoch": 0.09992,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.073548387096774e-05,
+      "loss": 0.1502,
+      "step": 6245
+    },
+    {
+      "epoch": 0.099936,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.073387096774194e-05,
+      "loss": 0.1977,
+      "step": 6246
+    },
+    {
+      "epoch": 0.099952,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.073225806451612e-05,
+      "loss": 0.1972,
+      "step": 6247
+    },
+    {
+      "epoch": 0.099968,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.073064516129032e-05,
+      "loss": 0.1658,
+      "step": 6248
+    },
+    {
+      "epoch": 0.099984,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.072903225806451e-05,
+      "loss": 0.1918,
+      "step": 6249
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.072741935483871e-05,
+      "loss": 0.1581,
+      "step": 6250
+    },
+    {
+      "epoch": 0.100016,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.072580645161291e-05,
+      "loss": 0.1839,
+      "step": 6251
+    },
+    {
+      "epoch": 0.100032,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.072419354838711e-05,
+      "loss": 0.1846,
+      "step": 6252
+    },
+    {
+      "epoch": 0.100048,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.07225806451613e-05,
+      "loss": 0.1594,
+      "step": 6253
+    },
+    {
+      "epoch": 0.100064,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.07209677419355e-05,
+      "loss": 0.1949,
+      "step": 6254
+    },
+    {
+      "epoch": 0.10008,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.071935483870968e-05,
+      "loss": 0.1733,
+      "step": 6255
+    },
+    {
+      "epoch": 0.100096,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.071774193548388e-05,
+      "loss": 0.2304,
+      "step": 6256
+    },
+    {
+      "epoch": 0.100112,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.071612903225807e-05,
+      "loss": 0.1746,
+      "step": 6257
+    },
+    {
+      "epoch": 0.100128,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.071451612903227e-05,
+      "loss": 0.2019,
+      "step": 6258
+    },
+    {
+      "epoch": 0.100144,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.071290322580645e-05,
+      "loss": 0.1453,
+      "step": 6259
+    },
+    {
+      "epoch": 0.10016,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.071129032258064e-05,
+      "loss": 0.1726,
+      "step": 6260
+    },
+    {
+      "epoch": 0.100176,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.070967741935484e-05,
+      "loss": 0.1891,
+      "step": 6261
+    },
+    {
+      "epoch": 0.100192,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.070806451612904e-05,
+      "loss": 0.1929,
+      "step": 6262
+    },
+    {
+      "epoch": 0.100208,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.070645161290324e-05,
+      "loss": 0.1835,
+      "step": 6263
+    },
+    {
+      "epoch": 0.100224,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.070483870967742e-05,
+      "loss": 0.1738,
+      "step": 6264
+    },
+    {
+      "epoch": 0.10024,
+      "grad_norm": 1.125,
+      "learning_rate": 9.070322580645162e-05,
+      "loss": 0.177,
+      "step": 6265
+    },
+    {
+      "epoch": 0.100256,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.070161290322581e-05,
+      "loss": 0.1537,
+      "step": 6266
+    },
+    {
+      "epoch": 0.100272,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.070000000000001e-05,
+      "loss": 0.1847,
+      "step": 6267
+    },
+    {
+      "epoch": 0.100288,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.06983870967742e-05,
+      "loss": 0.209,
+      "step": 6268
+    },
+    {
+      "epoch": 0.100304,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.06967741935484e-05,
+      "loss": 0.1296,
+      "step": 6269
+    },
+    {
+      "epoch": 0.10032,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.069516129032258e-05,
+      "loss": 0.1725,
+      "step": 6270
+    },
+    {
+      "epoch": 0.100336,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.069354838709678e-05,
+      "loss": 0.1351,
+      "step": 6271
+    },
+    {
+      "epoch": 0.100352,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.069193548387097e-05,
+      "loss": 0.1653,
+      "step": 6272
+    },
+    {
+      "epoch": 0.100368,
+      "grad_norm": 1.8671875,
+      "learning_rate": 9.069032258064517e-05,
+      "loss": 0.1844,
+      "step": 6273
+    },
+    {
+      "epoch": 0.100384,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.068870967741935e-05,
+      "loss": 0.1696,
+      "step": 6274
+    },
+    {
+      "epoch": 0.1004,
+      "grad_norm": 1.9609375,
+      "learning_rate": 9.068709677419355e-05,
+      "loss": 0.1939,
+      "step": 6275
+    },
+    {
+      "epoch": 0.100416,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.068548387096775e-05,
+      "loss": 0.1934,
+      "step": 6276
+    },
+    {
+      "epoch": 0.100432,
+      "grad_norm": 1.4921875,
+      "learning_rate": 9.068387096774194e-05,
+      "loss": 0.212,
+      "step": 6277
+    },
+    {
+      "epoch": 0.100448,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.068225806451614e-05,
+      "loss": 0.136,
+      "step": 6278
+    },
+    {
+      "epoch": 0.100464,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.068064516129032e-05,
+      "loss": 0.1982,
+      "step": 6279
+    },
+    {
+      "epoch": 0.10048,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.067903225806452e-05,
+      "loss": 0.1383,
+      "step": 6280
+    },
+    {
+      "epoch": 0.100496,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.067741935483871e-05,
+      "loss": 0.2333,
+      "step": 6281
+    },
+    {
+      "epoch": 0.100512,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.067580645161291e-05,
+      "loss": 0.1743,
+      "step": 6282
+    },
+    {
+      "epoch": 0.100528,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.06741935483871e-05,
+      "loss": 0.189,
+      "step": 6283
+    },
+    {
+      "epoch": 0.100544,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.06725806451613e-05,
+      "loss": 0.1685,
+      "step": 6284
+    },
+    {
+      "epoch": 0.10056,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.067096774193548e-05,
+      "loss": 0.1837,
+      "step": 6285
+    },
+    {
+      "epoch": 0.100576,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.066935483870968e-05,
+      "loss": 0.1543,
+      "step": 6286
+    },
+    {
+      "epoch": 0.100592,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.066774193548388e-05,
+      "loss": 0.2155,
+      "step": 6287
+    },
+    {
+      "epoch": 0.100608,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.066612903225808e-05,
+      "loss": 0.2497,
+      "step": 6288
+    },
+    {
+      "epoch": 0.100624,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.066451612903227e-05,
+      "loss": 0.1523,
+      "step": 6289
+    },
+    {
+      "epoch": 0.10064,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.066290322580647e-05,
+      "loss": 0.2225,
+      "step": 6290
+    },
+    {
+      "epoch": 0.100656,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.066129032258065e-05,
+      "loss": 0.1445,
+      "step": 6291
+    },
+    {
+      "epoch": 0.100672,
+      "grad_norm": 1.125,
+      "learning_rate": 9.065967741935484e-05,
+      "loss": 0.1434,
+      "step": 6292
+    },
+    {
+      "epoch": 0.100688,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.065806451612904e-05,
+      "loss": 0.1761,
+      "step": 6293
+    },
+    {
+      "epoch": 0.100704,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.065645161290322e-05,
+      "loss": 0.1987,
+      "step": 6294
+    },
+    {
+      "epoch": 0.10072,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.065483870967742e-05,
+      "loss": 0.2157,
+      "step": 6295
+    },
+    {
+      "epoch": 0.100736,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.065322580645161e-05,
+      "loss": 0.1783,
+      "step": 6296
+    },
+    {
+      "epoch": 0.100752,
+      "grad_norm": 1.3125,
+      "learning_rate": 9.065161290322581e-05,
+      "loss": 0.1638,
+      "step": 6297
+    },
+    {
+      "epoch": 0.100768,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.065000000000001e-05,
+      "loss": 0.1512,
+      "step": 6298
+    },
+    {
+      "epoch": 0.100784,
+      "grad_norm": 1.3984375,
+      "learning_rate": 9.064838709677421e-05,
+      "loss": 0.1839,
+      "step": 6299
+    },
+    {
+      "epoch": 0.1008,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.06467741935484e-05,
+      "loss": 0.1448,
+      "step": 6300
+    },
+    {
+      "epoch": 0.100816,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.06451612903226e-05,
+      "loss": 0.2027,
+      "step": 6301
+    },
+    {
+      "epoch": 0.100832,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.064354838709678e-05,
+      "loss": 0.2052,
+      "step": 6302
+    },
+    {
+      "epoch": 0.100848,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.064193548387098e-05,
+      "loss": 0.1691,
+      "step": 6303
+    },
+    {
+      "epoch": 0.100864,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.064032258064517e-05,
+      "loss": 0.1948,
+      "step": 6304
+    },
+    {
+      "epoch": 0.10088,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.063870967741936e-05,
+      "loss": 0.1578,
+      "step": 6305
+    },
+    {
+      "epoch": 0.100896,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.063709677419355e-05,
+      "loss": 0.1824,
+      "step": 6306
+    },
+    {
+      "epoch": 0.100912,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.063548387096774e-05,
+      "loss": 0.1575,
+      "step": 6307
+    },
+    {
+      "epoch": 0.100928,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.063387096774194e-05,
+      "loss": 0.1938,
+      "step": 6308
+    },
+    {
+      "epoch": 0.100944,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.063225806451612e-05,
+      "loss": 0.1888,
+      "step": 6309
+    },
+    {
+      "epoch": 0.10096,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.063064516129032e-05,
+      "loss": 0.1939,
+      "step": 6310
+    },
+    {
+      "epoch": 0.100976,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.062903225806452e-05,
+      "loss": 0.1642,
+      "step": 6311
+    },
+    {
+      "epoch": 0.100992,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.062741935483872e-05,
+      "loss": 0.1616,
+      "step": 6312
+    },
+    {
+      "epoch": 0.101008,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.062580645161291e-05,
+      "loss": 0.1599,
+      "step": 6313
+    },
+    {
+      "epoch": 0.101024,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.062419354838711e-05,
+      "loss": 0.1608,
+      "step": 6314
+    },
+    {
+      "epoch": 0.10104,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.062258064516129e-05,
+      "loss": 0.1365,
+      "step": 6315
+    },
+    {
+      "epoch": 0.101056,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.062096774193549e-05,
+      "loss": 0.1824,
+      "step": 6316
+    },
+    {
+      "epoch": 0.101072,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.061935483870968e-05,
+      "loss": 0.168,
+      "step": 6317
+    },
+    {
+      "epoch": 0.101088,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.061774193548388e-05,
+      "loss": 0.1781,
+      "step": 6318
+    },
+    {
+      "epoch": 0.101104,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.061612903225806e-05,
+      "loss": 0.1413,
+      "step": 6319
+    },
+    {
+      "epoch": 0.10112,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.061451612903226e-05,
+      "loss": 0.1932,
+      "step": 6320
+    },
+    {
+      "epoch": 0.101136,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.061290322580645e-05,
+      "loss": 0.1826,
+      "step": 6321
+    },
+    {
+      "epoch": 0.101152,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.061129032258065e-05,
+      "loss": 0.1946,
+      "step": 6322
+    },
+    {
+      "epoch": 0.101168,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.060967741935485e-05,
+      "loss": 0.1808,
+      "step": 6323
+    },
+    {
+      "epoch": 0.101184,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.060806451612904e-05,
+      "loss": 0.1824,
+      "step": 6324
+    },
+    {
+      "epoch": 0.1012,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.060645161290324e-05,
+      "loss": 0.1665,
+      "step": 6325
+    },
+    {
+      "epoch": 0.101216,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.060483870967742e-05,
+      "loss": 0.1782,
+      "step": 6326
+    },
+    {
+      "epoch": 0.101232,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.060322580645162e-05,
+      "loss": 0.1994,
+      "step": 6327
+    },
+    {
+      "epoch": 0.101248,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.060161290322581e-05,
+      "loss": 0.1995,
+      "step": 6328
+    },
+    {
+      "epoch": 0.101264,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.06e-05,
+      "loss": 0.1513,
+      "step": 6329
+    },
+    {
+      "epoch": 0.10128,
+      "grad_norm": 1.0,
+      "learning_rate": 9.059838709677419e-05,
+      "loss": 0.2057,
+      "step": 6330
+    },
+    {
+      "epoch": 0.101296,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.059677419354839e-05,
+      "loss": 0.1905,
+      "step": 6331
+    },
+    {
+      "epoch": 0.101312,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.059516129032258e-05,
+      "loss": 0.1907,
+      "step": 6332
+    },
+    {
+      "epoch": 0.101328,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.059354838709678e-05,
+      "loss": 0.1924,
+      "step": 6333
+    },
+    {
+      "epoch": 0.101344,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.059193548387098e-05,
+      "loss": 0.2158,
+      "step": 6334
+    },
+    {
+      "epoch": 0.10136,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.059032258064516e-05,
+      "loss": 0.1902,
+      "step": 6335
+    },
+    {
+      "epoch": 0.101376,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.058870967741936e-05,
+      "loss": 0.1864,
+      "step": 6336
+    },
+    {
+      "epoch": 0.101392,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.058709677419355e-05,
+      "loss": 0.185,
+      "step": 6337
+    },
+    {
+      "epoch": 0.101408,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.058548387096775e-05,
+      "loss": 0.1484,
+      "step": 6338
+    },
+    {
+      "epoch": 0.101424,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.058387096774194e-05,
+      "loss": 0.1821,
+      "step": 6339
+    },
+    {
+      "epoch": 0.10144,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.058225806451613e-05,
+      "loss": 0.1951,
+      "step": 6340
+    },
+    {
+      "epoch": 0.101456,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.058064516129032e-05,
+      "loss": 0.1399,
+      "step": 6341
+    },
+    {
+      "epoch": 0.101472,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.057903225806452e-05,
+      "loss": 0.1821,
+      "step": 6342
+    },
+    {
+      "epoch": 0.101488,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.05774193548387e-05,
+      "loss": 0.1824,
+      "step": 6343
+    },
+    {
+      "epoch": 0.101504,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.05758064516129e-05,
+      "loss": 0.1884,
+      "step": 6344
+    },
+    {
+      "epoch": 0.10152,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.057419354838709e-05,
+      "loss": 0.1915,
+      "step": 6345
+    },
+    {
+      "epoch": 0.101536,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.057258064516129e-05,
+      "loss": 0.1685,
+      "step": 6346
+    },
+    {
+      "epoch": 0.101552,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.057096774193549e-05,
+      "loss": 0.2102,
+      "step": 6347
+    },
+    {
+      "epoch": 0.101568,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.056935483870969e-05,
+      "loss": 0.143,
+      "step": 6348
+    },
+    {
+      "epoch": 0.101584,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.056774193548388e-05,
+      "loss": 0.1767,
+      "step": 6349
+    },
+    {
+      "epoch": 0.1016,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.056612903225808e-05,
+      "loss": 0.1728,
+      "step": 6350
+    },
+    {
+      "epoch": 0.101616,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.056451612903226e-05,
+      "loss": 0.1785,
+      "step": 6351
+    },
+    {
+      "epoch": 0.101632,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.056290322580646e-05,
+      "loss": 0.1591,
+      "step": 6352
+    },
+    {
+      "epoch": 0.101648,
+      "grad_norm": 1.515625,
+      "learning_rate": 9.056129032258065e-05,
+      "loss": 0.1548,
+      "step": 6353
+    },
+    {
+      "epoch": 0.101664,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.055967741935483e-05,
+      "loss": 0.2116,
+      "step": 6354
+    },
+    {
+      "epoch": 0.10168,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.055806451612903e-05,
+      "loss": 0.1938,
+      "step": 6355
+    },
+    {
+      "epoch": 0.101696,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.055645161290322e-05,
+      "loss": 0.1948,
+      "step": 6356
+    },
+    {
+      "epoch": 0.101712,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.055483870967742e-05,
+      "loss": 0.1352,
+      "step": 6357
+    },
+    {
+      "epoch": 0.101728,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.055322580645162e-05,
+      "loss": 0.1848,
+      "step": 6358
+    },
+    {
+      "epoch": 0.101744,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.055161290322582e-05,
+      "loss": 0.1896,
+      "step": 6359
+    },
+    {
+      "epoch": 0.10176,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.055e-05,
+      "loss": 0.1826,
+      "step": 6360
+    },
+    {
+      "epoch": 0.101776,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.05483870967742e-05,
+      "loss": 0.1637,
+      "step": 6361
+    },
+    {
+      "epoch": 0.101792,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.054677419354839e-05,
+      "loss": 0.1748,
+      "step": 6362
+    },
+    {
+      "epoch": 0.101808,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.054516129032259e-05,
+      "loss": 0.1377,
+      "step": 6363
+    },
+    {
+      "epoch": 0.101824,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.054354838709678e-05,
+      "loss": 0.2019,
+      "step": 6364
+    },
+    {
+      "epoch": 0.10184,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.054193548387098e-05,
+      "loss": 0.2056,
+      "step": 6365
+    },
+    {
+      "epoch": 0.101856,
+      "grad_norm": 1.359375,
+      "learning_rate": 9.054032258064516e-05,
+      "loss": 0.1835,
+      "step": 6366
+    },
+    {
+      "epoch": 0.101872,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.053870967741936e-05,
+      "loss": 0.1687,
+      "step": 6367
+    },
+    {
+      "epoch": 0.101888,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.053709677419355e-05,
+      "loss": 0.166,
+      "step": 6368
+    },
+    {
+      "epoch": 0.101904,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.053548387096773e-05,
+      "loss": 0.1855,
+      "step": 6369
+    },
+    {
+      "epoch": 0.10192,
+      "grad_norm": 1.4375,
+      "learning_rate": 9.053387096774193e-05,
+      "loss": 0.1609,
+      "step": 6370
+    },
+    {
+      "epoch": 0.101936,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.053225806451613e-05,
+      "loss": 0.1632,
+      "step": 6371
+    },
+    {
+      "epoch": 0.101952,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.053064516129033e-05,
+      "loss": 0.1947,
+      "step": 6372
+    },
+    {
+      "epoch": 0.101968,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.052903225806452e-05,
+      "loss": 0.1993,
+      "step": 6373
+    },
+    {
+      "epoch": 0.101984,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.052741935483872e-05,
+      "loss": 0.1802,
+      "step": 6374
+    },
+    {
+      "epoch": 0.102,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.05258064516129e-05,
+      "loss": 0.1821,
+      "step": 6375
+    },
+    {
+      "epoch": 0.102016,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.05241935483871e-05,
+      "loss": 0.1335,
+      "step": 6376
+    },
+    {
+      "epoch": 0.102032,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.052258064516129e-05,
+      "loss": 0.1276,
+      "step": 6377
+    },
+    {
+      "epoch": 0.102048,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.052096774193549e-05,
+      "loss": 0.2414,
+      "step": 6378
+    },
+    {
+      "epoch": 0.102064,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.051935483870968e-05,
+      "loss": 0.1538,
+      "step": 6379
+    },
+    {
+      "epoch": 0.10208,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.051774193548388e-05,
+      "loss": 0.1903,
+      "step": 6380
+    },
+    {
+      "epoch": 0.102096,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.051612903225806e-05,
+      "loss": 0.1834,
+      "step": 6381
+    },
+    {
+      "epoch": 0.102112,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.051451612903226e-05,
+      "loss": 0.1636,
+      "step": 6382
+    },
+    {
+      "epoch": 0.102128,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.051290322580646e-05,
+      "loss": 0.1774,
+      "step": 6383
+    },
+    {
+      "epoch": 0.102144,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.051129032258065e-05,
+      "loss": 0.1815,
+      "step": 6384
+    },
+    {
+      "epoch": 0.10216,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.050967741935485e-05,
+      "loss": 0.1611,
+      "step": 6385
+    },
+    {
+      "epoch": 0.102176,
+      "grad_norm": 1.4140625,
+      "learning_rate": 9.050806451612903e-05,
+      "loss": 0.1757,
+      "step": 6386
+    },
+    {
+      "epoch": 0.102192,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.050645161290323e-05,
+      "loss": 0.1644,
+      "step": 6387
+    },
+    {
+      "epoch": 0.102208,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.050483870967742e-05,
+      "loss": 0.1509,
+      "step": 6388
+    },
+    {
+      "epoch": 0.102224,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.050322580645162e-05,
+      "loss": 0.1438,
+      "step": 6389
+    },
+    {
+      "epoch": 0.10224,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.05016129032258e-05,
+      "loss": 0.2145,
+      "step": 6390
+    },
+    {
+      "epoch": 0.102256,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.05e-05,
+      "loss": 0.1704,
+      "step": 6391
+    },
+    {
+      "epoch": 0.102272,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.049838709677419e-05,
+      "loss": 0.1749,
+      "step": 6392
+    },
+    {
+      "epoch": 0.102288,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.049677419354839e-05,
+      "loss": 0.174,
+      "step": 6393
+    },
+    {
+      "epoch": 0.102304,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.049516129032259e-05,
+      "loss": 0.1701,
+      "step": 6394
+    },
+    {
+      "epoch": 0.10232,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.049354838709679e-05,
+      "loss": 0.2119,
+      "step": 6395
+    },
+    {
+      "epoch": 0.102336,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.049193548387098e-05,
+      "loss": 0.2086,
+      "step": 6396
+    },
+    {
+      "epoch": 0.102352,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.049032258064517e-05,
+      "loss": 0.1744,
+      "step": 6397
+    },
+    {
+      "epoch": 0.102368,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.048870967741936e-05,
+      "loss": 0.1748,
+      "step": 6398
+    },
+    {
+      "epoch": 0.102384,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.048709677419356e-05,
+      "loss": 0.1557,
+      "step": 6399
+    },
+    {
+      "epoch": 0.1024,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.048548387096775e-05,
+      "loss": 0.1864,
+      "step": 6400
+    },
+    {
+      "epoch": 0.102416,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.048387096774193e-05,
+      "loss": 0.1745,
+      "step": 6401
+    },
+    {
+      "epoch": 0.102432,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.048225806451613e-05,
+      "loss": 0.1616,
+      "step": 6402
+    },
+    {
+      "epoch": 0.102448,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.048064516129032e-05,
+      "loss": 0.2272,
+      "step": 6403
+    },
+    {
+      "epoch": 0.102464,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.047903225806452e-05,
+      "loss": 0.1825,
+      "step": 6404
+    },
+    {
+      "epoch": 0.10248,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.04774193548387e-05,
+      "loss": 0.1875,
+      "step": 6405
+    },
+    {
+      "epoch": 0.102496,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.04758064516129e-05,
+      "loss": 0.1871,
+      "step": 6406
+    },
+    {
+      "epoch": 0.102512,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.04741935483871e-05,
+      "loss": 0.1337,
+      "step": 6407
+    },
+    {
+      "epoch": 0.102528,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.04725806451613e-05,
+      "loss": 0.2119,
+      "step": 6408
+    },
+    {
+      "epoch": 0.102544,
+      "grad_norm": 1.3828125,
+      "learning_rate": 9.047096774193549e-05,
+      "loss": 0.1951,
+      "step": 6409
+    },
+    {
+      "epoch": 0.10256,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.046935483870969e-05,
+      "loss": 0.1504,
+      "step": 6410
+    },
+    {
+      "epoch": 0.102576,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.046774193548387e-05,
+      "loss": 0.1819,
+      "step": 6411
+    },
+    {
+      "epoch": 0.102592,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.046612903225807e-05,
+      "loss": 0.1489,
+      "step": 6412
+    },
+    {
+      "epoch": 0.102608,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.046451612903226e-05,
+      "loss": 0.1895,
+      "step": 6413
+    },
+    {
+      "epoch": 0.102624,
+      "grad_norm": 0.453125,
+      "learning_rate": 9.046290322580646e-05,
+      "loss": 0.1628,
+      "step": 6414
+    },
+    {
+      "epoch": 0.10264,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.046129032258065e-05,
+      "loss": 0.1272,
+      "step": 6415
+    },
+    {
+      "epoch": 0.102656,
+      "grad_norm": 1.0,
+      "learning_rate": 9.045967741935483e-05,
+      "loss": 0.1563,
+      "step": 6416
+    },
+    {
+      "epoch": 0.102672,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.045806451612903e-05,
+      "loss": 0.1779,
+      "step": 6417
+    },
+    {
+      "epoch": 0.102688,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.045645161290323e-05,
+      "loss": 0.2201,
+      "step": 6418
+    },
+    {
+      "epoch": 0.102704,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.045483870967743e-05,
+      "loss": 0.1605,
+      "step": 6419
+    },
+    {
+      "epoch": 0.10272,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.045322580645162e-05,
+      "loss": 0.1872,
+      "step": 6420
+    },
+    {
+      "epoch": 0.102736,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.045161290322582e-05,
+      "loss": 0.2092,
+      "step": 6421
+    },
+    {
+      "epoch": 0.102752,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.045e-05,
+      "loss": 0.1628,
+      "step": 6422
+    },
+    {
+      "epoch": 0.102768,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.04483870967742e-05,
+      "loss": 0.1638,
+      "step": 6423
+    },
+    {
+      "epoch": 0.102784,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.044677419354839e-05,
+      "loss": 0.2271,
+      "step": 6424
+    },
+    {
+      "epoch": 0.1028,
+      "grad_norm": 1.5,
+      "learning_rate": 9.044516129032259e-05,
+      "loss": 0.1994,
+      "step": 6425
+    },
+    {
+      "epoch": 0.102816,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.044354838709677e-05,
+      "loss": 0.1299,
+      "step": 6426
+    },
+    {
+      "epoch": 0.102832,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.044193548387097e-05,
+      "loss": 0.1618,
+      "step": 6427
+    },
+    {
+      "epoch": 0.102848,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.044032258064516e-05,
+      "loss": 0.1578,
+      "step": 6428
+    },
+    {
+      "epoch": 0.102864,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.043870967741936e-05,
+      "loss": 0.2047,
+      "step": 6429
+    },
+    {
+      "epoch": 0.10288,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.043709677419355e-05,
+      "loss": 0.1392,
+      "step": 6430
+    },
+    {
+      "epoch": 0.102896,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.043548387096775e-05,
+      "loss": 0.1154,
+      "step": 6431
+    },
+    {
+      "epoch": 0.102912,
+      "grad_norm": 1.4296875,
+      "learning_rate": 9.043387096774195e-05,
+      "loss": 0.2144,
+      "step": 6432
+    },
+    {
+      "epoch": 0.102928,
+      "grad_norm": 0.875,
+      "learning_rate": 9.043225806451613e-05,
+      "loss": 0.1864,
+      "step": 6433
+    },
+    {
+      "epoch": 0.102944,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.043064516129033e-05,
+      "loss": 0.2193,
+      "step": 6434
+    },
+    {
+      "epoch": 0.10296,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.042903225806452e-05,
+      "loss": 0.1679,
+      "step": 6435
+    },
+    {
+      "epoch": 0.102976,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.042741935483872e-05,
+      "loss": 0.1583,
+      "step": 6436
+    },
+    {
+      "epoch": 0.102992,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.04258064516129e-05,
+      "loss": 0.1613,
+      "step": 6437
+    },
+    {
+      "epoch": 0.103008,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.04241935483871e-05,
+      "loss": 0.1615,
+      "step": 6438
+    },
+    {
+      "epoch": 0.103024,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.042258064516129e-05,
+      "loss": 0.1833,
+      "step": 6439
+    },
+    {
+      "epoch": 0.10304,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.042096774193549e-05,
+      "loss": 0.248,
+      "step": 6440
+    },
+    {
+      "epoch": 0.103056,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.041935483870967e-05,
+      "loss": 0.1786,
+      "step": 6441
+    },
+    {
+      "epoch": 0.103072,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.041774193548387e-05,
+      "loss": 0.1761,
+      "step": 6442
+    },
+    {
+      "epoch": 0.103088,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.041612903225807e-05,
+      "loss": 0.1687,
+      "step": 6443
+    },
+    {
+      "epoch": 0.103104,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.041451612903227e-05,
+      "loss": 0.1616,
+      "step": 6444
+    },
+    {
+      "epoch": 0.10312,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.041290322580646e-05,
+      "loss": 0.1949,
+      "step": 6445
+    },
+    {
+      "epoch": 0.103136,
+      "grad_norm": 1.4921875,
+      "learning_rate": 9.041129032258065e-05,
+      "loss": 0.2253,
+      "step": 6446
+    },
+    {
+      "epoch": 0.103152,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.040967741935484e-05,
+      "loss": 0.1557,
+      "step": 6447
+    },
+    {
+      "epoch": 0.103168,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.040806451612903e-05,
+      "loss": 0.1712,
+      "step": 6448
+    },
+    {
+      "epoch": 0.103184,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.040645161290323e-05,
+      "loss": 0.2208,
+      "step": 6449
+    },
+    {
+      "epoch": 0.1032,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.040483870967742e-05,
+      "loss": 0.2017,
+      "step": 6450
+    },
+    {
+      "epoch": 0.103216,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.040322580645162e-05,
+      "loss": 0.1994,
+      "step": 6451
+    },
+    {
+      "epoch": 0.103232,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.04016129032258e-05,
+      "loss": 0.167,
+      "step": 6452
+    },
+    {
+      "epoch": 0.103248,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.04e-05,
+      "loss": 0.1705,
+      "step": 6453
+    },
+    {
+      "epoch": 0.103264,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.03983870967742e-05,
+      "loss": 0.1895,
+      "step": 6454
+    },
+    {
+      "epoch": 0.10328,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.03967741935484e-05,
+      "loss": 0.2033,
+      "step": 6455
+    },
+    {
+      "epoch": 0.103296,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.039516129032259e-05,
+      "loss": 0.1558,
+      "step": 6456
+    },
+    {
+      "epoch": 0.103312,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.039354838709679e-05,
+      "loss": 0.1577,
+      "step": 6457
+    },
+    {
+      "epoch": 0.103328,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.039193548387097e-05,
+      "loss": 0.1678,
+      "step": 6458
+    },
+    {
+      "epoch": 0.103344,
+      "grad_norm": 1.4921875,
+      "learning_rate": 9.039032258064517e-05,
+      "loss": 0.1946,
+      "step": 6459
+    },
+    {
+      "epoch": 0.10336,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.038870967741936e-05,
+      "loss": 0.1663,
+      "step": 6460
+    },
+    {
+      "epoch": 0.103376,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.038709677419356e-05,
+      "loss": 0.1743,
+      "step": 6461
+    },
+    {
+      "epoch": 0.103392,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.038548387096774e-05,
+      "loss": 0.1537,
+      "step": 6462
+    },
+    {
+      "epoch": 0.103408,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.038387096774193e-05,
+      "loss": 0.1276,
+      "step": 6463
+    },
+    {
+      "epoch": 0.103424,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.038225806451613e-05,
+      "loss": 0.1838,
+      "step": 6464
+    },
+    {
+      "epoch": 0.10344,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.038064516129032e-05,
+      "loss": 0.1514,
+      "step": 6465
+    },
+    {
+      "epoch": 0.103456,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.037903225806452e-05,
+      "loss": 0.181,
+      "step": 6466
+    },
+    {
+      "epoch": 0.103472,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.037741935483872e-05,
+      "loss": 0.1669,
+      "step": 6467
+    },
+    {
+      "epoch": 0.103488,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.037580645161291e-05,
+      "loss": 0.1712,
+      "step": 6468
+    },
+    {
+      "epoch": 0.103504,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.03741935483871e-05,
+      "loss": 0.1557,
+      "step": 6469
+    },
+    {
+      "epoch": 0.10352,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.03725806451613e-05,
+      "loss": 0.1651,
+      "step": 6470
+    },
+    {
+      "epoch": 0.103536,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.037096774193549e-05,
+      "loss": 0.2017,
+      "step": 6471
+    },
+    {
+      "epoch": 0.103552,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.036935483870969e-05,
+      "loss": 0.1431,
+      "step": 6472
+    },
+    {
+      "epoch": 0.103568,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.036774193548387e-05,
+      "loss": 0.1729,
+      "step": 6473
+    },
+    {
+      "epoch": 0.103584,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.036612903225807e-05,
+      "loss": 0.1584,
+      "step": 6474
+    },
+    {
+      "epoch": 0.1036,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.036451612903226e-05,
+      "loss": 0.1673,
+      "step": 6475
+    },
+    {
+      "epoch": 0.103616,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.036290322580646e-05,
+      "loss": 0.1356,
+      "step": 6476
+    },
+    {
+      "epoch": 0.103632,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.036129032258064e-05,
+      "loss": 0.1529,
+      "step": 6477
+    },
+    {
+      "epoch": 0.103648,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.035967741935484e-05,
+      "loss": 0.2063,
+      "step": 6478
+    },
+    {
+      "epoch": 0.103664,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.035806451612904e-05,
+      "loss": 0.1962,
+      "step": 6479
+    },
+    {
+      "epoch": 0.10368,
+      "grad_norm": 1.8125,
+      "learning_rate": 9.035645161290323e-05,
+      "loss": 0.2151,
+      "step": 6480
+    },
+    {
+      "epoch": 0.103696,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.035483870967743e-05,
+      "loss": 0.1781,
+      "step": 6481
+    },
+    {
+      "epoch": 0.103712,
+      "grad_norm": 1.25,
+      "learning_rate": 9.035322580645161e-05,
+      "loss": 0.1548,
+      "step": 6482
+    },
+    {
+      "epoch": 0.103728,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.035161290322581e-05,
+      "loss": 0.16,
+      "step": 6483
+    },
+    {
+      "epoch": 0.103744,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.035e-05,
+      "loss": 0.1416,
+      "step": 6484
+    },
+    {
+      "epoch": 0.10376,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.03483870967742e-05,
+      "loss": 0.1565,
+      "step": 6485
+    },
+    {
+      "epoch": 0.103776,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.034677419354839e-05,
+      "loss": 0.2127,
+      "step": 6486
+    },
+    {
+      "epoch": 0.103792,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.034516129032259e-05,
+      "loss": 0.1821,
+      "step": 6487
+    },
+    {
+      "epoch": 0.103808,
+      "grad_norm": 1.3515625,
+      "learning_rate": 9.034354838709677e-05,
+      "loss": 0.2078,
+      "step": 6488
+    },
+    {
+      "epoch": 0.103824,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.034193548387097e-05,
+      "loss": 0.1985,
+      "step": 6489
+    },
+    {
+      "epoch": 0.10384,
+      "grad_norm": 1.7890625,
+      "learning_rate": 9.034032258064517e-05,
+      "loss": 0.2126,
+      "step": 6490
+    },
+    {
+      "epoch": 0.103856,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.033870967741937e-05,
+      "loss": 0.1512,
+      "step": 6491
+    },
+    {
+      "epoch": 0.103872,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.033709677419356e-05,
+      "loss": 0.1615,
+      "step": 6492
+    },
+    {
+      "epoch": 0.103888,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.033548387096774e-05,
+      "loss": 0.1716,
+      "step": 6493
+    },
+    {
+      "epoch": 0.103904,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.033387096774194e-05,
+      "loss": 0.1962,
+      "step": 6494
+    },
+    {
+      "epoch": 0.10392,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.033225806451613e-05,
+      "loss": 0.1775,
+      "step": 6495
+    },
+    {
+      "epoch": 0.103936,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.033064516129033e-05,
+      "loss": 0.1833,
+      "step": 6496
+    },
+    {
+      "epoch": 0.103952,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.032903225806451e-05,
+      "loss": 0.2017,
+      "step": 6497
+    },
+    {
+      "epoch": 0.103968,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.032741935483871e-05,
+      "loss": 0.1326,
+      "step": 6498
+    },
+    {
+      "epoch": 0.103984,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.03258064516129e-05,
+      "loss": 0.1447,
+      "step": 6499
+    },
+    {
+      "epoch": 0.104,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.03241935483871e-05,
+      "loss": 0.1618,
+      "step": 6500
+    },
+    {
+      "epoch": 0.104016,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.032258064516129e-05,
+      "loss": 0.1587,
+      "step": 6501
+    },
+    {
+      "epoch": 0.104032,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.032096774193549e-05,
+      "loss": 0.1936,
+      "step": 6502
+    },
+    {
+      "epoch": 0.104048,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.031935483870969e-05,
+      "loss": 0.1676,
+      "step": 6503
+    },
+    {
+      "epoch": 0.104064,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.031774193548388e-05,
+      "loss": 0.157,
+      "step": 6504
+    },
+    {
+      "epoch": 0.10408,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.031612903225807e-05,
+      "loss": 0.1966,
+      "step": 6505
+    },
+    {
+      "epoch": 0.104096,
+      "grad_norm": 0.5078125,
+      "learning_rate": 9.031451612903227e-05,
+      "loss": 0.158,
+      "step": 6506
+    },
+    {
+      "epoch": 0.104112,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.031290322580646e-05,
+      "loss": 0.188,
+      "step": 6507
+    },
+    {
+      "epoch": 0.104128,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.031129032258066e-05,
+      "loss": 0.1675,
+      "step": 6508
+    },
+    {
+      "epoch": 0.104144,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.030967741935484e-05,
+      "loss": 0.1396,
+      "step": 6509
+    },
+    {
+      "epoch": 0.10416,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.030806451612903e-05,
+      "loss": 0.1458,
+      "step": 6510
+    },
+    {
+      "epoch": 0.104176,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.030645161290323e-05,
+      "loss": 0.2115,
+      "step": 6511
+    },
+    {
+      "epoch": 0.104192,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.030483870967741e-05,
+      "loss": 0.1678,
+      "step": 6512
+    },
+    {
+      "epoch": 0.104208,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.030322580645161e-05,
+      "loss": 0.1594,
+      "step": 6513
+    },
+    {
+      "epoch": 0.104224,
+      "grad_norm": 1.0,
+      "learning_rate": 9.030161290322581e-05,
+      "loss": 0.1484,
+      "step": 6514
+    },
+    {
+      "epoch": 0.10424,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.030000000000001e-05,
+      "loss": 0.1668,
+      "step": 6515
+    },
+    {
+      "epoch": 0.104256,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.02983870967742e-05,
+      "loss": 0.2031,
+      "step": 6516
+    },
+    {
+      "epoch": 0.104272,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.02967741935484e-05,
+      "loss": 0.2165,
+      "step": 6517
+    },
+    {
+      "epoch": 0.104288,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.029516129032258e-05,
+      "loss": 0.1914,
+      "step": 6518
+    },
+    {
+      "epoch": 0.104304,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.029354838709678e-05,
+      "loss": 0.224,
+      "step": 6519
+    },
+    {
+      "epoch": 0.10432,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.029193548387097e-05,
+      "loss": 0.215,
+      "step": 6520
+    },
+    {
+      "epoch": 0.104336,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.029032258064517e-05,
+      "loss": 0.1678,
+      "step": 6521
+    },
+    {
+      "epoch": 0.104352,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.028870967741936e-05,
+      "loss": 0.1613,
+      "step": 6522
+    },
+    {
+      "epoch": 0.104368,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.028709677419356e-05,
+      "loss": 0.1624,
+      "step": 6523
+    },
+    {
+      "epoch": 0.104384,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.028548387096774e-05,
+      "loss": 0.2028,
+      "step": 6524
+    },
+    {
+      "epoch": 0.1044,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.028387096774194e-05,
+      "loss": 0.1731,
+      "step": 6525
+    },
+    {
+      "epoch": 0.104416,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.028225806451613e-05,
+      "loss": 0.1639,
+      "step": 6526
+    },
+    {
+      "epoch": 0.104432,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.028064516129033e-05,
+      "loss": 0.1467,
+      "step": 6527
+    },
+    {
+      "epoch": 0.104448,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.027903225806453e-05,
+      "loss": 0.1909,
+      "step": 6528
+    },
+    {
+      "epoch": 0.104464,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.027741935483871e-05,
+      "loss": 0.1754,
+      "step": 6529
+    },
+    {
+      "epoch": 0.10448,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.027580645161291e-05,
+      "loss": 0.2116,
+      "step": 6530
+    },
+    {
+      "epoch": 0.104496,
+      "grad_norm": 0.75,
+      "learning_rate": 9.02741935483871e-05,
+      "loss": 0.187,
+      "step": 6531
+    },
+    {
+      "epoch": 0.104512,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.02725806451613e-05,
+      "loss": 0.1784,
+      "step": 6532
+    },
+    {
+      "epoch": 0.104528,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.027096774193548e-05,
+      "loss": 0.18,
+      "step": 6533
+    },
+    {
+      "epoch": 0.104544,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.026935483870968e-05,
+      "loss": 0.1471,
+      "step": 6534
+    },
+    {
+      "epoch": 0.10456,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.026774193548387e-05,
+      "loss": 0.1951,
+      "step": 6535
+    },
+    {
+      "epoch": 0.104576,
+      "grad_norm": 0.48828125,
+      "learning_rate": 9.026612903225807e-05,
+      "loss": 0.1363,
+      "step": 6536
+    },
+    {
+      "epoch": 0.104592,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.026451612903226e-05,
+      "loss": 0.1957,
+      "step": 6537
+    },
+    {
+      "epoch": 0.104608,
+      "grad_norm": 0.625,
+      "learning_rate": 9.026290322580646e-05,
+      "loss": 0.1635,
+      "step": 6538
+    },
+    {
+      "epoch": 0.104624,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.026129032258065e-05,
+      "loss": 0.1499,
+      "step": 6539
+    },
+    {
+      "epoch": 0.10464,
+      "grad_norm": 1.40625,
+      "learning_rate": 9.025967741935484e-05,
+      "loss": 0.2084,
+      "step": 6540
+    },
+    {
+      "epoch": 0.104656,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.025806451612904e-05,
+      "loss": 0.193,
+      "step": 6541
+    },
+    {
+      "epoch": 0.104672,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.025645161290323e-05,
+      "loss": 0.1922,
+      "step": 6542
+    },
+    {
+      "epoch": 0.104688,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.025483870967743e-05,
+      "loss": 0.1833,
+      "step": 6543
+    },
+    {
+      "epoch": 0.104704,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.025322580645161e-05,
+      "loss": 0.1757,
+      "step": 6544
+    },
+    {
+      "epoch": 0.10472,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.025161290322581e-05,
+      "loss": 0.2048,
+      "step": 6545
+    },
+    {
+      "epoch": 0.104736,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.025e-05,
+      "loss": 0.1928,
+      "step": 6546
+    },
+    {
+      "epoch": 0.104752,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.02483870967742e-05,
+      "loss": 0.2162,
+      "step": 6547
+    },
+    {
+      "epoch": 0.104768,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.024677419354838e-05,
+      "loss": 0.1771,
+      "step": 6548
+    },
+    {
+      "epoch": 0.104784,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.024516129032258e-05,
+      "loss": 0.1721,
+      "step": 6549
+    },
+    {
+      "epoch": 0.1048,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.024354838709678e-05,
+      "loss": 0.1983,
+      "step": 6550
+    },
+    {
+      "epoch": 0.104816,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.024193548387098e-05,
+      "loss": 0.2093,
+      "step": 6551
+    },
+    {
+      "epoch": 0.104832,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.024032258064517e-05,
+      "loss": 0.188,
+      "step": 6552
+    },
+    {
+      "epoch": 0.104848,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.023870967741937e-05,
+      "loss": 0.1996,
+      "step": 6553
+    },
+    {
+      "epoch": 0.104864,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.023709677419355e-05,
+      "loss": 0.2192,
+      "step": 6554
+    },
+    {
+      "epoch": 0.10488,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.023548387096774e-05,
+      "loss": 0.1839,
+      "step": 6555
+    },
+    {
+      "epoch": 0.104896,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.023387096774194e-05,
+      "loss": 0.1577,
+      "step": 6556
+    },
+    {
+      "epoch": 0.104912,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.023225806451613e-05,
+      "loss": 0.1896,
+      "step": 6557
+    },
+    {
+      "epoch": 0.104928,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.023064516129033e-05,
+      "loss": 0.1909,
+      "step": 6558
+    },
+    {
+      "epoch": 0.104944,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.022903225806451e-05,
+      "loss": 0.1838,
+      "step": 6559
+    },
+    {
+      "epoch": 0.10496,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.022741935483871e-05,
+      "loss": 0.1993,
+      "step": 6560
+    },
+    {
+      "epoch": 0.104976,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.02258064516129e-05,
+      "loss": 0.1594,
+      "step": 6561
+    },
+    {
+      "epoch": 0.104992,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.02241935483871e-05,
+      "loss": 0.1862,
+      "step": 6562
+    },
+    {
+      "epoch": 0.105008,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.02225806451613e-05,
+      "loss": 0.1941,
+      "step": 6563
+    },
+    {
+      "epoch": 0.105024,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.02209677419355e-05,
+      "loss": 0.1428,
+      "step": 6564
+    },
+    {
+      "epoch": 0.10504,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.021935483870968e-05,
+      "loss": 0.2243,
+      "step": 6565
+    },
+    {
+      "epoch": 0.105056,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.021774193548388e-05,
+      "loss": 0.2169,
+      "step": 6566
+    },
+    {
+      "epoch": 0.105072,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.021612903225807e-05,
+      "loss": 0.1591,
+      "step": 6567
+    },
+    {
+      "epoch": 0.105088,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.021451612903227e-05,
+      "loss": 0.1722,
+      "step": 6568
+    },
+    {
+      "epoch": 0.105104,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.021290322580645e-05,
+      "loss": 0.1828,
+      "step": 6569
+    },
+    {
+      "epoch": 0.10512,
+      "grad_norm": 1.5703125,
+      "learning_rate": 9.021129032258065e-05,
+      "loss": 0.1795,
+      "step": 6570
+    },
+    {
+      "epoch": 0.105136,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.020967741935484e-05,
+      "loss": 0.1851,
+      "step": 6571
+    },
+    {
+      "epoch": 0.105152,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.020806451612903e-05,
+      "loss": 0.164,
+      "step": 6572
+    },
+    {
+      "epoch": 0.105168,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.020645161290323e-05,
+      "loss": 0.1942,
+      "step": 6573
+    },
+    {
+      "epoch": 0.105184,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.020483870967743e-05,
+      "loss": 0.1662,
+      "step": 6574
+    },
+    {
+      "epoch": 0.1052,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.020322580645162e-05,
+      "loss": 0.1676,
+      "step": 6575
+    },
+    {
+      "epoch": 0.105216,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.020161290322581e-05,
+      "loss": 0.1522,
+      "step": 6576
+    },
+    {
+      "epoch": 0.105232,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.020000000000001e-05,
+      "loss": 0.1837,
+      "step": 6577
+    },
+    {
+      "epoch": 0.105248,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.01983870967742e-05,
+      "loss": 0.165,
+      "step": 6578
+    },
+    {
+      "epoch": 0.105264,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.01967741935484e-05,
+      "loss": 0.2135,
+      "step": 6579
+    },
+    {
+      "epoch": 0.10528,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.019516129032258e-05,
+      "loss": 0.1963,
+      "step": 6580
+    },
+    {
+      "epoch": 0.105296,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.019354838709678e-05,
+      "loss": 0.187,
+      "step": 6581
+    },
+    {
+      "epoch": 0.105312,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.019193548387097e-05,
+      "loss": 0.1833,
+      "step": 6582
+    },
+    {
+      "epoch": 0.105328,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.019032258064517e-05,
+      "loss": 0.158,
+      "step": 6583
+    },
+    {
+      "epoch": 0.105344,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.018870967741935e-05,
+      "loss": 0.1655,
+      "step": 6584
+    },
+    {
+      "epoch": 0.10536,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.018709677419355e-05,
+      "loss": 0.2092,
+      "step": 6585
+    },
+    {
+      "epoch": 0.105376,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.018548387096775e-05,
+      "loss": 0.1966,
+      "step": 6586
+    },
+    {
+      "epoch": 0.105392,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.018387096774194e-05,
+      "loss": 0.195,
+      "step": 6587
+    },
+    {
+      "epoch": 0.105408,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.018225806451614e-05,
+      "loss": 0.1791,
+      "step": 6588
+    },
+    {
+      "epoch": 0.105424,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.018064516129032e-05,
+      "loss": 0.1927,
+      "step": 6589
+    },
+    {
+      "epoch": 0.10544,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.017903225806452e-05,
+      "loss": 0.1721,
+      "step": 6590
+    },
+    {
+      "epoch": 0.105456,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.017741935483871e-05,
+      "loss": 0.1715,
+      "step": 6591
+    },
+    {
+      "epoch": 0.105472,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.017580645161291e-05,
+      "loss": 0.1618,
+      "step": 6592
+    },
+    {
+      "epoch": 0.105488,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.01741935483871e-05,
+      "loss": 0.1905,
+      "step": 6593
+    },
+    {
+      "epoch": 0.105504,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.01725806451613e-05,
+      "loss": 0.1777,
+      "step": 6594
+    },
+    {
+      "epoch": 0.10552,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.017096774193548e-05,
+      "loss": 0.1676,
+      "step": 6595
+    },
+    {
+      "epoch": 0.105536,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.016935483870968e-05,
+      "loss": 0.1726,
+      "step": 6596
+    },
+    {
+      "epoch": 0.105552,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.016774193548387e-05,
+      "loss": 0.1781,
+      "step": 6597
+    },
+    {
+      "epoch": 0.105568,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.016612903225807e-05,
+      "loss": 0.1285,
+      "step": 6598
+    },
+    {
+      "epoch": 0.105584,
+      "grad_norm": 0.625,
+      "learning_rate": 9.016451612903227e-05,
+      "loss": 0.1476,
+      "step": 6599
+    },
+    {
+      "epoch": 0.1056,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.016290322580647e-05,
+      "loss": 0.2011,
+      "step": 6600
+    },
+    {
+      "epoch": 0.105616,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.016129032258065e-05,
+      "loss": 0.1509,
+      "step": 6601
+    },
+    {
+      "epoch": 0.105632,
+      "grad_norm": 0.46484375,
+      "learning_rate": 9.015967741935484e-05,
+      "loss": 0.1457,
+      "step": 6602
+    },
+    {
+      "epoch": 0.105648,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.015806451612904e-05,
+      "loss": 0.1991,
+      "step": 6603
+    },
+    {
+      "epoch": 0.105664,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.015645161290322e-05,
+      "loss": 0.1649,
+      "step": 6604
+    },
+    {
+      "epoch": 0.10568,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.015483870967742e-05,
+      "loss": 0.2096,
+      "step": 6605
+    },
+    {
+      "epoch": 0.105696,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.015322580645161e-05,
+      "loss": 0.138,
+      "step": 6606
+    },
+    {
+      "epoch": 0.105712,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.015161290322581e-05,
+      "loss": 0.1855,
+      "step": 6607
+    },
+    {
+      "epoch": 0.105728,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.015e-05,
+      "loss": 0.1578,
+      "step": 6608
+    },
+    {
+      "epoch": 0.105744,
+      "grad_norm": 1.3125,
+      "learning_rate": 9.01483870967742e-05,
+      "loss": 0.1628,
+      "step": 6609
+    },
+    {
+      "epoch": 0.10576,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.01467741935484e-05,
+      "loss": 0.1829,
+      "step": 6610
+    },
+    {
+      "epoch": 0.105776,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.01451612903226e-05,
+      "loss": 0.2018,
+      "step": 6611
+    },
+    {
+      "epoch": 0.105792,
+      "grad_norm": 0.875,
+      "learning_rate": 9.014354838709678e-05,
+      "loss": 0.2042,
+      "step": 6612
+    },
+    {
+      "epoch": 0.105808,
+      "grad_norm": 1.25,
+      "learning_rate": 9.014193548387098e-05,
+      "loss": 0.1887,
+      "step": 6613
+    },
+    {
+      "epoch": 0.105824,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.014032258064517e-05,
+      "loss": 0.1892,
+      "step": 6614
+    },
+    {
+      "epoch": 0.10584,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.013870967741937e-05,
+      "loss": 0.1933,
+      "step": 6615
+    },
+    {
+      "epoch": 0.105856,
+      "grad_norm": 1.0,
+      "learning_rate": 9.013709677419355e-05,
+      "loss": 0.1484,
+      "step": 6616
+    },
+    {
+      "epoch": 0.105872,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.013548387096774e-05,
+      "loss": 0.1474,
+      "step": 6617
+    },
+    {
+      "epoch": 0.105888,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.013387096774194e-05,
+      "loss": 0.1654,
+      "step": 6618
+    },
+    {
+      "epoch": 0.105904,
+      "grad_norm": 1.359375,
+      "learning_rate": 9.013225806451612e-05,
+      "loss": 0.1805,
+      "step": 6619
+    },
+    {
+      "epoch": 0.10592,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.013064516129032e-05,
+      "loss": 0.1626,
+      "step": 6620
+    },
+    {
+      "epoch": 0.105936,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.012903225806451e-05,
+      "loss": 0.1679,
+      "step": 6621
+    },
+    {
+      "epoch": 0.105952,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.012741935483871e-05,
+      "loss": 0.1991,
+      "step": 6622
+    },
+    {
+      "epoch": 0.105968,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.012580645161291e-05,
+      "loss": 0.1707,
+      "step": 6623
+    },
+    {
+      "epoch": 0.105984,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.012419354838711e-05,
+      "loss": 0.219,
+      "step": 6624
+    },
+    {
+      "epoch": 0.106,
+      "grad_norm": 1.5078125,
+      "learning_rate": 9.01225806451613e-05,
+      "loss": 0.2014,
+      "step": 6625
+    },
+    {
+      "epoch": 0.106016,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.01209677419355e-05,
+      "loss": 0.1529,
+      "step": 6626
+    },
+    {
+      "epoch": 0.106032,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.011935483870968e-05,
+      "loss": 0.1957,
+      "step": 6627
+    },
+    {
+      "epoch": 0.106048,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.011774193548388e-05,
+      "loss": 0.188,
+      "step": 6628
+    },
+    {
+      "epoch": 0.106064,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.011612903225807e-05,
+      "loss": 0.1728,
+      "step": 6629
+    },
+    {
+      "epoch": 0.10608,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.011451612903227e-05,
+      "loss": 0.2003,
+      "step": 6630
+    },
+    {
+      "epoch": 0.106096,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.011290322580645e-05,
+      "loss": 0.184,
+      "step": 6631
+    },
+    {
+      "epoch": 0.106112,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.011129032258065e-05,
+      "loss": 0.1578,
+      "step": 6632
+    },
+    {
+      "epoch": 0.106128,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.010967741935484e-05,
+      "loss": 0.2216,
+      "step": 6633
+    },
+    {
+      "epoch": 0.106144,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.010806451612904e-05,
+      "loss": 0.1664,
+      "step": 6634
+    },
+    {
+      "epoch": 0.10616,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.010645161290324e-05,
+      "loss": 0.2108,
+      "step": 6635
+    },
+    {
+      "epoch": 0.106176,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.010483870967742e-05,
+      "loss": 0.1611,
+      "step": 6636
+    },
+    {
+      "epoch": 0.106192,
+      "grad_norm": 1.34375,
+      "learning_rate": 9.010322580645162e-05,
+      "loss": 0.1264,
+      "step": 6637
+    },
+    {
+      "epoch": 0.106208,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.010161290322581e-05,
+      "loss": 0.1584,
+      "step": 6638
+    },
+    {
+      "epoch": 0.106224,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.010000000000001e-05,
+      "loss": 0.1541,
+      "step": 6639
+    },
+    {
+      "epoch": 0.10624,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.00983870967742e-05,
+      "loss": 0.1278,
+      "step": 6640
+    },
+    {
+      "epoch": 0.106256,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.00967741935484e-05,
+      "loss": 0.1788,
+      "step": 6641
+    },
+    {
+      "epoch": 0.106272,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.009516129032258e-05,
+      "loss": 0.1904,
+      "step": 6642
+    },
+    {
+      "epoch": 0.106288,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.009354838709678e-05,
+      "loss": 0.1945,
+      "step": 6643
+    },
+    {
+      "epoch": 0.106304,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.009193548387097e-05,
+      "loss": 0.2313,
+      "step": 6644
+    },
+    {
+      "epoch": 0.10632,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.009032258064517e-05,
+      "loss": 0.2015,
+      "step": 6645
+    },
+    {
+      "epoch": 0.106336,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.008870967741936e-05,
+      "loss": 0.1967,
+      "step": 6646
+    },
+    {
+      "epoch": 0.106352,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.008709677419356e-05,
+      "loss": 0.2053,
+      "step": 6647
+    },
+    {
+      "epoch": 0.106368,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.008548387096775e-05,
+      "loss": 0.1879,
+      "step": 6648
+    },
+    {
+      "epoch": 0.106384,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.008387096774194e-05,
+      "loss": 0.2122,
+      "step": 6649
+    },
+    {
+      "epoch": 0.1064,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.008225806451614e-05,
+      "loss": 0.1569,
+      "step": 6650
+    },
+    {
+      "epoch": 0.106416,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.008064516129032e-05,
+      "loss": 0.2115,
+      "step": 6651
+    },
+    {
+      "epoch": 0.106432,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.007903225806452e-05,
+      "loss": 0.1942,
+      "step": 6652
+    },
+    {
+      "epoch": 0.106448,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.007741935483871e-05,
+      "loss": 0.1669,
+      "step": 6653
+    },
+    {
+      "epoch": 0.106464,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.007580645161291e-05,
+      "loss": 0.2116,
+      "step": 6654
+    },
+    {
+      "epoch": 0.10648,
+      "grad_norm": 1.3515625,
+      "learning_rate": 9.00741935483871e-05,
+      "loss": 0.1854,
+      "step": 6655
+    },
+    {
+      "epoch": 0.106496,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.00725806451613e-05,
+      "loss": 0.187,
+      "step": 6656
+    },
+    {
+      "epoch": 0.106512,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.007096774193548e-05,
+      "loss": 0.1499,
+      "step": 6657
+    },
+    {
+      "epoch": 0.106528,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.006935483870968e-05,
+      "loss": 0.1911,
+      "step": 6658
+    },
+    {
+      "epoch": 0.106544,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.006774193548388e-05,
+      "loss": 0.167,
+      "step": 6659
+    },
+    {
+      "epoch": 0.10656,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.006612903225808e-05,
+      "loss": 0.2066,
+      "step": 6660
+    },
+    {
+      "epoch": 0.106576,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.006451612903226e-05,
+      "loss": 0.1644,
+      "step": 6661
+    },
+    {
+      "epoch": 0.106592,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.006290322580646e-05,
+      "loss": 0.1778,
+      "step": 6662
+    },
+    {
+      "epoch": 0.106608,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.006129032258065e-05,
+      "loss": 0.1557,
+      "step": 6663
+    },
+    {
+      "epoch": 0.106624,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.005967741935484e-05,
+      "loss": 0.1362,
+      "step": 6664
+    },
+    {
+      "epoch": 0.10664,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.005806451612904e-05,
+      "loss": 0.1873,
+      "step": 6665
+    },
+    {
+      "epoch": 0.106656,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.005645161290322e-05,
+      "loss": 0.1845,
+      "step": 6666
+    },
+    {
+      "epoch": 0.106672,
+      "grad_norm": 1.7578125,
+      "learning_rate": 9.005483870967742e-05,
+      "loss": 0.2364,
+      "step": 6667
+    },
+    {
+      "epoch": 0.106688,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.005322580645161e-05,
+      "loss": 0.1747,
+      "step": 6668
+    },
+    {
+      "epoch": 0.106704,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.005161290322581e-05,
+      "loss": 0.1825,
+      "step": 6669
+    },
+    {
+      "epoch": 0.10672,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.005000000000001e-05,
+      "loss": 0.2082,
+      "step": 6670
+    },
+    {
+      "epoch": 0.106736,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.00483870967742e-05,
+      "loss": 0.1786,
+      "step": 6671
+    },
+    {
+      "epoch": 0.106752,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.004677419354839e-05,
+      "loss": 0.206,
+      "step": 6672
+    },
+    {
+      "epoch": 0.106768,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.004516129032259e-05,
+      "loss": 0.1718,
+      "step": 6673
+    },
+    {
+      "epoch": 0.106784,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.004354838709678e-05,
+      "loss": 0.147,
+      "step": 6674
+    },
+    {
+      "epoch": 0.1068,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.004193548387098e-05,
+      "loss": 0.1555,
+      "step": 6675
+    },
+    {
+      "epoch": 0.106816,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.004032258064516e-05,
+      "loss": 0.1827,
+      "step": 6676
+    },
+    {
+      "epoch": 0.106832,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.003870967741936e-05,
+      "loss": 0.1941,
+      "step": 6677
+    },
+    {
+      "epoch": 0.106848,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.003709677419355e-05,
+      "loss": 0.2028,
+      "step": 6678
+    },
+    {
+      "epoch": 0.106864,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.003548387096775e-05,
+      "loss": 0.149,
+      "step": 6679
+    },
+    {
+      "epoch": 0.10688,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.003387096774194e-05,
+      "loss": 0.1914,
+      "step": 6680
+    },
+    {
+      "epoch": 0.106896,
+      "grad_norm": 1.0,
+      "learning_rate": 9.003225806451614e-05,
+      "loss": 0.2235,
+      "step": 6681
+    },
+    {
+      "epoch": 0.106912,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.003064516129032e-05,
+      "loss": 0.1661,
+      "step": 6682
+    },
+    {
+      "epoch": 0.106928,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.002903225806452e-05,
+      "loss": 0.1905,
+      "step": 6683
+    },
+    {
+      "epoch": 0.106944,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.002741935483872e-05,
+      "loss": 0.146,
+      "step": 6684
+    },
+    {
+      "epoch": 0.10696,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.00258064516129e-05,
+      "loss": 0.1834,
+      "step": 6685
+    },
+    {
+      "epoch": 0.106976,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.00241935483871e-05,
+      "loss": 0.1948,
+      "step": 6686
+    },
+    {
+      "epoch": 0.106992,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.002258064516129e-05,
+      "loss": 0.1848,
+      "step": 6687
+    },
+    {
+      "epoch": 0.107008,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.002096774193549e-05,
+      "loss": 0.1612,
+      "step": 6688
+    },
+    {
+      "epoch": 0.107024,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.001935483870968e-05,
+      "loss": 0.1942,
+      "step": 6689
+    },
+    {
+      "epoch": 0.10704,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.001774193548388e-05,
+      "loss": 0.1825,
+      "step": 6690
+    },
+    {
+      "epoch": 0.107056,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.001612903225806e-05,
+      "loss": 0.157,
+      "step": 6691
+    },
+    {
+      "epoch": 0.107072,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.001451612903226e-05,
+      "loss": 0.1969,
+      "step": 6692
+    },
+    {
+      "epoch": 0.107088,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.001290322580645e-05,
+      "loss": 0.2219,
+      "step": 6693
+    },
+    {
+      "epoch": 0.107104,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.001129032258065e-05,
+      "loss": 0.1861,
+      "step": 6694
+    },
+    {
+      "epoch": 0.10712,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.000967741935485e-05,
+      "loss": 0.2049,
+      "step": 6695
+    },
+    {
+      "epoch": 0.107136,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.000806451612903e-05,
+      "loss": 0.1661,
+      "step": 6696
+    },
+    {
+      "epoch": 0.107152,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.000645161290323e-05,
+      "loss": 0.1404,
+      "step": 6697
+    },
+    {
+      "epoch": 0.107168,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.000483870967742e-05,
+      "loss": 0.2318,
+      "step": 6698
+    },
+    {
+      "epoch": 0.107184,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.000322580645162e-05,
+      "loss": 0.1627,
+      "step": 6699
+    },
+    {
+      "epoch": 0.1072,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.00016129032258e-05,
+      "loss": 0.1831,
+      "step": 6700
+    },
+    {
+      "epoch": 0.107216,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9e-05,
+      "loss": 0.1608,
+      "step": 6701
+    },
+    {
+      "epoch": 0.107232,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.999838709677419e-05,
+      "loss": 0.1693,
+      "step": 6702
+    },
+    {
+      "epoch": 0.107248,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.999677419354839e-05,
+      "loss": 0.1635,
+      "step": 6703
+    },
+    {
+      "epoch": 0.107264,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.999516129032258e-05,
+      "loss": 0.1949,
+      "step": 6704
+    },
+    {
+      "epoch": 0.10728,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.999354838709678e-05,
+      "loss": 0.2034,
+      "step": 6705
+    },
+    {
+      "epoch": 0.107296,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.999193548387098e-05,
+      "loss": 0.1579,
+      "step": 6706
+    },
+    {
+      "epoch": 0.107312,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.999032258064518e-05,
+      "loss": 0.1988,
+      "step": 6707
+    },
+    {
+      "epoch": 0.107328,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.998870967741936e-05,
+      "loss": 0.1553,
+      "step": 6708
+    },
+    {
+      "epoch": 0.107344,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.998709677419356e-05,
+      "loss": 0.1324,
+      "step": 6709
+    },
+    {
+      "epoch": 0.10736,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.998548387096775e-05,
+      "loss": 0.1536,
+      "step": 6710
+    },
+    {
+      "epoch": 0.107376,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.998387096774193e-05,
+      "loss": 0.2035,
+      "step": 6711
+    },
+    {
+      "epoch": 0.107392,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.998225806451613e-05,
+      "loss": 0.1622,
+      "step": 6712
+    },
+    {
+      "epoch": 0.107408,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.998064516129032e-05,
+      "loss": 0.2198,
+      "step": 6713
+    },
+    {
+      "epoch": 0.107424,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.997903225806452e-05,
+      "loss": 0.1722,
+      "step": 6714
+    },
+    {
+      "epoch": 0.10744,
+      "grad_norm": 0.75,
+      "learning_rate": 8.99774193548387e-05,
+      "loss": 0.1469,
+      "step": 6715
+    },
+    {
+      "epoch": 0.107456,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.99758064516129e-05,
+      "loss": 0.1873,
+      "step": 6716
+    },
+    {
+      "epoch": 0.107472,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.997419354838709e-05,
+      "loss": 0.1703,
+      "step": 6717
+    },
+    {
+      "epoch": 0.107488,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.997258064516129e-05,
+      "loss": 0.199,
+      "step": 6718
+    },
+    {
+      "epoch": 0.107504,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.997096774193549e-05,
+      "loss": 0.154,
+      "step": 6719
+    },
+    {
+      "epoch": 0.10752,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.996935483870969e-05,
+      "loss": 0.1516,
+      "step": 6720
+    },
+    {
+      "epoch": 0.107536,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.996774193548388e-05,
+      "loss": 0.119,
+      "step": 6721
+    },
+    {
+      "epoch": 0.107552,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.996612903225808e-05,
+      "loss": 0.1488,
+      "step": 6722
+    },
+    {
+      "epoch": 0.107568,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.996451612903226e-05,
+      "loss": 0.1971,
+      "step": 6723
+    },
+    {
+      "epoch": 0.107584,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.996290322580646e-05,
+      "loss": 0.2122,
+      "step": 6724
+    },
+    {
+      "epoch": 0.1076,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.996129032258065e-05,
+      "loss": 0.1845,
+      "step": 6725
+    },
+    {
+      "epoch": 0.107616,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.995967741935483e-05,
+      "loss": 0.1589,
+      "step": 6726
+    },
+    {
+      "epoch": 0.107632,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.995806451612903e-05,
+      "loss": 0.1536,
+      "step": 6727
+    },
+    {
+      "epoch": 0.107648,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.995645161290322e-05,
+      "loss": 0.2136,
+      "step": 6728
+    },
+    {
+      "epoch": 0.107664,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.995483870967742e-05,
+      "loss": 0.1749,
+      "step": 6729
+    },
+    {
+      "epoch": 0.10768,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.995322580645162e-05,
+      "loss": 0.1982,
+      "step": 6730
+    },
+    {
+      "epoch": 0.107696,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.995161290322582e-05,
+      "loss": 0.1901,
+      "step": 6731
+    },
+    {
+      "epoch": 0.107712,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.995e-05,
+      "loss": 0.1754,
+      "step": 6732
+    },
+    {
+      "epoch": 0.107728,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.99483870967742e-05,
+      "loss": 0.2366,
+      "step": 6733
+    },
+    {
+      "epoch": 0.107744,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.994677419354839e-05,
+      "loss": 0.1352,
+      "step": 6734
+    },
+    {
+      "epoch": 0.10776,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.994516129032259e-05,
+      "loss": 0.2044,
+      "step": 6735
+    },
+    {
+      "epoch": 0.107776,
+      "grad_norm": 1.2734375,
+      "learning_rate": 8.994354838709678e-05,
+      "loss": 0.1554,
+      "step": 6736
+    },
+    {
+      "epoch": 0.107792,
+      "grad_norm": 1.2890625,
+      "learning_rate": 8.994193548387098e-05,
+      "loss": 0.1946,
+      "step": 6737
+    },
+    {
+      "epoch": 0.107808,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.994032258064516e-05,
+      "loss": 0.168,
+      "step": 6738
+    },
+    {
+      "epoch": 0.107824,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.993870967741936e-05,
+      "loss": 0.1986,
+      "step": 6739
+    },
+    {
+      "epoch": 0.10784,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.993709677419355e-05,
+      "loss": 0.174,
+      "step": 6740
+    },
+    {
+      "epoch": 0.107856,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.993548387096775e-05,
+      "loss": 0.1635,
+      "step": 6741
+    },
+    {
+      "epoch": 0.107872,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.993387096774195e-05,
+      "loss": 0.1826,
+      "step": 6742
+    },
+    {
+      "epoch": 0.107888,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.993225806451613e-05,
+      "loss": 0.1591,
+      "step": 6743
+    },
+    {
+      "epoch": 0.107904,
+      "grad_norm": 1.3828125,
+      "learning_rate": 8.993064516129033e-05,
+      "loss": 0.1588,
+      "step": 6744
+    },
+    {
+      "epoch": 0.10792,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.992903225806452e-05,
+      "loss": 0.1717,
+      "step": 6745
+    },
+    {
+      "epoch": 0.107936,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.992741935483872e-05,
+      "loss": 0.1867,
+      "step": 6746
+    },
+    {
+      "epoch": 0.107952,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.99258064516129e-05,
+      "loss": 0.1954,
+      "step": 6747
+    },
+    {
+      "epoch": 0.107968,
+      "grad_norm": 1.25,
+      "learning_rate": 8.99241935483871e-05,
+      "loss": 0.1852,
+      "step": 6748
+    },
+    {
+      "epoch": 0.107984,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.992258064516129e-05,
+      "loss": 0.1613,
+      "step": 6749
+    },
+    {
+      "epoch": 0.108,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.992096774193549e-05,
+      "loss": 0.2051,
+      "step": 6750
+    },
+    {
+      "epoch": 0.108016,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.991935483870968e-05,
+      "loss": 0.1816,
+      "step": 6751
+    },
+    {
+      "epoch": 0.108032,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.991774193548388e-05,
+      "loss": 0.161,
+      "step": 6752
+    },
+    {
+      "epoch": 0.108048,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.991612903225806e-05,
+      "loss": 0.1748,
+      "step": 6753
+    },
+    {
+      "epoch": 0.108064,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.991451612903226e-05,
+      "loss": 0.1588,
+      "step": 6754
+    },
+    {
+      "epoch": 0.10808,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.991290322580646e-05,
+      "loss": 0.1852,
+      "step": 6755
+    },
+    {
+      "epoch": 0.108096,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.991129032258066e-05,
+      "loss": 0.1249,
+      "step": 6756
+    },
+    {
+      "epoch": 0.108112,
+      "grad_norm": 0.875,
+      "learning_rate": 8.990967741935485e-05,
+      "loss": 0.1631,
+      "step": 6757
+    },
+    {
+      "epoch": 0.108128,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.990806451612903e-05,
+      "loss": 0.1748,
+      "step": 6758
+    },
+    {
+      "epoch": 0.108144,
+      "grad_norm": 1.5,
+      "learning_rate": 8.990645161290323e-05,
+      "loss": 0.202,
+      "step": 6759
+    },
+    {
+      "epoch": 0.10816,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.990483870967742e-05,
+      "loss": 0.2047,
+      "step": 6760
+    },
+    {
+      "epoch": 0.108176,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.990322580645162e-05,
+      "loss": 0.1561,
+      "step": 6761
+    },
+    {
+      "epoch": 0.108192,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.99016129032258e-05,
+      "loss": 0.1831,
+      "step": 6762
+    },
+    {
+      "epoch": 0.108208,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.99e-05,
+      "loss": 0.1661,
+      "step": 6763
+    },
+    {
+      "epoch": 0.108224,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.989838709677419e-05,
+      "loss": 0.1793,
+      "step": 6764
+    },
+    {
+      "epoch": 0.10824,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.989677419354839e-05,
+      "loss": 0.1825,
+      "step": 6765
+    },
+    {
+      "epoch": 0.108256,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.989516129032259e-05,
+      "loss": 0.1632,
+      "step": 6766
+    },
+    {
+      "epoch": 0.108272,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.989354838709679e-05,
+      "loss": 0.1971,
+      "step": 6767
+    },
+    {
+      "epoch": 0.108288,
+      "grad_norm": 1.0,
+      "learning_rate": 8.989193548387097e-05,
+      "loss": 0.2028,
+      "step": 6768
+    },
+    {
+      "epoch": 0.108304,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.989032258064517e-05,
+      "loss": 0.1679,
+      "step": 6769
+    },
+    {
+      "epoch": 0.10832,
+      "grad_norm": 0.875,
+      "learning_rate": 8.988870967741936e-05,
+      "loss": 0.1628,
+      "step": 6770
+    },
+    {
+      "epoch": 0.108336,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.988709677419356e-05,
+      "loss": 0.153,
+      "step": 6771
+    },
+    {
+      "epoch": 0.108352,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.988548387096775e-05,
+      "loss": 0.2071,
+      "step": 6772
+    },
+    {
+      "epoch": 0.108368,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.988387096774193e-05,
+      "loss": 0.1978,
+      "step": 6773
+    },
+    {
+      "epoch": 0.108384,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.988225806451613e-05,
+      "loss": 0.1857,
+      "step": 6774
+    },
+    {
+      "epoch": 0.1084,
+      "grad_norm": 0.875,
+      "learning_rate": 8.988064516129032e-05,
+      "loss": 0.1819,
+      "step": 6775
+    },
+    {
+      "epoch": 0.108416,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.987903225806452e-05,
+      "loss": 0.1563,
+      "step": 6776
+    },
+    {
+      "epoch": 0.108432,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.987741935483872e-05,
+      "loss": 0.1825,
+      "step": 6777
+    },
+    {
+      "epoch": 0.108448,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.98758064516129e-05,
+      "loss": 0.151,
+      "step": 6778
+    },
+    {
+      "epoch": 0.108464,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.98741935483871e-05,
+      "loss": 0.1682,
+      "step": 6779
+    },
+    {
+      "epoch": 0.10848,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.98725806451613e-05,
+      "loss": 0.1785,
+      "step": 6780
+    },
+    {
+      "epoch": 0.108496,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.987096774193549e-05,
+      "loss": 0.222,
+      "step": 6781
+    },
+    {
+      "epoch": 0.108512,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.986935483870969e-05,
+      "loss": 0.2143,
+      "step": 6782
+    },
+    {
+      "epoch": 0.108528,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.986774193548387e-05,
+      "loss": 0.1739,
+      "step": 6783
+    },
+    {
+      "epoch": 0.108544,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.986612903225807e-05,
+      "loss": 0.1682,
+      "step": 6784
+    },
+    {
+      "epoch": 0.10856,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.986451612903226e-05,
+      "loss": 0.1604,
+      "step": 6785
+    },
+    {
+      "epoch": 0.108576,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.986290322580646e-05,
+      "loss": 0.1288,
+      "step": 6786
+    },
+    {
+      "epoch": 0.108592,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.986129032258065e-05,
+      "loss": 0.1891,
+      "step": 6787
+    },
+    {
+      "epoch": 0.108608,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.985967741935484e-05,
+      "loss": 0.2031,
+      "step": 6788
+    },
+    {
+      "epoch": 0.108624,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.985806451612903e-05,
+      "loss": 0.1586,
+      "step": 6789
+    },
+    {
+      "epoch": 0.10864,
+      "grad_norm": 1.7734375,
+      "learning_rate": 8.985645161290323e-05,
+      "loss": 0.2179,
+      "step": 6790
+    },
+    {
+      "epoch": 0.108656,
+      "grad_norm": 1.78125,
+      "learning_rate": 8.985483870967743e-05,
+      "loss": 0.2045,
+      "step": 6791
+    },
+    {
+      "epoch": 0.108672,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.985322580645162e-05,
+      "loss": 0.1595,
+      "step": 6792
+    },
+    {
+      "epoch": 0.108688,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.985161290322582e-05,
+      "loss": 0.1584,
+      "step": 6793
+    },
+    {
+      "epoch": 0.108704,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.985e-05,
+      "loss": 0.1612,
+      "step": 6794
+    },
+    {
+      "epoch": 0.10872,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.98483870967742e-05,
+      "loss": 0.1968,
+      "step": 6795
+    },
+    {
+      "epoch": 0.108736,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.984677419354839e-05,
+      "loss": 0.1521,
+      "step": 6796
+    },
+    {
+      "epoch": 0.108752,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.984516129032259e-05,
+      "loss": 0.1434,
+      "step": 6797
+    },
+    {
+      "epoch": 0.108768,
+      "grad_norm": 1.390625,
+      "learning_rate": 8.984354838709677e-05,
+      "loss": 0.1823,
+      "step": 6798
+    },
+    {
+      "epoch": 0.108784,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.984193548387097e-05,
+      "loss": 0.1989,
+      "step": 6799
+    },
+    {
+      "epoch": 0.1088,
+      "grad_norm": 1.328125,
+      "learning_rate": 8.984032258064516e-05,
+      "loss": 0.2484,
+      "step": 6800
+    },
+    {
+      "epoch": 0.108816,
+      "grad_norm": 1.2890625,
+      "learning_rate": 8.983870967741936e-05,
+      "loss": 0.1844,
+      "step": 6801
+    },
+    {
+      "epoch": 0.108832,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.983709677419356e-05,
+      "loss": 0.1452,
+      "step": 6802
+    },
+    {
+      "epoch": 0.108848,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.983548387096776e-05,
+      "loss": 0.1398,
+      "step": 6803
+    },
+    {
+      "epoch": 0.108864,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.983387096774194e-05,
+      "loss": 0.1459,
+      "step": 6804
+    },
+    {
+      "epoch": 0.10888,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.983225806451613e-05,
+      "loss": 0.1949,
+      "step": 6805
+    },
+    {
+      "epoch": 0.108896,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.983064516129033e-05,
+      "loss": 0.174,
+      "step": 6806
+    },
+    {
+      "epoch": 0.108912,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.982903225806452e-05,
+      "loss": 0.1578,
+      "step": 6807
+    },
+    {
+      "epoch": 0.108928,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.982741935483872e-05,
+      "loss": 0.1716,
+      "step": 6808
+    },
+    {
+      "epoch": 0.108944,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.98258064516129e-05,
+      "loss": 0.185,
+      "step": 6809
+    },
+    {
+      "epoch": 0.10896,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.98241935483871e-05,
+      "loss": 0.1935,
+      "step": 6810
+    },
+    {
+      "epoch": 0.108976,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.982258064516129e-05,
+      "loss": 0.2096,
+      "step": 6811
+    },
+    {
+      "epoch": 0.108992,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.982096774193549e-05,
+      "loss": 0.1773,
+      "step": 6812
+    },
+    {
+      "epoch": 0.109008,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.981935483870967e-05,
+      "loss": 0.1907,
+      "step": 6813
+    },
+    {
+      "epoch": 0.109024,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.981774193548387e-05,
+      "loss": 0.1907,
+      "step": 6814
+    },
+    {
+      "epoch": 0.10904,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.981612903225807e-05,
+      "loss": 0.2201,
+      "step": 6815
+    },
+    {
+      "epoch": 0.109056,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.981451612903227e-05,
+      "loss": 0.1786,
+      "step": 6816
+    },
+    {
+      "epoch": 0.109072,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.981290322580646e-05,
+      "loss": 0.1611,
+      "step": 6817
+    },
+    {
+      "epoch": 0.109088,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.981129032258066e-05,
+      "loss": 0.1538,
+      "step": 6818
+    },
+    {
+      "epoch": 0.109104,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.980967741935484e-05,
+      "loss": 0.1463,
+      "step": 6819
+    },
+    {
+      "epoch": 0.10912,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.980806451612903e-05,
+      "loss": 0.1994,
+      "step": 6820
+    },
+    {
+      "epoch": 0.109136,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.980645161290323e-05,
+      "loss": 0.1662,
+      "step": 6821
+    },
+    {
+      "epoch": 0.109152,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.980483870967742e-05,
+      "loss": 0.2195,
+      "step": 6822
+    },
+    {
+      "epoch": 0.109168,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.980322580645162e-05,
+      "loss": 0.1473,
+      "step": 6823
+    },
+    {
+      "epoch": 0.109184,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.98016129032258e-05,
+      "loss": 0.1741,
+      "step": 6824
+    },
+    {
+      "epoch": 0.1092,
+      "grad_norm": 1.3984375,
+      "learning_rate": 8.98e-05,
+      "loss": 0.1692,
+      "step": 6825
+    },
+    {
+      "epoch": 0.109216,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.97983870967742e-05,
+      "loss": 0.1871,
+      "step": 6826
+    },
+    {
+      "epoch": 0.109232,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.97967741935484e-05,
+      "loss": 0.2088,
+      "step": 6827
+    },
+    {
+      "epoch": 0.109248,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.979516129032259e-05,
+      "loss": 0.1753,
+      "step": 6828
+    },
+    {
+      "epoch": 0.109264,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.979354838709679e-05,
+      "loss": 0.2345,
+      "step": 6829
+    },
+    {
+      "epoch": 0.10928,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.979193548387097e-05,
+      "loss": 0.181,
+      "step": 6830
+    },
+    {
+      "epoch": 0.109296,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.979032258064517e-05,
+      "loss": 0.1892,
+      "step": 6831
+    },
+    {
+      "epoch": 0.109312,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.978870967741936e-05,
+      "loss": 0.1748,
+      "step": 6832
+    },
+    {
+      "epoch": 0.109328,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.978709677419356e-05,
+      "loss": 0.1691,
+      "step": 6833
+    },
+    {
+      "epoch": 0.109344,
+      "grad_norm": 1.453125,
+      "learning_rate": 8.978548387096774e-05,
+      "loss": 0.2418,
+      "step": 6834
+    },
+    {
+      "epoch": 0.10936,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.978387096774193e-05,
+      "loss": 0.204,
+      "step": 6835
+    },
+    {
+      "epoch": 0.109376,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.978225806451613e-05,
+      "loss": 0.1721,
+      "step": 6836
+    },
+    {
+      "epoch": 0.109392,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.978064516129033e-05,
+      "loss": 0.1778,
+      "step": 6837
+    },
+    {
+      "epoch": 0.109408,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.977903225806453e-05,
+      "loss": 0.199,
+      "step": 6838
+    },
+    {
+      "epoch": 0.109424,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.977741935483871e-05,
+      "loss": 0.1766,
+      "step": 6839
+    },
+    {
+      "epoch": 0.10944,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.977580645161291e-05,
+      "loss": 0.2033,
+      "step": 6840
+    },
+    {
+      "epoch": 0.109456,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.97741935483871e-05,
+      "loss": 0.1751,
+      "step": 6841
+    },
+    {
+      "epoch": 0.109472,
+      "grad_norm": 1.265625,
+      "learning_rate": 8.97725806451613e-05,
+      "loss": 0.1663,
+      "step": 6842
+    },
+    {
+      "epoch": 0.109488,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.977096774193549e-05,
+      "loss": 0.1936,
+      "step": 6843
+    },
+    {
+      "epoch": 0.109504,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.976935483870969e-05,
+      "loss": 0.1812,
+      "step": 6844
+    },
+    {
+      "epoch": 0.10952,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.976774193548387e-05,
+      "loss": 0.2296,
+      "step": 6845
+    },
+    {
+      "epoch": 0.109536,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.976612903225807e-05,
+      "loss": 0.1786,
+      "step": 6846
+    },
+    {
+      "epoch": 0.109552,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.976451612903226e-05,
+      "loss": 0.1644,
+      "step": 6847
+    },
+    {
+      "epoch": 0.109568,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.976290322580646e-05,
+      "loss": 0.1904,
+      "step": 6848
+    },
+    {
+      "epoch": 0.109584,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.976129032258064e-05,
+      "loss": 0.171,
+      "step": 6849
+    },
+    {
+      "epoch": 0.1096,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.975967741935484e-05,
+      "loss": 0.1897,
+      "step": 6850
+    },
+    {
+      "epoch": 0.109616,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.975806451612904e-05,
+      "loss": 0.1563,
+      "step": 6851
+    },
+    {
+      "epoch": 0.109632,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.975645161290323e-05,
+      "loss": 0.1713,
+      "step": 6852
+    },
+    {
+      "epoch": 0.109648,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.975483870967743e-05,
+      "loss": 0.176,
+      "step": 6853
+    },
+    {
+      "epoch": 0.109664,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.975322580645161e-05,
+      "loss": 0.225,
+      "step": 6854
+    },
+    {
+      "epoch": 0.10968,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.975161290322581e-05,
+      "loss": 0.1397,
+      "step": 6855
+    },
+    {
+      "epoch": 0.109696,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.975e-05,
+      "loss": 0.1628,
+      "step": 6856
+    },
+    {
+      "epoch": 0.109712,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.97483870967742e-05,
+      "loss": 0.1909,
+      "step": 6857
+    },
+    {
+      "epoch": 0.109728,
+      "grad_norm": 1.4296875,
+      "learning_rate": 8.974677419354839e-05,
+      "loss": 0.1845,
+      "step": 6858
+    },
+    {
+      "epoch": 0.109744,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.974516129032258e-05,
+      "loss": 0.165,
+      "step": 6859
+    },
+    {
+      "epoch": 0.10976,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.974354838709677e-05,
+      "loss": 0.1975,
+      "step": 6860
+    },
+    {
+      "epoch": 0.109776,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.974193548387097e-05,
+      "loss": 0.1633,
+      "step": 6861
+    },
+    {
+      "epoch": 0.109792,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.974032258064517e-05,
+      "loss": 0.1473,
+      "step": 6862
+    },
+    {
+      "epoch": 0.109808,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.973870967741937e-05,
+      "loss": 0.2226,
+      "step": 6863
+    },
+    {
+      "epoch": 0.109824,
+      "grad_norm": 0.75,
+      "learning_rate": 8.973709677419356e-05,
+      "loss": 0.1853,
+      "step": 6864
+    },
+    {
+      "epoch": 0.10984,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.973548387096776e-05,
+      "loss": 0.1655,
+      "step": 6865
+    },
+    {
+      "epoch": 0.109856,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.973387096774194e-05,
+      "loss": 0.2008,
+      "step": 6866
+    },
+    {
+      "epoch": 0.109872,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.973225806451613e-05,
+      "loss": 0.1339,
+      "step": 6867
+    },
+    {
+      "epoch": 0.109888,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.973064516129033e-05,
+      "loss": 0.1922,
+      "step": 6868
+    },
+    {
+      "epoch": 0.109904,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.972903225806451e-05,
+      "loss": 0.1933,
+      "step": 6869
+    },
+    {
+      "epoch": 0.10992,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.972741935483871e-05,
+      "loss": 0.1709,
+      "step": 6870
+    },
+    {
+      "epoch": 0.109936,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.97258064516129e-05,
+      "loss": 0.1955,
+      "step": 6871
+    },
+    {
+      "epoch": 0.109952,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.97241935483871e-05,
+      "loss": 0.1647,
+      "step": 6872
+    },
+    {
+      "epoch": 0.109968,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.972258064516128e-05,
+      "loss": 0.1613,
+      "step": 6873
+    },
+    {
+      "epoch": 0.109984,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.972096774193548e-05,
+      "loss": 0.1897,
+      "step": 6874
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.971935483870968e-05,
+      "loss": 0.2076,
+      "step": 6875
+    },
+    {
+      "epoch": 0.110016,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.971774193548388e-05,
+      "loss": 0.1945,
+      "step": 6876
+    },
+    {
+      "epoch": 0.110032,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.971612903225807e-05,
+      "loss": 0.193,
+      "step": 6877
+    },
+    {
+      "epoch": 0.110048,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.971451612903227e-05,
+      "loss": 0.1806,
+      "step": 6878
+    },
+    {
+      "epoch": 0.110064,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.971290322580646e-05,
+      "loss": 0.1135,
+      "step": 6879
+    },
+    {
+      "epoch": 0.11008,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.971129032258066e-05,
+      "loss": 0.2066,
+      "step": 6880
+    },
+    {
+      "epoch": 0.110096,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.970967741935484e-05,
+      "loss": 0.1355,
+      "step": 6881
+    },
+    {
+      "epoch": 0.110112,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.970806451612903e-05,
+      "loss": 0.203,
+      "step": 6882
+    },
+    {
+      "epoch": 0.110128,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.970645161290323e-05,
+      "loss": 0.2134,
+      "step": 6883
+    },
+    {
+      "epoch": 0.110144,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.970483870967741e-05,
+      "loss": 0.2113,
+      "step": 6884
+    },
+    {
+      "epoch": 0.11016,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.970322580645161e-05,
+      "loss": 0.1654,
+      "step": 6885
+    },
+    {
+      "epoch": 0.110176,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.970161290322581e-05,
+      "loss": 0.1708,
+      "step": 6886
+    },
+    {
+      "epoch": 0.110192,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.970000000000001e-05,
+      "loss": 0.1701,
+      "step": 6887
+    },
+    {
+      "epoch": 0.110208,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.96983870967742e-05,
+      "loss": 0.1975,
+      "step": 6888
+    },
+    {
+      "epoch": 0.110224,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.96967741935484e-05,
+      "loss": 0.2004,
+      "step": 6889
+    },
+    {
+      "epoch": 0.11024,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.969516129032258e-05,
+      "loss": 0.1573,
+      "step": 6890
+    },
+    {
+      "epoch": 0.110256,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.969354838709678e-05,
+      "loss": 0.1914,
+      "step": 6891
+    },
+    {
+      "epoch": 0.110272,
+      "grad_norm": 1.0,
+      "learning_rate": 8.969193548387097e-05,
+      "loss": 0.1934,
+      "step": 6892
+    },
+    {
+      "epoch": 0.110288,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.969032258064517e-05,
+      "loss": 0.1497,
+      "step": 6893
+    },
+    {
+      "epoch": 0.110304,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.968870967741936e-05,
+      "loss": 0.1849,
+      "step": 6894
+    },
+    {
+      "epoch": 0.11032,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.968709677419355e-05,
+      "loss": 0.1432,
+      "step": 6895
+    },
+    {
+      "epoch": 0.110336,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.968548387096774e-05,
+      "loss": 0.1561,
+      "step": 6896
+    },
+    {
+      "epoch": 0.110352,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.968387096774194e-05,
+      "loss": 0.197,
+      "step": 6897
+    },
+    {
+      "epoch": 0.110368,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.968225806451614e-05,
+      "loss": 0.1485,
+      "step": 6898
+    },
+    {
+      "epoch": 0.110384,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.968064516129033e-05,
+      "loss": 0.1656,
+      "step": 6899
+    },
+    {
+      "epoch": 0.1104,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.967903225806453e-05,
+      "loss": 0.2083,
+      "step": 6900
+    },
+    {
+      "epoch": 0.110416,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.967741935483871e-05,
+      "loss": 0.1289,
+      "step": 6901
+    },
+    {
+      "epoch": 0.110432,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.967580645161291e-05,
+      "loss": 0.1888,
+      "step": 6902
+    },
+    {
+      "epoch": 0.110448,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.96741935483871e-05,
+      "loss": 0.2085,
+      "step": 6903
+    },
+    {
+      "epoch": 0.110464,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.96725806451613e-05,
+      "loss": 0.1459,
+      "step": 6904
+    },
+    {
+      "epoch": 0.11048,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.967096774193548e-05,
+      "loss": 0.154,
+      "step": 6905
+    },
+    {
+      "epoch": 0.110496,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.966935483870968e-05,
+      "loss": 0.1363,
+      "step": 6906
+    },
+    {
+      "epoch": 0.110512,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.966774193548387e-05,
+      "loss": 0.2047,
+      "step": 6907
+    },
+    {
+      "epoch": 0.110528,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.966612903225807e-05,
+      "loss": 0.1941,
+      "step": 6908
+    },
+    {
+      "epoch": 0.110544,
+      "grad_norm": 1.296875,
+      "learning_rate": 8.966451612903225e-05,
+      "loss": 0.1766,
+      "step": 6909
+    },
+    {
+      "epoch": 0.11056,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.966290322580645e-05,
+      "loss": 0.1837,
+      "step": 6910
+    },
+    {
+      "epoch": 0.110576,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.966129032258065e-05,
+      "loss": 0.1802,
+      "step": 6911
+    },
+    {
+      "epoch": 0.110592,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.965967741935485e-05,
+      "loss": 0.1663,
+      "step": 6912
+    },
+    {
+      "epoch": 0.110608,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.965806451612904e-05,
+      "loss": 0.1717,
+      "step": 6913
+    },
+    {
+      "epoch": 0.110624,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.965645161290323e-05,
+      "loss": 0.1795,
+      "step": 6914
+    },
+    {
+      "epoch": 0.11064,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.965483870967743e-05,
+      "loss": 0.192,
+      "step": 6915
+    },
+    {
+      "epoch": 0.110656,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.965322580645161e-05,
+      "loss": 0.2337,
+      "step": 6916
+    },
+    {
+      "epoch": 0.110672,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.965161290322581e-05,
+      "loss": 0.1748,
+      "step": 6917
+    },
+    {
+      "epoch": 0.110688,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.965e-05,
+      "loss": 0.1445,
+      "step": 6918
+    },
+    {
+      "epoch": 0.110704,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.96483870967742e-05,
+      "loss": 0.161,
+      "step": 6919
+    },
+    {
+      "epoch": 0.11072,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.964677419354838e-05,
+      "loss": 0.2044,
+      "step": 6920
+    },
+    {
+      "epoch": 0.110736,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.964516129032258e-05,
+      "loss": 0.1729,
+      "step": 6921
+    },
+    {
+      "epoch": 0.110752,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.964354838709678e-05,
+      "loss": 0.1649,
+      "step": 6922
+    },
+    {
+      "epoch": 0.110768,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.964193548387098e-05,
+      "loss": 0.1504,
+      "step": 6923
+    },
+    {
+      "epoch": 0.110784,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.964032258064517e-05,
+      "loss": 0.1713,
+      "step": 6924
+    },
+    {
+      "epoch": 0.1108,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.963870967741937e-05,
+      "loss": 0.2185,
+      "step": 6925
+    },
+    {
+      "epoch": 0.110816,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.963709677419355e-05,
+      "loss": 0.1975,
+      "step": 6926
+    },
+    {
+      "epoch": 0.110832,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.963548387096775e-05,
+      "loss": 0.1636,
+      "step": 6927
+    },
+    {
+      "epoch": 0.110848,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.963387096774194e-05,
+      "loss": 0.1657,
+      "step": 6928
+    },
+    {
+      "epoch": 0.110864,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.963225806451613e-05,
+      "loss": 0.1666,
+      "step": 6929
+    },
+    {
+      "epoch": 0.11088,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.963064516129032e-05,
+      "loss": 0.2165,
+      "step": 6930
+    },
+    {
+      "epoch": 0.110896,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.962903225806451e-05,
+      "loss": 0.1963,
+      "step": 6931
+    },
+    {
+      "epoch": 0.110912,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.962741935483871e-05,
+      "loss": 0.2068,
+      "step": 6932
+    },
+    {
+      "epoch": 0.110928,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.962580645161291e-05,
+      "loss": 0.193,
+      "step": 6933
+    },
+    {
+      "epoch": 0.110944,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.96241935483871e-05,
+      "loss": 0.193,
+      "step": 6934
+    },
+    {
+      "epoch": 0.11096,
+      "grad_norm": 0.4609375,
+      "learning_rate": 8.96225806451613e-05,
+      "loss": 0.1399,
+      "step": 6935
+    },
+    {
+      "epoch": 0.110976,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.96209677419355e-05,
+      "loss": 0.1644,
+      "step": 6936
+    },
+    {
+      "epoch": 0.110992,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.961935483870968e-05,
+      "loss": 0.1749,
+      "step": 6937
+    },
+    {
+      "epoch": 0.111008,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.961774193548388e-05,
+      "loss": 0.2324,
+      "step": 6938
+    },
+    {
+      "epoch": 0.111024,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.961612903225807e-05,
+      "loss": 0.1947,
+      "step": 6939
+    },
+    {
+      "epoch": 0.11104,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.961451612903227e-05,
+      "loss": 0.2396,
+      "step": 6940
+    },
+    {
+      "epoch": 0.111056,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.961290322580645e-05,
+      "loss": 0.1743,
+      "step": 6941
+    },
+    {
+      "epoch": 0.111072,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.961129032258065e-05,
+      "loss": 0.1726,
+      "step": 6942
+    },
+    {
+      "epoch": 0.111088,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.960967741935484e-05,
+      "loss": 0.1661,
+      "step": 6943
+    },
+    {
+      "epoch": 0.111104,
+      "grad_norm": 1.328125,
+      "learning_rate": 8.960806451612902e-05,
+      "loss": 0.1948,
+      "step": 6944
+    },
+    {
+      "epoch": 0.11112,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.960645161290322e-05,
+      "loss": 0.174,
+      "step": 6945
+    },
+    {
+      "epoch": 0.111136,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.960483870967742e-05,
+      "loss": 0.1544,
+      "step": 6946
+    },
+    {
+      "epoch": 0.111152,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.960322580645162e-05,
+      "loss": 0.1805,
+      "step": 6947
+    },
+    {
+      "epoch": 0.111168,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.960161290322581e-05,
+      "loss": 0.177,
+      "step": 6948
+    },
+    {
+      "epoch": 0.111184,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.960000000000001e-05,
+      "loss": 0.1812,
+      "step": 6949
+    },
+    {
+      "epoch": 0.1112,
+      "grad_norm": 0.875,
+      "learning_rate": 8.95983870967742e-05,
+      "loss": 0.1697,
+      "step": 6950
+    },
+    {
+      "epoch": 0.111216,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.95967741935484e-05,
+      "loss": 0.1829,
+      "step": 6951
+    },
+    {
+      "epoch": 0.111232,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.959516129032258e-05,
+      "loss": 0.1583,
+      "step": 6952
+    },
+    {
+      "epoch": 0.111248,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.959354838709678e-05,
+      "loss": 0.1822,
+      "step": 6953
+    },
+    {
+      "epoch": 0.111264,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.959193548387097e-05,
+      "loss": 0.1541,
+      "step": 6954
+    },
+    {
+      "epoch": 0.11128,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.959032258064517e-05,
+      "loss": 0.1684,
+      "step": 6955
+    },
+    {
+      "epoch": 0.111296,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.958870967741935e-05,
+      "loss": 0.1738,
+      "step": 6956
+    },
+    {
+      "epoch": 0.111312,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.958709677419355e-05,
+      "loss": 0.1802,
+      "step": 6957
+    },
+    {
+      "epoch": 0.111328,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.958548387096775e-05,
+      "loss": 0.1908,
+      "step": 6958
+    },
+    {
+      "epoch": 0.111344,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.958387096774195e-05,
+      "loss": 0.1879,
+      "step": 6959
+    },
+    {
+      "epoch": 0.11136,
+      "grad_norm": 1.0,
+      "learning_rate": 8.958225806451614e-05,
+      "loss": 0.2099,
+      "step": 6960
+    },
+    {
+      "epoch": 0.111376,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.958064516129032e-05,
+      "loss": 0.1907,
+      "step": 6961
+    },
+    {
+      "epoch": 0.111392,
+      "grad_norm": 0.625,
+      "learning_rate": 8.957903225806452e-05,
+      "loss": 0.16,
+      "step": 6962
+    },
+    {
+      "epoch": 0.111408,
+      "grad_norm": 1.390625,
+      "learning_rate": 8.957741935483871e-05,
+      "loss": 0.179,
+      "step": 6963
+    },
+    {
+      "epoch": 0.111424,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.957580645161291e-05,
+      "loss": 0.1417,
+      "step": 6964
+    },
+    {
+      "epoch": 0.11144,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.95741935483871e-05,
+      "loss": 0.1694,
+      "step": 6965
+    },
+    {
+      "epoch": 0.111456,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.95725806451613e-05,
+      "loss": 0.154,
+      "step": 6966
+    },
+    {
+      "epoch": 0.111472,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.957096774193548e-05,
+      "loss": 0.1711,
+      "step": 6967
+    },
+    {
+      "epoch": 0.111488,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.956935483870968e-05,
+      "loss": 0.189,
+      "step": 6968
+    },
+    {
+      "epoch": 0.111504,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.956774193548387e-05,
+      "loss": 0.1905,
+      "step": 6969
+    },
+    {
+      "epoch": 0.11152,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.956612903225807e-05,
+      "loss": 0.1875,
+      "step": 6970
+    },
+    {
+      "epoch": 0.111536,
+      "grad_norm": 1.390625,
+      "learning_rate": 8.956451612903227e-05,
+      "loss": 0.1971,
+      "step": 6971
+    },
+    {
+      "epoch": 0.111552,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.956290322580647e-05,
+      "loss": 0.1659,
+      "step": 6972
+    },
+    {
+      "epoch": 0.111568,
+      "grad_norm": 0.462890625,
+      "learning_rate": 8.956129032258065e-05,
+      "loss": 0.148,
+      "step": 6973
+    },
+    {
+      "epoch": 0.111584,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.955967741935485e-05,
+      "loss": 0.2014,
+      "step": 6974
+    },
+    {
+      "epoch": 0.1116,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.955806451612904e-05,
+      "loss": 0.163,
+      "step": 6975
+    },
+    {
+      "epoch": 0.111616,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.955645161290322e-05,
+      "loss": 0.1963,
+      "step": 6976
+    },
+    {
+      "epoch": 0.111632,
+      "grad_norm": 1.5390625,
+      "learning_rate": 8.955483870967742e-05,
+      "loss": 0.2143,
+      "step": 6977
+    },
+    {
+      "epoch": 0.111648,
+      "grad_norm": 0.75,
+      "learning_rate": 8.955322580645161e-05,
+      "loss": 0.1882,
+      "step": 6978
+    },
+    {
+      "epoch": 0.111664,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.955161290322581e-05,
+      "loss": 0.1521,
+      "step": 6979
+    },
+    {
+      "epoch": 0.11168,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.955e-05,
+      "loss": 0.1923,
+      "step": 6980
+    },
+    {
+      "epoch": 0.111696,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.95483870967742e-05,
+      "loss": 0.189,
+      "step": 6981
+    },
+    {
+      "epoch": 0.111712,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.95467741935484e-05,
+      "loss": 0.195,
+      "step": 6982
+    },
+    {
+      "epoch": 0.111728,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.95451612903226e-05,
+      "loss": 0.1794,
+      "step": 6983
+    },
+    {
+      "epoch": 0.111744,
+      "grad_norm": 1.5390625,
+      "learning_rate": 8.954354838709678e-05,
+      "loss": 0.1687,
+      "step": 6984
+    },
+    {
+      "epoch": 0.11176,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.954193548387098e-05,
+      "loss": 0.1884,
+      "step": 6985
+    },
+    {
+      "epoch": 0.111776,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.954032258064517e-05,
+      "loss": 0.1997,
+      "step": 6986
+    },
+    {
+      "epoch": 0.111792,
+      "grad_norm": 1.4765625,
+      "learning_rate": 8.953870967741937e-05,
+      "loss": 0.2151,
+      "step": 6987
+    },
+    {
+      "epoch": 0.111808,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.953709677419355e-05,
+      "loss": 0.174,
+      "step": 6988
+    },
+    {
+      "epoch": 0.111824,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.953548387096775e-05,
+      "loss": 0.1678,
+      "step": 6989
+    },
+    {
+      "epoch": 0.11184,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.953387096774194e-05,
+      "loss": 0.1472,
+      "step": 6990
+    },
+    {
+      "epoch": 0.111856,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.953225806451612e-05,
+      "loss": 0.142,
+      "step": 6991
+    },
+    {
+      "epoch": 0.111872,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.953064516129032e-05,
+      "loss": 0.1743,
+      "step": 6992
+    },
+    {
+      "epoch": 0.111888,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.952903225806452e-05,
+      "loss": 0.1898,
+      "step": 6993
+    },
+    {
+      "epoch": 0.111904,
+      "grad_norm": 1.2109375,
+      "learning_rate": 8.952741935483872e-05,
+      "loss": 0.2283,
+      "step": 6994
+    },
+    {
+      "epoch": 0.11192,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.952580645161291e-05,
+      "loss": 0.1857,
+      "step": 6995
+    },
+    {
+      "epoch": 0.111936,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.952419354838711e-05,
+      "loss": 0.1524,
+      "step": 6996
+    },
+    {
+      "epoch": 0.111952,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.95225806451613e-05,
+      "loss": 0.2097,
+      "step": 6997
+    },
+    {
+      "epoch": 0.111968,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.952096774193549e-05,
+      "loss": 0.1644,
+      "step": 6998
+    },
+    {
+      "epoch": 0.111984,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.951935483870968e-05,
+      "loss": 0.145,
+      "step": 6999
+    },
+    {
+      "epoch": 0.112,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.951774193548388e-05,
+      "loss": 0.1918,
+      "step": 7000
+    },
+    {
+      "epoch": 0.112016,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.951612903225806e-05,
+      "loss": 0.1793,
+      "step": 7001
+    },
+    {
+      "epoch": 0.112032,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.951451612903226e-05,
+      "loss": 0.1403,
+      "step": 7002
+    },
+    {
+      "epoch": 0.112048,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.951290322580645e-05,
+      "loss": 0.1959,
+      "step": 7003
+    },
+    {
+      "epoch": 0.112064,
+      "grad_norm": 1.5546875,
+      "learning_rate": 8.951129032258065e-05,
+      "loss": 0.1748,
+      "step": 7004
+    },
+    {
+      "epoch": 0.11208,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.950967741935484e-05,
+      "loss": 0.1739,
+      "step": 7005
+    },
+    {
+      "epoch": 0.112096,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.950806451612904e-05,
+      "loss": 0.2037,
+      "step": 7006
+    },
+    {
+      "epoch": 0.112112,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.950645161290324e-05,
+      "loss": 0.2094,
+      "step": 7007
+    },
+    {
+      "epoch": 0.112128,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.950483870967742e-05,
+      "loss": 0.1729,
+      "step": 7008
+    },
+    {
+      "epoch": 0.112144,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.950322580645162e-05,
+      "loss": 0.184,
+      "step": 7009
+    },
+    {
+      "epoch": 0.11216,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.950161290322581e-05,
+      "loss": 0.1578,
+      "step": 7010
+    },
+    {
+      "epoch": 0.112176,
+      "grad_norm": 1.875,
+      "learning_rate": 8.950000000000001e-05,
+      "loss": 0.1822,
+      "step": 7011
+    },
+    {
+      "epoch": 0.112192,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.949838709677419e-05,
+      "loss": 0.1683,
+      "step": 7012
+    },
+    {
+      "epoch": 0.112208,
+      "grad_norm": 1.4765625,
+      "learning_rate": 8.949677419354839e-05,
+      "loss": 0.151,
+      "step": 7013
+    },
+    {
+      "epoch": 0.112224,
+      "grad_norm": 1.515625,
+      "learning_rate": 8.949516129032258e-05,
+      "loss": 0.1786,
+      "step": 7014
+    },
+    {
+      "epoch": 0.11224,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.949354838709678e-05,
+      "loss": 0.1681,
+      "step": 7015
+    },
+    {
+      "epoch": 0.112256,
+      "grad_norm": 1.2109375,
+      "learning_rate": 8.949193548387096e-05,
+      "loss": 0.1534,
+      "step": 7016
+    },
+    {
+      "epoch": 0.112272,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.949032258064516e-05,
+      "loss": 0.1657,
+      "step": 7017
+    },
+    {
+      "epoch": 0.112288,
+      "grad_norm": 1.3671875,
+      "learning_rate": 8.948870967741936e-05,
+      "loss": 0.1999,
+      "step": 7018
+    },
+    {
+      "epoch": 0.112304,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.948709677419356e-05,
+      "loss": 0.1818,
+      "step": 7019
+    },
+    {
+      "epoch": 0.11232,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.948548387096775e-05,
+      "loss": 0.1789,
+      "step": 7020
+    },
+    {
+      "epoch": 0.112336,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.948387096774195e-05,
+      "loss": 0.1851,
+      "step": 7021
+    },
+    {
+      "epoch": 0.112352,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.948225806451614e-05,
+      "loss": 0.1552,
+      "step": 7022
+    },
+    {
+      "epoch": 0.112368,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.948064516129032e-05,
+      "loss": 0.1554,
+      "step": 7023
+    },
+    {
+      "epoch": 0.112384,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.947903225806452e-05,
+      "loss": 0.1357,
+      "step": 7024
+    },
+    {
+      "epoch": 0.1124,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.947741935483871e-05,
+      "loss": 0.1943,
+      "step": 7025
+    },
+    {
+      "epoch": 0.112416,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.94758064516129e-05,
+      "loss": 0.2256,
+      "step": 7026
+    },
+    {
+      "epoch": 0.112432,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.947419354838709e-05,
+      "loss": 0.1526,
+      "step": 7027
+    },
+    {
+      "epoch": 0.112448,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.947258064516129e-05,
+      "loss": 0.1914,
+      "step": 7028
+    },
+    {
+      "epoch": 0.112464,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.947096774193549e-05,
+      "loss": 0.1552,
+      "step": 7029
+    },
+    {
+      "epoch": 0.11248,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.946935483870968e-05,
+      "loss": 0.139,
+      "step": 7030
+    },
+    {
+      "epoch": 0.112496,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.946774193548388e-05,
+      "loss": 0.2127,
+      "step": 7031
+    },
+    {
+      "epoch": 0.112512,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.946612903225808e-05,
+      "loss": 0.1888,
+      "step": 7032
+    },
+    {
+      "epoch": 0.112528,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.946451612903226e-05,
+      "loss": 0.1719,
+      "step": 7033
+    },
+    {
+      "epoch": 0.112544,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.946290322580646e-05,
+      "loss": 0.1843,
+      "step": 7034
+    },
+    {
+      "epoch": 0.11256,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.946129032258065e-05,
+      "loss": 0.1795,
+      "step": 7035
+    },
+    {
+      "epoch": 0.112576,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.945967741935485e-05,
+      "loss": 0.2081,
+      "step": 7036
+    },
+    {
+      "epoch": 0.112592,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.945806451612903e-05,
+      "loss": 0.1913,
+      "step": 7037
+    },
+    {
+      "epoch": 0.112608,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.945645161290322e-05,
+      "loss": 0.1947,
+      "step": 7038
+    },
+    {
+      "epoch": 0.112624,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.945483870967742e-05,
+      "loss": 0.1582,
+      "step": 7039
+    },
+    {
+      "epoch": 0.11264,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.94532258064516e-05,
+      "loss": 0.1512,
+      "step": 7040
+    },
+    {
+      "epoch": 0.112656,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.94516129032258e-05,
+      "loss": 0.1768,
+      "step": 7041
+    },
+    {
+      "epoch": 0.112672,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.945e-05,
+      "loss": 0.2232,
+      "step": 7042
+    },
+    {
+      "epoch": 0.112688,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.94483870967742e-05,
+      "loss": 0.2216,
+      "step": 7043
+    },
+    {
+      "epoch": 0.112704,
+      "grad_norm": 1.0,
+      "learning_rate": 8.944677419354839e-05,
+      "loss": 0.2209,
+      "step": 7044
+    },
+    {
+      "epoch": 0.11272,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.944516129032259e-05,
+      "loss": 0.2062,
+      "step": 7045
+    },
+    {
+      "epoch": 0.112736,
+      "grad_norm": 0.95703125,
+      "learning_rate": 8.944354838709678e-05,
+      "loss": 0.1858,
+      "step": 7046
+    },
+    {
+      "epoch": 0.112752,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.944193548387098e-05,
+      "loss": 0.1502,
+      "step": 7047
+    },
+    {
+      "epoch": 0.112768,
+      "grad_norm": 1.3828125,
+      "learning_rate": 8.944032258064516e-05,
+      "loss": 0.1648,
+      "step": 7048
+    },
+    {
+      "epoch": 0.112784,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.943870967741936e-05,
+      "loss": 0.2302,
+      "step": 7049
+    },
+    {
+      "epoch": 0.1128,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.943709677419355e-05,
+      "loss": 0.1947,
+      "step": 7050
+    },
+    {
+      "epoch": 0.112816,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.943548387096775e-05,
+      "loss": 0.1701,
+      "step": 7051
+    },
+    {
+      "epoch": 0.112832,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.943387096774193e-05,
+      "loss": 0.2338,
+      "step": 7052
+    },
+    {
+      "epoch": 0.112848,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.943225806451613e-05,
+      "loss": 0.2021,
+      "step": 7053
+    },
+    {
+      "epoch": 0.112864,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.943064516129033e-05,
+      "loss": 0.158,
+      "step": 7054
+    },
+    {
+      "epoch": 0.11288,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.942903225806452e-05,
+      "loss": 0.1996,
+      "step": 7055
+    },
+    {
+      "epoch": 0.112896,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.942741935483872e-05,
+      "loss": 0.182,
+      "step": 7056
+    },
+    {
+      "epoch": 0.112912,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.94258064516129e-05,
+      "loss": 0.1832,
+      "step": 7057
+    },
+    {
+      "epoch": 0.112928,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.94241935483871e-05,
+      "loss": 0.1998,
+      "step": 7058
+    },
+    {
+      "epoch": 0.112944,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.942258064516129e-05,
+      "loss": 0.1691,
+      "step": 7059
+    },
+    {
+      "epoch": 0.11296,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.942096774193549e-05,
+      "loss": 0.1765,
+      "step": 7060
+    },
+    {
+      "epoch": 0.112976,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.941935483870968e-05,
+      "loss": 0.1824,
+      "step": 7061
+    },
+    {
+      "epoch": 0.112992,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.941774193548388e-05,
+      "loss": 0.151,
+      "step": 7062
+    },
+    {
+      "epoch": 0.113008,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.941612903225806e-05,
+      "loss": 0.2011,
+      "step": 7063
+    },
+    {
+      "epoch": 0.113024,
+      "grad_norm": 1.4453125,
+      "learning_rate": 8.941451612903226e-05,
+      "loss": 0.2159,
+      "step": 7064
+    },
+    {
+      "epoch": 0.11304,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.941290322580645e-05,
+      "loss": 0.1896,
+      "step": 7065
+    },
+    {
+      "epoch": 0.113056,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.941129032258065e-05,
+      "loss": 0.1614,
+      "step": 7066
+    },
+    {
+      "epoch": 0.113072,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.940967741935485e-05,
+      "loss": 0.1795,
+      "step": 7067
+    },
+    {
+      "epoch": 0.113088,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.940806451612905e-05,
+      "loss": 0.183,
+      "step": 7068
+    },
+    {
+      "epoch": 0.113104,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.940645161290323e-05,
+      "loss": 0.1821,
+      "step": 7069
+    },
+    {
+      "epoch": 0.11312,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.940483870967742e-05,
+      "loss": 0.1953,
+      "step": 7070
+    },
+    {
+      "epoch": 0.113136,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.940322580645162e-05,
+      "loss": 0.1572,
+      "step": 7071
+    },
+    {
+      "epoch": 0.113152,
+      "grad_norm": 0.46875,
+      "learning_rate": 8.94016129032258e-05,
+      "loss": 0.1821,
+      "step": 7072
+    },
+    {
+      "epoch": 0.113168,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.94e-05,
+      "loss": 0.2009,
+      "step": 7073
+    },
+    {
+      "epoch": 0.113184,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.939838709677419e-05,
+      "loss": 0.1869,
+      "step": 7074
+    },
+    {
+      "epoch": 0.1132,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.939677419354839e-05,
+      "loss": 0.1773,
+      "step": 7075
+    },
+    {
+      "epoch": 0.113216,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.939516129032258e-05,
+      "loss": 0.2172,
+      "step": 7076
+    },
+    {
+      "epoch": 0.113232,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.939354838709678e-05,
+      "loss": 0.178,
+      "step": 7077
+    },
+    {
+      "epoch": 0.113248,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.939193548387098e-05,
+      "loss": 0.1915,
+      "step": 7078
+    },
+    {
+      "epoch": 0.113264,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.939032258064518e-05,
+      "loss": 0.1643,
+      "step": 7079
+    },
+    {
+      "epoch": 0.11328,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.938870967741936e-05,
+      "loss": 0.1519,
+      "step": 7080
+    },
+    {
+      "epoch": 0.113296,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.938709677419356e-05,
+      "loss": 0.1908,
+      "step": 7081
+    },
+    {
+      "epoch": 0.113312,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.938548387096775e-05,
+      "loss": 0.1764,
+      "step": 7082
+    },
+    {
+      "epoch": 0.113328,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.938387096774195e-05,
+      "loss": 0.167,
+      "step": 7083
+    },
+    {
+      "epoch": 0.113344,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.938225806451613e-05,
+      "loss": 0.1415,
+      "step": 7084
+    },
+    {
+      "epoch": 0.11336,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.938064516129032e-05,
+      "loss": 0.1256,
+      "step": 7085
+    },
+    {
+      "epoch": 0.113376,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.937903225806452e-05,
+      "loss": 0.1365,
+      "step": 7086
+    },
+    {
+      "epoch": 0.113392,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.93774193548387e-05,
+      "loss": 0.1616,
+      "step": 7087
+    },
+    {
+      "epoch": 0.113408,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.93758064516129e-05,
+      "loss": 0.1969,
+      "step": 7088
+    },
+    {
+      "epoch": 0.113424,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.93741935483871e-05,
+      "loss": 0.2016,
+      "step": 7089
+    },
+    {
+      "epoch": 0.11344,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.93725806451613e-05,
+      "loss": 0.1637,
+      "step": 7090
+    },
+    {
+      "epoch": 0.113456,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.937096774193549e-05,
+      "loss": 0.2007,
+      "step": 7091
+    },
+    {
+      "epoch": 0.113472,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.936935483870969e-05,
+      "loss": 0.1536,
+      "step": 7092
+    },
+    {
+      "epoch": 0.113488,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.936774193548388e-05,
+      "loss": 0.184,
+      "step": 7093
+    },
+    {
+      "epoch": 0.113504,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.936612903225807e-05,
+      "loss": 0.1961,
+      "step": 7094
+    },
+    {
+      "epoch": 0.11352,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.936451612903226e-05,
+      "loss": 0.1889,
+      "step": 7095
+    },
+    {
+      "epoch": 0.113536,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.936290322580646e-05,
+      "loss": 0.1721,
+      "step": 7096
+    },
+    {
+      "epoch": 0.113552,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.936129032258065e-05,
+      "loss": 0.1718,
+      "step": 7097
+    },
+    {
+      "epoch": 0.113568,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.935967741935485e-05,
+      "loss": 0.2018,
+      "step": 7098
+    },
+    {
+      "epoch": 0.113584,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.935806451612903e-05,
+      "loss": 0.1553,
+      "step": 7099
+    },
+    {
+      "epoch": 0.1136,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.935645161290322e-05,
+      "loss": 0.1856,
+      "step": 7100
+    },
+    {
+      "epoch": 0.113616,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.935483870967742e-05,
+      "loss": 0.1801,
+      "step": 7101
+    },
+    {
+      "epoch": 0.113632,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.935322580645162e-05,
+      "loss": 0.212,
+      "step": 7102
+    },
+    {
+      "epoch": 0.113648,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.935161290322582e-05,
+      "loss": 0.1778,
+      "step": 7103
+    },
+    {
+      "epoch": 0.113664,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.935e-05,
+      "loss": 0.205,
+      "step": 7104
+    },
+    {
+      "epoch": 0.11368,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.93483870967742e-05,
+      "loss": 0.2084,
+      "step": 7105
+    },
+    {
+      "epoch": 0.113696,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.934677419354839e-05,
+      "loss": 0.146,
+      "step": 7106
+    },
+    {
+      "epoch": 0.113712,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.934516129032259e-05,
+      "loss": 0.1927,
+      "step": 7107
+    },
+    {
+      "epoch": 0.113728,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.934354838709677e-05,
+      "loss": 0.1724,
+      "step": 7108
+    },
+    {
+      "epoch": 0.113744,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.934193548387097e-05,
+      "loss": 0.1823,
+      "step": 7109
+    },
+    {
+      "epoch": 0.11376,
+      "grad_norm": 1.2109375,
+      "learning_rate": 8.934032258064516e-05,
+      "loss": 0.1483,
+      "step": 7110
+    },
+    {
+      "epoch": 0.113776,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.933870967741936e-05,
+      "loss": 0.1944,
+      "step": 7111
+    },
+    {
+      "epoch": 0.113792,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.933709677419355e-05,
+      "loss": 0.1591,
+      "step": 7112
+    },
+    {
+      "epoch": 0.113808,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.933548387096775e-05,
+      "loss": 0.168,
+      "step": 7113
+    },
+    {
+      "epoch": 0.113824,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.933387096774195e-05,
+      "loss": 0.1478,
+      "step": 7114
+    },
+    {
+      "epoch": 0.11384,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.933225806451613e-05,
+      "loss": 0.1876,
+      "step": 7115
+    },
+    {
+      "epoch": 0.113856,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.933064516129033e-05,
+      "loss": 0.2037,
+      "step": 7116
+    },
+    {
+      "epoch": 0.113872,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.932903225806452e-05,
+      "loss": 0.1441,
+      "step": 7117
+    },
+    {
+      "epoch": 0.113888,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.932741935483872e-05,
+      "loss": 0.2117,
+      "step": 7118
+    },
+    {
+      "epoch": 0.113904,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.93258064516129e-05,
+      "loss": 0.1666,
+      "step": 7119
+    },
+    {
+      "epoch": 0.11392,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.93241935483871e-05,
+      "loss": 0.1497,
+      "step": 7120
+    },
+    {
+      "epoch": 0.113936,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.932258064516129e-05,
+      "loss": 0.1925,
+      "step": 7121
+    },
+    {
+      "epoch": 0.113952,
+      "grad_norm": 2.125,
+      "learning_rate": 8.932096774193549e-05,
+      "loss": 0.2082,
+      "step": 7122
+    },
+    {
+      "epoch": 0.113968,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.931935483870967e-05,
+      "loss": 0.1503,
+      "step": 7123
+    },
+    {
+      "epoch": 0.113984,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.931774193548387e-05,
+      "loss": 0.1957,
+      "step": 7124
+    },
+    {
+      "epoch": 0.114,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.931612903225806e-05,
+      "loss": 0.1664,
+      "step": 7125
+    },
+    {
+      "epoch": 0.114016,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.931451612903226e-05,
+      "loss": 0.1689,
+      "step": 7126
+    },
+    {
+      "epoch": 0.114032,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.931290322580646e-05,
+      "loss": 0.1742,
+      "step": 7127
+    },
+    {
+      "epoch": 0.114048,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.931129032258066e-05,
+      "loss": 0.2017,
+      "step": 7128
+    },
+    {
+      "epoch": 0.114064,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.930967741935485e-05,
+      "loss": 0.1966,
+      "step": 7129
+    },
+    {
+      "epoch": 0.11408,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.930806451612904e-05,
+      "loss": 0.1661,
+      "step": 7130
+    },
+    {
+      "epoch": 0.114096,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.930645161290323e-05,
+      "loss": 0.1623,
+      "step": 7131
+    },
+    {
+      "epoch": 0.114112,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.930483870967742e-05,
+      "loss": 0.1818,
+      "step": 7132
+    },
+    {
+      "epoch": 0.114128,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.930322580645162e-05,
+      "loss": 0.1615,
+      "step": 7133
+    },
+    {
+      "epoch": 0.114144,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.93016129032258e-05,
+      "loss": 0.1711,
+      "step": 7134
+    },
+    {
+      "epoch": 0.11416,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.93e-05,
+      "loss": 0.1804,
+      "step": 7135
+    },
+    {
+      "epoch": 0.114176,
+      "grad_norm": 1.421875,
+      "learning_rate": 8.929838709677419e-05,
+      "loss": 0.222,
+      "step": 7136
+    },
+    {
+      "epoch": 0.114192,
+      "grad_norm": 1.84375,
+      "learning_rate": 8.929677419354839e-05,
+      "loss": 0.2287,
+      "step": 7137
+    },
+    {
+      "epoch": 0.114208,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.929516129032259e-05,
+      "loss": 0.1902,
+      "step": 7138
+    },
+    {
+      "epoch": 0.114224,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.929354838709679e-05,
+      "loss": 0.1753,
+      "step": 7139
+    },
+    {
+      "epoch": 0.11424,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.929193548387097e-05,
+      "loss": 0.1905,
+      "step": 7140
+    },
+    {
+      "epoch": 0.114256,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.929032258064517e-05,
+      "loss": 0.1547,
+      "step": 7141
+    },
+    {
+      "epoch": 0.114272,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.928870967741936e-05,
+      "loss": 0.1859,
+      "step": 7142
+    },
+    {
+      "epoch": 0.114288,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.928709677419356e-05,
+      "loss": 0.1964,
+      "step": 7143
+    },
+    {
+      "epoch": 0.114304,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.928548387096774e-05,
+      "loss": 0.1897,
+      "step": 7144
+    },
+    {
+      "epoch": 0.11432,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.928387096774194e-05,
+      "loss": 0.1594,
+      "step": 7145
+    },
+    {
+      "epoch": 0.114336,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.928225806451613e-05,
+      "loss": 0.1962,
+      "step": 7146
+    },
+    {
+      "epoch": 0.114352,
+      "grad_norm": 0.875,
+      "learning_rate": 8.928064516129032e-05,
+      "loss": 0.1553,
+      "step": 7147
+    },
+    {
+      "epoch": 0.114368,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.927903225806452e-05,
+      "loss": 0.1542,
+      "step": 7148
+    },
+    {
+      "epoch": 0.114384,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.927741935483872e-05,
+      "loss": 0.1951,
+      "step": 7149
+    },
+    {
+      "epoch": 0.1144,
+      "grad_norm": 1.125,
+      "learning_rate": 8.927580645161292e-05,
+      "loss": 0.2121,
+      "step": 7150
+    },
+    {
+      "epoch": 0.114416,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.92741935483871e-05,
+      "loss": 0.1528,
+      "step": 7151
+    },
+    {
+      "epoch": 0.114432,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.92725806451613e-05,
+      "loss": 0.2075,
+      "step": 7152
+    },
+    {
+      "epoch": 0.114448,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.927096774193549e-05,
+      "loss": 0.1746,
+      "step": 7153
+    },
+    {
+      "epoch": 0.114464,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.926935483870969e-05,
+      "loss": 0.1803,
+      "step": 7154
+    },
+    {
+      "epoch": 0.11448,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.926774193548387e-05,
+      "loss": 0.1997,
+      "step": 7155
+    },
+    {
+      "epoch": 0.114496,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.926612903225807e-05,
+      "loss": 0.1767,
+      "step": 7156
+    },
+    {
+      "epoch": 0.114512,
+      "grad_norm": 1.3671875,
+      "learning_rate": 8.926451612903226e-05,
+      "loss": 0.1717,
+      "step": 7157
+    },
+    {
+      "epoch": 0.114528,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.926290322580646e-05,
+      "loss": 0.1664,
+      "step": 7158
+    },
+    {
+      "epoch": 0.114544,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.926129032258064e-05,
+      "loss": 0.1553,
+      "step": 7159
+    },
+    {
+      "epoch": 0.11456,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.925967741935484e-05,
+      "loss": 0.1833,
+      "step": 7160
+    },
+    {
+      "epoch": 0.114576,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.925806451612903e-05,
+      "loss": 0.181,
+      "step": 7161
+    },
+    {
+      "epoch": 0.114592,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.925645161290323e-05,
+      "loss": 0.1649,
+      "step": 7162
+    },
+    {
+      "epoch": 0.114608,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.925483870967743e-05,
+      "loss": 0.1554,
+      "step": 7163
+    },
+    {
+      "epoch": 0.114624,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.925322580645162e-05,
+      "loss": 0.2164,
+      "step": 7164
+    },
+    {
+      "epoch": 0.11464,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.925161290322581e-05,
+      "loss": 0.1941,
+      "step": 7165
+    },
+    {
+      "epoch": 0.114656,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.925e-05,
+      "loss": 0.184,
+      "step": 7166
+    },
+    {
+      "epoch": 0.114672,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.92483870967742e-05,
+      "loss": 0.1605,
+      "step": 7167
+    },
+    {
+      "epoch": 0.114688,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.924677419354839e-05,
+      "loss": 0.1415,
+      "step": 7168
+    },
+    {
+      "epoch": 0.114704,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.924516129032259e-05,
+      "loss": 0.1603,
+      "step": 7169
+    },
+    {
+      "epoch": 0.11472,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.924354838709677e-05,
+      "loss": 0.1748,
+      "step": 7170
+    },
+    {
+      "epoch": 0.114736,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.924193548387097e-05,
+      "loss": 0.2033,
+      "step": 7171
+    },
+    {
+      "epoch": 0.114752,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.924032258064516e-05,
+      "loss": 0.1808,
+      "step": 7172
+    },
+    {
+      "epoch": 0.114768,
+      "grad_norm": 1.0,
+      "learning_rate": 8.923870967741936e-05,
+      "loss": 0.1637,
+      "step": 7173
+    },
+    {
+      "epoch": 0.114784,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.923709677419356e-05,
+      "loss": 0.2137,
+      "step": 7174
+    },
+    {
+      "epoch": 0.1148,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.923548387096776e-05,
+      "loss": 0.2017,
+      "step": 7175
+    },
+    {
+      "epoch": 0.114816,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.923387096774194e-05,
+      "loss": 0.1818,
+      "step": 7176
+    },
+    {
+      "epoch": 0.114832,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.923225806451614e-05,
+      "loss": 0.2574,
+      "step": 7177
+    },
+    {
+      "epoch": 0.114848,
+      "grad_norm": 1.484375,
+      "learning_rate": 8.923064516129033e-05,
+      "loss": 0.237,
+      "step": 7178
+    },
+    {
+      "epoch": 0.114864,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.922903225806451e-05,
+      "loss": 0.209,
+      "step": 7179
+    },
+    {
+      "epoch": 0.11488,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.922741935483871e-05,
+      "loss": 0.2045,
+      "step": 7180
+    },
+    {
+      "epoch": 0.114896,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.92258064516129e-05,
+      "loss": 0.1674,
+      "step": 7181
+    },
+    {
+      "epoch": 0.114912,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.92241935483871e-05,
+      "loss": 0.1338,
+      "step": 7182
+    },
+    {
+      "epoch": 0.114928,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.922258064516129e-05,
+      "loss": 0.179,
+      "step": 7183
+    },
+    {
+      "epoch": 0.114944,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.922096774193549e-05,
+      "loss": 0.1827,
+      "step": 7184
+    },
+    {
+      "epoch": 0.11496,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.921935483870969e-05,
+      "loss": 0.1836,
+      "step": 7185
+    },
+    {
+      "epoch": 0.114976,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.921774193548387e-05,
+      "loss": 0.1293,
+      "step": 7186
+    },
+    {
+      "epoch": 0.114992,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.921612903225807e-05,
+      "loss": 0.1275,
+      "step": 7187
+    },
+    {
+      "epoch": 0.115008,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.921451612903227e-05,
+      "loss": 0.1899,
+      "step": 7188
+    },
+    {
+      "epoch": 0.115024,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.921290322580646e-05,
+      "loss": 0.171,
+      "step": 7189
+    },
+    {
+      "epoch": 0.11504,
+      "grad_norm": 1.125,
+      "learning_rate": 8.921129032258066e-05,
+      "loss": 0.1834,
+      "step": 7190
+    },
+    {
+      "epoch": 0.115056,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.920967741935484e-05,
+      "loss": 0.1474,
+      "step": 7191
+    },
+    {
+      "epoch": 0.115072,
+      "grad_norm": 1.359375,
+      "learning_rate": 8.920806451612904e-05,
+      "loss": 0.1745,
+      "step": 7192
+    },
+    {
+      "epoch": 0.115088,
+      "grad_norm": 1.53125,
+      "learning_rate": 8.920645161290323e-05,
+      "loss": 0.266,
+      "step": 7193
+    },
+    {
+      "epoch": 0.115104,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.920483870967741e-05,
+      "loss": 0.1739,
+      "step": 7194
+    },
+    {
+      "epoch": 0.11512,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.920322580645161e-05,
+      "loss": 0.1762,
+      "step": 7195
+    },
+    {
+      "epoch": 0.115136,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.92016129032258e-05,
+      "loss": 0.1928,
+      "step": 7196
+    },
+    {
+      "epoch": 0.115152,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.92e-05,
+      "loss": 0.1443,
+      "step": 7197
+    },
+    {
+      "epoch": 0.115168,
+      "grad_norm": 1.3359375,
+      "learning_rate": 8.91983870967742e-05,
+      "loss": 0.1638,
+      "step": 7198
+    },
+    {
+      "epoch": 0.115184,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.91967741935484e-05,
+      "loss": 0.1839,
+      "step": 7199
+    },
+    {
+      "epoch": 0.1152,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.919516129032259e-05,
+      "loss": 0.1737,
+      "step": 7200
+    },
+    {
+      "epoch": 0.115216,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.919354838709678e-05,
+      "loss": 0.1597,
+      "step": 7201
+    },
+    {
+      "epoch": 0.115232,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.919193548387097e-05,
+      "loss": 0.1816,
+      "step": 7202
+    },
+    {
+      "epoch": 0.115248,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.919032258064517e-05,
+      "loss": 0.1901,
+      "step": 7203
+    },
+    {
+      "epoch": 0.115264,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.918870967741936e-05,
+      "loss": 0.1459,
+      "step": 7204
+    },
+    {
+      "epoch": 0.11528,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.918709677419356e-05,
+      "loss": 0.2012,
+      "step": 7205
+    },
+    {
+      "epoch": 0.115296,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.918548387096774e-05,
+      "loss": 0.2197,
+      "step": 7206
+    },
+    {
+      "epoch": 0.115312,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.918387096774194e-05,
+      "loss": 0.1871,
+      "step": 7207
+    },
+    {
+      "epoch": 0.115328,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.918225806451613e-05,
+      "loss": 0.1443,
+      "step": 7208
+    },
+    {
+      "epoch": 0.115344,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.918064516129033e-05,
+      "loss": 0.176,
+      "step": 7209
+    },
+    {
+      "epoch": 0.11536,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.917903225806453e-05,
+      "loss": 0.1585,
+      "step": 7210
+    },
+    {
+      "epoch": 0.115376,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.917741935483871e-05,
+      "loss": 0.176,
+      "step": 7211
+    },
+    {
+      "epoch": 0.115392,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.917580645161291e-05,
+      "loss": 0.1525,
+      "step": 7212
+    },
+    {
+      "epoch": 0.115408,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.91741935483871e-05,
+      "loss": 0.1807,
+      "step": 7213
+    },
+    {
+      "epoch": 0.115424,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.91725806451613e-05,
+      "loss": 0.125,
+      "step": 7214
+    },
+    {
+      "epoch": 0.11544,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.917096774193548e-05,
+      "loss": 0.1807,
+      "step": 7215
+    },
+    {
+      "epoch": 0.115456,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.916935483870968e-05,
+      "loss": 0.1681,
+      "step": 7216
+    },
+    {
+      "epoch": 0.115472,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.916774193548387e-05,
+      "loss": 0.1571,
+      "step": 7217
+    },
+    {
+      "epoch": 0.115488,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.916612903225807e-05,
+      "loss": 0.1418,
+      "step": 7218
+    },
+    {
+      "epoch": 0.115504,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.916451612903226e-05,
+      "loss": 0.162,
+      "step": 7219
+    },
+    {
+      "epoch": 0.11552,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.916290322580646e-05,
+      "loss": 0.2171,
+      "step": 7220
+    },
+    {
+      "epoch": 0.115536,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.916129032258064e-05,
+      "loss": 0.1708,
+      "step": 7221
+    },
+    {
+      "epoch": 0.115552,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.915967741935484e-05,
+      "loss": 0.2286,
+      "step": 7222
+    },
+    {
+      "epoch": 0.115568,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.915806451612904e-05,
+      "loss": 0.2092,
+      "step": 7223
+    },
+    {
+      "epoch": 0.115584,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.915645161290323e-05,
+      "loss": 0.2266,
+      "step": 7224
+    },
+    {
+      "epoch": 0.1156,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.915483870967743e-05,
+      "loss": 0.1897,
+      "step": 7225
+    },
+    {
+      "epoch": 0.115616,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.915322580645161e-05,
+      "loss": 0.1676,
+      "step": 7226
+    },
+    {
+      "epoch": 0.115632,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.915161290322581e-05,
+      "loss": 0.1759,
+      "step": 7227
+    },
+    {
+      "epoch": 0.115648,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.915e-05,
+      "loss": 0.1641,
+      "step": 7228
+    },
+    {
+      "epoch": 0.115664,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.91483870967742e-05,
+      "loss": 0.1643,
+      "step": 7229
+    },
+    {
+      "epoch": 0.11568,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.914677419354838e-05,
+      "loss": 0.1467,
+      "step": 7230
+    },
+    {
+      "epoch": 0.115696,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.914516129032258e-05,
+      "loss": 0.1225,
+      "step": 7231
+    },
+    {
+      "epoch": 0.115712,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.914354838709677e-05,
+      "loss": 0.1311,
+      "step": 7232
+    },
+    {
+      "epoch": 0.115728,
+      "grad_norm": 1.328125,
+      "learning_rate": 8.914193548387097e-05,
+      "loss": 0.2322,
+      "step": 7233
+    },
+    {
+      "epoch": 0.115744,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.914032258064517e-05,
+      "loss": 0.1719,
+      "step": 7234
+    },
+    {
+      "epoch": 0.11576,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.913870967741937e-05,
+      "loss": 0.1667,
+      "step": 7235
+    },
+    {
+      "epoch": 0.115776,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.913709677419355e-05,
+      "loss": 0.157,
+      "step": 7236
+    },
+    {
+      "epoch": 0.115792,
+      "grad_norm": 1.5,
+      "learning_rate": 8.913548387096775e-05,
+      "loss": 0.1545,
+      "step": 7237
+    },
+    {
+      "epoch": 0.115808,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.913387096774194e-05,
+      "loss": 0.1772,
+      "step": 7238
+    },
+    {
+      "epoch": 0.115824,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.913225806451614e-05,
+      "loss": 0.1487,
+      "step": 7239
+    },
+    {
+      "epoch": 0.11584,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.913064516129033e-05,
+      "loss": 0.1654,
+      "step": 7240
+    },
+    {
+      "epoch": 0.115856,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.912903225806451e-05,
+      "loss": 0.2039,
+      "step": 7241
+    },
+    {
+      "epoch": 0.115872,
+      "grad_norm": 1.5625,
+      "learning_rate": 8.912741935483871e-05,
+      "loss": 0.2055,
+      "step": 7242
+    },
+    {
+      "epoch": 0.115888,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.91258064516129e-05,
+      "loss": 0.1798,
+      "step": 7243
+    },
+    {
+      "epoch": 0.115904,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.91241935483871e-05,
+      "loss": 0.1288,
+      "step": 7244
+    },
+    {
+      "epoch": 0.11592,
+      "grad_norm": 1.296875,
+      "learning_rate": 8.91225806451613e-05,
+      "loss": 0.2343,
+      "step": 7245
+    },
+    {
+      "epoch": 0.115936,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.91209677419355e-05,
+      "loss": 0.2206,
+      "step": 7246
+    },
+    {
+      "epoch": 0.115952,
+      "grad_norm": 0.95703125,
+      "learning_rate": 8.911935483870968e-05,
+      "loss": 0.1958,
+      "step": 7247
+    },
+    {
+      "epoch": 0.115968,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.911774193548388e-05,
+      "loss": 0.1371,
+      "step": 7248
+    },
+    {
+      "epoch": 0.115984,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.911612903225807e-05,
+      "loss": 0.1836,
+      "step": 7249
+    },
+    {
+      "epoch": 0.116,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.911451612903227e-05,
+      "loss": 0.1534,
+      "step": 7250
+    },
+    {
+      "epoch": 0.116016,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.911290322580645e-05,
+      "loss": 0.1571,
+      "step": 7251
+    },
+    {
+      "epoch": 0.116032,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.911129032258065e-05,
+      "loss": 0.215,
+      "step": 7252
+    },
+    {
+      "epoch": 0.116048,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.910967741935484e-05,
+      "loss": 0.171,
+      "step": 7253
+    },
+    {
+      "epoch": 0.116064,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.910806451612904e-05,
+      "loss": 0.1395,
+      "step": 7254
+    },
+    {
+      "epoch": 0.11608,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.910645161290323e-05,
+      "loss": 0.2327,
+      "step": 7255
+    },
+    {
+      "epoch": 0.116096,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.910483870967741e-05,
+      "loss": 0.1708,
+      "step": 7256
+    },
+    {
+      "epoch": 0.116112,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.910322580645161e-05,
+      "loss": 0.179,
+      "step": 7257
+    },
+    {
+      "epoch": 0.116128,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.910161290322581e-05,
+      "loss": 0.1799,
+      "step": 7258
+    },
+    {
+      "epoch": 0.116144,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.910000000000001e-05,
+      "loss": 0.203,
+      "step": 7259
+    },
+    {
+      "epoch": 0.11616,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.90983870967742e-05,
+      "loss": 0.1881,
+      "step": 7260
+    },
+    {
+      "epoch": 0.116176,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.90967741935484e-05,
+      "loss": 0.157,
+      "step": 7261
+    },
+    {
+      "epoch": 0.116192,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.909516129032258e-05,
+      "loss": 0.1887,
+      "step": 7262
+    },
+    {
+      "epoch": 0.116208,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.909354838709678e-05,
+      "loss": 0.1359,
+      "step": 7263
+    },
+    {
+      "epoch": 0.116224,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.909193548387097e-05,
+      "loss": 0.1636,
+      "step": 7264
+    },
+    {
+      "epoch": 0.11624,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.909032258064517e-05,
+      "loss": 0.1254,
+      "step": 7265
+    },
+    {
+      "epoch": 0.116256,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.908870967741935e-05,
+      "loss": 0.1662,
+      "step": 7266
+    },
+    {
+      "epoch": 0.116272,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.908709677419355e-05,
+      "loss": 0.1863,
+      "step": 7267
+    },
+    {
+      "epoch": 0.116288,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.908548387096774e-05,
+      "loss": 0.1972,
+      "step": 7268
+    },
+    {
+      "epoch": 0.116304,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.908387096774194e-05,
+      "loss": 0.1734,
+      "step": 7269
+    },
+    {
+      "epoch": 0.11632,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.908225806451614e-05,
+      "loss": 0.1693,
+      "step": 7270
+    },
+    {
+      "epoch": 0.116336,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.908064516129033e-05,
+      "loss": 0.1962,
+      "step": 7271
+    },
+    {
+      "epoch": 0.116352,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.907903225806452e-05,
+      "loss": 0.1797,
+      "step": 7272
+    },
+    {
+      "epoch": 0.116368,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.907741935483871e-05,
+      "loss": 0.1945,
+      "step": 7273
+    },
+    {
+      "epoch": 0.116384,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.907580645161291e-05,
+      "loss": 0.2011,
+      "step": 7274
+    },
+    {
+      "epoch": 0.1164,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.90741935483871e-05,
+      "loss": 0.1682,
+      "step": 7275
+    },
+    {
+      "epoch": 0.116416,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.90725806451613e-05,
+      "loss": 0.1529,
+      "step": 7276
+    },
+    {
+      "epoch": 0.116432,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.907096774193548e-05,
+      "loss": 0.18,
+      "step": 7277
+    },
+    {
+      "epoch": 0.116448,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.906935483870968e-05,
+      "loss": 0.1741,
+      "step": 7278
+    },
+    {
+      "epoch": 0.116464,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.906774193548387e-05,
+      "loss": 0.1565,
+      "step": 7279
+    },
+    {
+      "epoch": 0.11648,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.906612903225807e-05,
+      "loss": 0.1782,
+      "step": 7280
+    },
+    {
+      "epoch": 0.116496,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.906451612903227e-05,
+      "loss": 0.1787,
+      "step": 7281
+    },
+    {
+      "epoch": 0.116512,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.906290322580645e-05,
+      "loss": 0.1706,
+      "step": 7282
+    },
+    {
+      "epoch": 0.116528,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.906129032258065e-05,
+      "loss": 0.1781,
+      "step": 7283
+    },
+    {
+      "epoch": 0.116544,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.905967741935485e-05,
+      "loss": 0.191,
+      "step": 7284
+    },
+    {
+      "epoch": 0.11656,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.905806451612904e-05,
+      "loss": 0.2171,
+      "step": 7285
+    },
+    {
+      "epoch": 0.116576,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.905645161290324e-05,
+      "loss": 0.1733,
+      "step": 7286
+    },
+    {
+      "epoch": 0.116592,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.905483870967742e-05,
+      "loss": 0.1802,
+      "step": 7287
+    },
+    {
+      "epoch": 0.116608,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.905322580645161e-05,
+      "loss": 0.1974,
+      "step": 7288
+    },
+    {
+      "epoch": 0.116624,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.905161290322581e-05,
+      "loss": 0.2005,
+      "step": 7289
+    },
+    {
+      "epoch": 0.11664,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.905e-05,
+      "loss": 0.1945,
+      "step": 7290
+    },
+    {
+      "epoch": 0.116656,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.90483870967742e-05,
+      "loss": 0.1491,
+      "step": 7291
+    },
+    {
+      "epoch": 0.116672,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.904677419354838e-05,
+      "loss": 0.2105,
+      "step": 7292
+    },
+    {
+      "epoch": 0.116688,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.904516129032258e-05,
+      "loss": 0.1995,
+      "step": 7293
+    },
+    {
+      "epoch": 0.116704,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.904354838709678e-05,
+      "loss": 0.1918,
+      "step": 7294
+    },
+    {
+      "epoch": 0.11672,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.904193548387098e-05,
+      "loss": 0.1284,
+      "step": 7295
+    },
+    {
+      "epoch": 0.116736,
+      "grad_norm": 1.0,
+      "learning_rate": 8.904032258064517e-05,
+      "loss": 0.1688,
+      "step": 7296
+    },
+    {
+      "epoch": 0.116752,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.903870967741937e-05,
+      "loss": 0.1563,
+      "step": 7297
+    },
+    {
+      "epoch": 0.116768,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.903709677419355e-05,
+      "loss": 0.1892,
+      "step": 7298
+    },
+    {
+      "epoch": 0.116784,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.903548387096775e-05,
+      "loss": 0.1751,
+      "step": 7299
+    },
+    {
+      "epoch": 0.1168,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.903387096774194e-05,
+      "loss": 0.19,
+      "step": 7300
+    },
+    {
+      "epoch": 0.116816,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.903225806451614e-05,
+      "loss": 0.1805,
+      "step": 7301
+    },
+    {
+      "epoch": 0.116832,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.903064516129032e-05,
+      "loss": 0.2033,
+      "step": 7302
+    },
+    {
+      "epoch": 0.116848,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.902903225806451e-05,
+      "loss": 0.1527,
+      "step": 7303
+    },
+    {
+      "epoch": 0.116864,
+      "grad_norm": 0.75,
+      "learning_rate": 8.902741935483871e-05,
+      "loss": 0.1834,
+      "step": 7304
+    },
+    {
+      "epoch": 0.11688,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.902580645161291e-05,
+      "loss": 0.19,
+      "step": 7305
+    },
+    {
+      "epoch": 0.116896,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.902419354838711e-05,
+      "loss": 0.2019,
+      "step": 7306
+    },
+    {
+      "epoch": 0.116912,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.90225806451613e-05,
+      "loss": 0.1994,
+      "step": 7307
+    },
+    {
+      "epoch": 0.116928,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.90209677419355e-05,
+      "loss": 0.1434,
+      "step": 7308
+    },
+    {
+      "epoch": 0.116944,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.901935483870968e-05,
+      "loss": 0.1979,
+      "step": 7309
+    },
+    {
+      "epoch": 0.11696,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.901774193548388e-05,
+      "loss": 0.1974,
+      "step": 7310
+    },
+    {
+      "epoch": 0.116976,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.901612903225807e-05,
+      "loss": 0.1728,
+      "step": 7311
+    },
+    {
+      "epoch": 0.116992,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.901451612903227e-05,
+      "loss": 0.1658,
+      "step": 7312
+    },
+    {
+      "epoch": 0.117008,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.901290322580645e-05,
+      "loss": 0.1526,
+      "step": 7313
+    },
+    {
+      "epoch": 0.117024,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.901129032258065e-05,
+      "loss": 0.1558,
+      "step": 7314
+    },
+    {
+      "epoch": 0.11704,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.900967741935484e-05,
+      "loss": 0.1642,
+      "step": 7315
+    },
+    {
+      "epoch": 0.117056,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.900806451612904e-05,
+      "loss": 0.1751,
+      "step": 7316
+    },
+    {
+      "epoch": 0.117072,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.900645161290322e-05,
+      "loss": 0.2094,
+      "step": 7317
+    },
+    {
+      "epoch": 0.117088,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.900483870967742e-05,
+      "loss": 0.1736,
+      "step": 7318
+    },
+    {
+      "epoch": 0.117104,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.900322580645162e-05,
+      "loss": 0.1971,
+      "step": 7319
+    },
+    {
+      "epoch": 0.11712,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.900161290322581e-05,
+      "loss": 0.1887,
+      "step": 7320
+    },
+    {
+      "epoch": 0.117136,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.900000000000001e-05,
+      "loss": 0.2097,
+      "step": 7321
+    },
+    {
+      "epoch": 0.117152,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.89983870967742e-05,
+      "loss": 0.1898,
+      "step": 7322
+    },
+    {
+      "epoch": 0.117168,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.89967741935484e-05,
+      "loss": 0.1781,
+      "step": 7323
+    },
+    {
+      "epoch": 0.117184,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.899516129032258e-05,
+      "loss": 0.1691,
+      "step": 7324
+    },
+    {
+      "epoch": 0.1172,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.899354838709678e-05,
+      "loss": 0.1614,
+      "step": 7325
+    },
+    {
+      "epoch": 0.117216,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.899193548387097e-05,
+      "loss": 0.1944,
+      "step": 7326
+    },
+    {
+      "epoch": 0.117232,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.899032258064517e-05,
+      "loss": 0.2247,
+      "step": 7327
+    },
+    {
+      "epoch": 0.117248,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.898870967741935e-05,
+      "loss": 0.1861,
+      "step": 7328
+    },
+    {
+      "epoch": 0.117264,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.898709677419355e-05,
+      "loss": 0.1799,
+      "step": 7329
+    },
+    {
+      "epoch": 0.11728,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.898548387096775e-05,
+      "loss": 0.1608,
+      "step": 7330
+    },
+    {
+      "epoch": 0.117296,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.898387096774195e-05,
+      "loss": 0.163,
+      "step": 7331
+    },
+    {
+      "epoch": 0.117312,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.898225806451614e-05,
+      "loss": 0.1881,
+      "step": 7332
+    },
+    {
+      "epoch": 0.117328,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.898064516129032e-05,
+      "loss": 0.1926,
+      "step": 7333
+    },
+    {
+      "epoch": 0.117344,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.897903225806452e-05,
+      "loss": 0.1547,
+      "step": 7334
+    },
+    {
+      "epoch": 0.11736,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.897741935483871e-05,
+      "loss": 0.1763,
+      "step": 7335
+    },
+    {
+      "epoch": 0.117376,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.897580645161291e-05,
+      "loss": 0.2081,
+      "step": 7336
+    },
+    {
+      "epoch": 0.117392,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.89741935483871e-05,
+      "loss": 0.1804,
+      "step": 7337
+    },
+    {
+      "epoch": 0.117408,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.89725806451613e-05,
+      "loss": 0.1755,
+      "step": 7338
+    },
+    {
+      "epoch": 0.117424,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.897096774193548e-05,
+      "loss": 0.1986,
+      "step": 7339
+    },
+    {
+      "epoch": 0.11744,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.896935483870968e-05,
+      "loss": 0.1512,
+      "step": 7340
+    },
+    {
+      "epoch": 0.117456,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.896774193548388e-05,
+      "loss": 0.1717,
+      "step": 7341
+    },
+    {
+      "epoch": 0.117472,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.896612903225808e-05,
+      "loss": 0.1645,
+      "step": 7342
+    },
+    {
+      "epoch": 0.117488,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.896451612903226e-05,
+      "loss": 0.1859,
+      "step": 7343
+    },
+    {
+      "epoch": 0.117504,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.896290322580646e-05,
+      "loss": 0.1781,
+      "step": 7344
+    },
+    {
+      "epoch": 0.11752,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.896129032258065e-05,
+      "loss": 0.1543,
+      "step": 7345
+    },
+    {
+      "epoch": 0.117536,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.895967741935485e-05,
+      "loss": 0.2112,
+      "step": 7346
+    },
+    {
+      "epoch": 0.117552,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.895806451612904e-05,
+      "loss": 0.1759,
+      "step": 7347
+    },
+    {
+      "epoch": 0.117568,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.895645161290324e-05,
+      "loss": 0.17,
+      "step": 7348
+    },
+    {
+      "epoch": 0.117584,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.895483870967742e-05,
+      "loss": 0.2091,
+      "step": 7349
+    },
+    {
+      "epoch": 0.1176,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.895322580645161e-05,
+      "loss": 0.2365,
+      "step": 7350
+    },
+    {
+      "epoch": 0.117616,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.895161290322581e-05,
+      "loss": 0.2081,
+      "step": 7351
+    },
+    {
+      "epoch": 0.117632,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.895e-05,
+      "loss": 0.1375,
+      "step": 7352
+    },
+    {
+      "epoch": 0.117648,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.89483870967742e-05,
+      "loss": 0.1867,
+      "step": 7353
+    },
+    {
+      "epoch": 0.117664,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.894677419354839e-05,
+      "loss": 0.1968,
+      "step": 7354
+    },
+    {
+      "epoch": 0.11768,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.894516129032259e-05,
+      "loss": 0.1574,
+      "step": 7355
+    },
+    {
+      "epoch": 0.117696,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.894354838709678e-05,
+      "loss": 0.2203,
+      "step": 7356
+    },
+    {
+      "epoch": 0.117712,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.894193548387098e-05,
+      "loss": 0.1342,
+      "step": 7357
+    },
+    {
+      "epoch": 0.117728,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.894032258064516e-05,
+      "loss": 0.1475,
+      "step": 7358
+    },
+    {
+      "epoch": 0.117744,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.893870967741936e-05,
+      "loss": 0.1902,
+      "step": 7359
+    },
+    {
+      "epoch": 0.11776,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.893709677419355e-05,
+      "loss": 0.1712,
+      "step": 7360
+    },
+    {
+      "epoch": 0.117776,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.893548387096775e-05,
+      "loss": 0.154,
+      "step": 7361
+    },
+    {
+      "epoch": 0.117792,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.893387096774194e-05,
+      "loss": 0.1885,
+      "step": 7362
+    },
+    {
+      "epoch": 0.117808,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.893225806451614e-05,
+      "loss": 0.1908,
+      "step": 7363
+    },
+    {
+      "epoch": 0.117824,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.893064516129032e-05,
+      "loss": 0.2123,
+      "step": 7364
+    },
+    {
+      "epoch": 0.11784,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.892903225806452e-05,
+      "loss": 0.1823,
+      "step": 7365
+    },
+    {
+      "epoch": 0.117856,
+      "grad_norm": 1.0,
+      "learning_rate": 8.892741935483872e-05,
+      "loss": 0.1282,
+      "step": 7366
+    },
+    {
+      "epoch": 0.117872,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.892580645161291e-05,
+      "loss": 0.1667,
+      "step": 7367
+    },
+    {
+      "epoch": 0.117888,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.89241935483871e-05,
+      "loss": 0.1594,
+      "step": 7368
+    },
+    {
+      "epoch": 0.117904,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.892258064516129e-05,
+      "loss": 0.2157,
+      "step": 7369
+    },
+    {
+      "epoch": 0.11792,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.892096774193549e-05,
+      "loss": 0.1288,
+      "step": 7370
+    },
+    {
+      "epoch": 0.117936,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.891935483870968e-05,
+      "loss": 0.1728,
+      "step": 7371
+    },
+    {
+      "epoch": 0.117952,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.891774193548388e-05,
+      "loss": 0.1643,
+      "step": 7372
+    },
+    {
+      "epoch": 0.117968,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.891612903225806e-05,
+      "loss": 0.195,
+      "step": 7373
+    },
+    {
+      "epoch": 0.117984,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.891451612903226e-05,
+      "loss": 0.158,
+      "step": 7374
+    },
+    {
+      "epoch": 0.118,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.891290322580645e-05,
+      "loss": 0.1718,
+      "step": 7375
+    },
+    {
+      "epoch": 0.118016,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.891129032258065e-05,
+      "loss": 0.1827,
+      "step": 7376
+    },
+    {
+      "epoch": 0.118032,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.890967741935484e-05,
+      "loss": 0.1814,
+      "step": 7377
+    },
+    {
+      "epoch": 0.118048,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.890806451612903e-05,
+      "loss": 0.175,
+      "step": 7378
+    },
+    {
+      "epoch": 0.118064,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.890645161290323e-05,
+      "loss": 0.1862,
+      "step": 7379
+    },
+    {
+      "epoch": 0.11808,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.890483870967742e-05,
+      "loss": 0.2109,
+      "step": 7380
+    },
+    {
+      "epoch": 0.118096,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.890322580645162e-05,
+      "loss": 0.1465,
+      "step": 7381
+    },
+    {
+      "epoch": 0.118112,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.89016129032258e-05,
+      "loss": 0.2075,
+      "step": 7382
+    },
+    {
+      "epoch": 0.118128,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.89e-05,
+      "loss": 0.1828,
+      "step": 7383
+    },
+    {
+      "epoch": 0.118144,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.889838709677419e-05,
+      "loss": 0.2135,
+      "step": 7384
+    },
+    {
+      "epoch": 0.11816,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.889677419354839e-05,
+      "loss": 0.2198,
+      "step": 7385
+    },
+    {
+      "epoch": 0.118176,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.889516129032258e-05,
+      "loss": 0.177,
+      "step": 7386
+    },
+    {
+      "epoch": 0.118192,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.889354838709678e-05,
+      "loss": 0.1487,
+      "step": 7387
+    },
+    {
+      "epoch": 0.118208,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.889193548387096e-05,
+      "loss": 0.1546,
+      "step": 7388
+    },
+    {
+      "epoch": 0.118224,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.889032258064516e-05,
+      "loss": 0.1709,
+      "step": 7389
+    },
+    {
+      "epoch": 0.11824,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.888870967741936e-05,
+      "loss": 0.1549,
+      "step": 7390
+    },
+    {
+      "epoch": 0.118256,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.888709677419356e-05,
+      "loss": 0.1733,
+      "step": 7391
+    },
+    {
+      "epoch": 0.118272,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.888548387096775e-05,
+      "loss": 0.188,
+      "step": 7392
+    },
+    {
+      "epoch": 0.118288,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.888387096774195e-05,
+      "loss": 0.2022,
+      "step": 7393
+    },
+    {
+      "epoch": 0.118304,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.888225806451613e-05,
+      "loss": 0.1987,
+      "step": 7394
+    },
+    {
+      "epoch": 0.11832,
+      "grad_norm": 0.474609375,
+      "learning_rate": 8.888064516129033e-05,
+      "loss": 0.1425,
+      "step": 7395
+    },
+    {
+      "epoch": 0.118336,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.887903225806452e-05,
+      "loss": 0.1623,
+      "step": 7396
+    },
+    {
+      "epoch": 0.118352,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.88774193548387e-05,
+      "loss": 0.1565,
+      "step": 7397
+    },
+    {
+      "epoch": 0.118368,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.88758064516129e-05,
+      "loss": 0.1614,
+      "step": 7398
+    },
+    {
+      "epoch": 0.118384,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.887419354838709e-05,
+      "loss": 0.1836,
+      "step": 7399
+    },
+    {
+      "epoch": 0.1184,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.887258064516129e-05,
+      "loss": 0.1615,
+      "step": 7400
+    },
+    {
+      "epoch": 0.118416,
+      "grad_norm": 1.234375,
+      "learning_rate": 8.887096774193549e-05,
+      "loss": 0.1656,
+      "step": 7401
+    },
+    {
+      "epoch": 0.118432,
+      "grad_norm": 1.765625,
+      "learning_rate": 8.886935483870969e-05,
+      "loss": 0.1902,
+      "step": 7402
+    },
+    {
+      "epoch": 0.118448,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.886774193548388e-05,
+      "loss": 0.1788,
+      "step": 7403
+    },
+    {
+      "epoch": 0.118464,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.886612903225808e-05,
+      "loss": 0.1973,
+      "step": 7404
+    },
+    {
+      "epoch": 0.11848,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.886451612903226e-05,
+      "loss": 0.1948,
+      "step": 7405
+    },
+    {
+      "epoch": 0.118496,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.886290322580646e-05,
+      "loss": 0.1577,
+      "step": 7406
+    },
+    {
+      "epoch": 0.118512,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.886129032258065e-05,
+      "loss": 0.1876,
+      "step": 7407
+    },
+    {
+      "epoch": 0.118528,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.885967741935485e-05,
+      "loss": 0.2084,
+      "step": 7408
+    },
+    {
+      "epoch": 0.118544,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.885806451612903e-05,
+      "loss": 0.1569,
+      "step": 7409
+    },
+    {
+      "epoch": 0.11856,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.885645161290323e-05,
+      "loss": 0.2114,
+      "step": 7410
+    },
+    {
+      "epoch": 0.118576,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.885483870967742e-05,
+      "loss": 0.2038,
+      "step": 7411
+    },
+    {
+      "epoch": 0.118592,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.88532258064516e-05,
+      "loss": 0.1846,
+      "step": 7412
+    },
+    {
+      "epoch": 0.118608,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.88516129032258e-05,
+      "loss": 0.1913,
+      "step": 7413
+    },
+    {
+      "epoch": 0.118624,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.885e-05,
+      "loss": 0.1794,
+      "step": 7414
+    },
+    {
+      "epoch": 0.11864,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.88483870967742e-05,
+      "loss": 0.1827,
+      "step": 7415
+    },
+    {
+      "epoch": 0.118656,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.884677419354839e-05,
+      "loss": 0.1528,
+      "step": 7416
+    },
+    {
+      "epoch": 0.118672,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.884516129032259e-05,
+      "loss": 0.2039,
+      "step": 7417
+    },
+    {
+      "epoch": 0.118688,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.884354838709678e-05,
+      "loss": 0.1483,
+      "step": 7418
+    },
+    {
+      "epoch": 0.118704,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.884193548387098e-05,
+      "loss": 0.2197,
+      "step": 7419
+    },
+    {
+      "epoch": 0.11872,
+      "grad_norm": 1.125,
+      "learning_rate": 8.884032258064516e-05,
+      "loss": 0.1987,
+      "step": 7420
+    },
+    {
+      "epoch": 0.118736,
+      "grad_norm": 0.5,
+      "learning_rate": 8.883870967741936e-05,
+      "loss": 0.1381,
+      "step": 7421
+    },
+    {
+      "epoch": 0.118752,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.883709677419355e-05,
+      "loss": 0.1569,
+      "step": 7422
+    },
+    {
+      "epoch": 0.118768,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.883548387096775e-05,
+      "loss": 0.1838,
+      "step": 7423
+    },
+    {
+      "epoch": 0.118784,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.883387096774193e-05,
+      "loss": 0.1864,
+      "step": 7424
+    },
+    {
+      "epoch": 0.1188,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.883225806451613e-05,
+      "loss": 0.1946,
+      "step": 7425
+    },
+    {
+      "epoch": 0.118816,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.883064516129033e-05,
+      "loss": 0.2104,
+      "step": 7426
+    },
+    {
+      "epoch": 0.118832,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.882903225806452e-05,
+      "loss": 0.1576,
+      "step": 7427
+    },
+    {
+      "epoch": 0.118848,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.882741935483872e-05,
+      "loss": 0.1585,
+      "step": 7428
+    },
+    {
+      "epoch": 0.118864,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.88258064516129e-05,
+      "loss": 0.2087,
+      "step": 7429
+    },
+    {
+      "epoch": 0.11888,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.88241935483871e-05,
+      "loss": 0.165,
+      "step": 7430
+    },
+    {
+      "epoch": 0.118896,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.882258064516129e-05,
+      "loss": 0.1516,
+      "step": 7431
+    },
+    {
+      "epoch": 0.118912,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.882096774193549e-05,
+      "loss": 0.1939,
+      "step": 7432
+    },
+    {
+      "epoch": 0.118928,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.881935483870968e-05,
+      "loss": 0.2002,
+      "step": 7433
+    },
+    {
+      "epoch": 0.118944,
+      "grad_norm": 1.125,
+      "learning_rate": 8.881774193548388e-05,
+      "loss": 0.1627,
+      "step": 7434
+    },
+    {
+      "epoch": 0.11896,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.881612903225806e-05,
+      "loss": 0.1972,
+      "step": 7435
+    },
+    {
+      "epoch": 0.118976,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.881451612903226e-05,
+      "loss": 0.1877,
+      "step": 7436
+    },
+    {
+      "epoch": 0.118992,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.881290322580646e-05,
+      "loss": 0.1634,
+      "step": 7437
+    },
+    {
+      "epoch": 0.119008,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.881129032258065e-05,
+      "loss": 0.1784,
+      "step": 7438
+    },
+    {
+      "epoch": 0.119024,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.880967741935485e-05,
+      "loss": 0.2125,
+      "step": 7439
+    },
+    {
+      "epoch": 0.11904,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.880806451612905e-05,
+      "loss": 0.1803,
+      "step": 7440
+    },
+    {
+      "epoch": 0.119056,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.880645161290323e-05,
+      "loss": 0.156,
+      "step": 7441
+    },
+    {
+      "epoch": 0.119072,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.880483870967742e-05,
+      "loss": 0.2115,
+      "step": 7442
+    },
+    {
+      "epoch": 0.119088,
+      "grad_norm": 1.2578125,
+      "learning_rate": 8.880322580645162e-05,
+      "loss": 0.178,
+      "step": 7443
+    },
+    {
+      "epoch": 0.119104,
+      "grad_norm": 2.375,
+      "learning_rate": 8.88016129032258e-05,
+      "loss": 0.2285,
+      "step": 7444
+    },
+    {
+      "epoch": 0.11912,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.88e-05,
+      "loss": 0.1892,
+      "step": 7445
+    },
+    {
+      "epoch": 0.119136,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.879838709677419e-05,
+      "loss": 0.1802,
+      "step": 7446
+    },
+    {
+      "epoch": 0.119152,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.879677419354839e-05,
+      "loss": 0.1997,
+      "step": 7447
+    },
+    {
+      "epoch": 0.119168,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.879516129032258e-05,
+      "loss": 0.2127,
+      "step": 7448
+    },
+    {
+      "epoch": 0.119184,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.879354838709678e-05,
+      "loss": 0.1857,
+      "step": 7449
+    },
+    {
+      "epoch": 0.1192,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.879193548387097e-05,
+      "loss": 0.1967,
+      "step": 7450
+    },
+    {
+      "epoch": 0.119216,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.879032258064517e-05,
+      "loss": 0.1524,
+      "step": 7451
+    },
+    {
+      "epoch": 0.119232,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.878870967741936e-05,
+      "loss": 0.1818,
+      "step": 7452
+    },
+    {
+      "epoch": 0.119248,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.878709677419356e-05,
+      "loss": 0.18,
+      "step": 7453
+    },
+    {
+      "epoch": 0.119264,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.878548387096775e-05,
+      "loss": 0.1606,
+      "step": 7454
+    },
+    {
+      "epoch": 0.11928,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.878387096774195e-05,
+      "loss": 0.1889,
+      "step": 7455
+    },
+    {
+      "epoch": 0.119296,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.878225806451613e-05,
+      "loss": 0.1917,
+      "step": 7456
+    },
+    {
+      "epoch": 0.119312,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.878064516129033e-05,
+      "loss": 0.1641,
+      "step": 7457
+    },
+    {
+      "epoch": 0.119328,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.877903225806452e-05,
+      "loss": 0.1797,
+      "step": 7458
+    },
+    {
+      "epoch": 0.119344,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.87774193548387e-05,
+      "loss": 0.1887,
+      "step": 7459
+    },
+    {
+      "epoch": 0.11936,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.87758064516129e-05,
+      "loss": 0.1858,
+      "step": 7460
+    },
+    {
+      "epoch": 0.119376,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.87741935483871e-05,
+      "loss": 0.2125,
+      "step": 7461
+    },
+    {
+      "epoch": 0.119392,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.87725806451613e-05,
+      "loss": 0.2162,
+      "step": 7462
+    },
+    {
+      "epoch": 0.119408,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.877096774193549e-05,
+      "loss": 0.2304,
+      "step": 7463
+    },
+    {
+      "epoch": 0.119424,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.876935483870969e-05,
+      "loss": 0.1582,
+      "step": 7464
+    },
+    {
+      "epoch": 0.11944,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.876774193548387e-05,
+      "loss": 0.1693,
+      "step": 7465
+    },
+    {
+      "epoch": 0.119456,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.876612903225807e-05,
+      "loss": 0.1394,
+      "step": 7466
+    },
+    {
+      "epoch": 0.119472,
+      "grad_norm": 1.5390625,
+      "learning_rate": 8.876451612903226e-05,
+      "loss": 0.1747,
+      "step": 7467
+    },
+    {
+      "epoch": 0.119488,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.876290322580646e-05,
+      "loss": 0.1795,
+      "step": 7468
+    },
+    {
+      "epoch": 0.119504,
+      "grad_norm": 1.296875,
+      "learning_rate": 8.876129032258065e-05,
+      "loss": 0.1988,
+      "step": 7469
+    },
+    {
+      "epoch": 0.11952,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.875967741935485e-05,
+      "loss": 0.1554,
+      "step": 7470
+    },
+    {
+      "epoch": 0.119536,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.875806451612903e-05,
+      "loss": 0.1956,
+      "step": 7471
+    },
+    {
+      "epoch": 0.119552,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.875645161290323e-05,
+      "loss": 0.1896,
+      "step": 7472
+    },
+    {
+      "epoch": 0.119568,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.875483870967742e-05,
+      "loss": 0.2071,
+      "step": 7473
+    },
+    {
+      "epoch": 0.119584,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.875322580645162e-05,
+      "loss": 0.1654,
+      "step": 7474
+    },
+    {
+      "epoch": 0.1196,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.875161290322582e-05,
+      "loss": 0.1723,
+      "step": 7475
+    },
+    {
+      "epoch": 0.119616,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.875e-05,
+      "loss": 0.1507,
+      "step": 7476
+    },
+    {
+      "epoch": 0.119632,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.87483870967742e-05,
+      "loss": 0.1909,
+      "step": 7477
+    },
+    {
+      "epoch": 0.119648,
+      "grad_norm": 0.95703125,
+      "learning_rate": 8.874677419354839e-05,
+      "loss": 0.2004,
+      "step": 7478
+    },
+    {
+      "epoch": 0.119664,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.874516129032259e-05,
+      "loss": 0.1677,
+      "step": 7479
+    },
+    {
+      "epoch": 0.11968,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.874354838709677e-05,
+      "loss": 0.1557,
+      "step": 7480
+    },
+    {
+      "epoch": 0.119696,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.874193548387097e-05,
+      "loss": 0.1523,
+      "step": 7481
+    },
+    {
+      "epoch": 0.119712,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.874032258064516e-05,
+      "loss": 0.1345,
+      "step": 7482
+    },
+    {
+      "epoch": 0.119728,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.873870967741936e-05,
+      "loss": 0.1641,
+      "step": 7483
+    },
+    {
+      "epoch": 0.119744,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.873709677419355e-05,
+      "loss": 0.1671,
+      "step": 7484
+    },
+    {
+      "epoch": 0.11976,
+      "grad_norm": 1.3203125,
+      "learning_rate": 8.873548387096774e-05,
+      "loss": 0.1982,
+      "step": 7485
+    },
+    {
+      "epoch": 0.119776,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.873387096774194e-05,
+      "loss": 0.1587,
+      "step": 7486
+    },
+    {
+      "epoch": 0.119792,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.873225806451614e-05,
+      "loss": 0.1834,
+      "step": 7487
+    },
+    {
+      "epoch": 0.119808,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.873064516129033e-05,
+      "loss": 0.1587,
+      "step": 7488
+    },
+    {
+      "epoch": 0.119824,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.872903225806452e-05,
+      "loss": 0.128,
+      "step": 7489
+    },
+    {
+      "epoch": 0.11984,
+      "grad_norm": 1.0,
+      "learning_rate": 8.872741935483872e-05,
+      "loss": 0.172,
+      "step": 7490
+    },
+    {
+      "epoch": 0.119856,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.87258064516129e-05,
+      "loss": 0.176,
+      "step": 7491
+    },
+    {
+      "epoch": 0.119872,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.87241935483871e-05,
+      "loss": 0.1941,
+      "step": 7492
+    },
+    {
+      "epoch": 0.119888,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.872258064516129e-05,
+      "loss": 0.1894,
+      "step": 7493
+    },
+    {
+      "epoch": 0.119904,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.872096774193549e-05,
+      "loss": 0.1423,
+      "step": 7494
+    },
+    {
+      "epoch": 0.11992,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.871935483870967e-05,
+      "loss": 0.217,
+      "step": 7495
+    },
+    {
+      "epoch": 0.119936,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.871774193548387e-05,
+      "loss": 0.1736,
+      "step": 7496
+    },
+    {
+      "epoch": 0.119952,
+      "grad_norm": 1.28125,
+      "learning_rate": 8.871612903225807e-05,
+      "loss": 0.199,
+      "step": 7497
+    },
+    {
+      "epoch": 0.119968,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.871451612903227e-05,
+      "loss": 0.2429,
+      "step": 7498
+    },
+    {
+      "epoch": 0.119984,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.871290322580646e-05,
+      "loss": 0.1867,
+      "step": 7499
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.871129032258066e-05,
+      "loss": 0.1944,
+      "step": 7500
+    },
+    {
+      "epoch": 0.120016,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.870967741935484e-05,
+      "loss": 0.1563,
+      "step": 7501
+    },
+    {
+      "epoch": 0.120032,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.870806451612904e-05,
+      "loss": 0.1856,
+      "step": 7502
+    },
+    {
+      "epoch": 0.120048,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.870645161290323e-05,
+      "loss": 0.199,
+      "step": 7503
+    },
+    {
+      "epoch": 0.120064,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.870483870967743e-05,
+      "loss": 0.2056,
+      "step": 7504
+    },
+    {
+      "epoch": 0.12008,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.870322580645162e-05,
+      "loss": 0.2003,
+      "step": 7505
+    },
+    {
+      "epoch": 0.120096,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.87016129032258e-05,
+      "loss": 0.1834,
+      "step": 7506
+    },
+    {
+      "epoch": 0.120112,
+      "grad_norm": 1.234375,
+      "learning_rate": 8.87e-05,
+      "loss": 0.1567,
+      "step": 7507
+    },
+    {
+      "epoch": 0.120128,
+      "grad_norm": 0.75,
+      "learning_rate": 8.869838709677419e-05,
+      "loss": 0.1194,
+      "step": 7508
+    },
+    {
+      "epoch": 0.120144,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.869677419354839e-05,
+      "loss": 0.1991,
+      "step": 7509
+    },
+    {
+      "epoch": 0.12016,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.869516129032259e-05,
+      "loss": 0.2093,
+      "step": 7510
+    },
+    {
+      "epoch": 0.120176,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.869354838709679e-05,
+      "loss": 0.1771,
+      "step": 7511
+    },
+    {
+      "epoch": 0.120192,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.869193548387097e-05,
+      "loss": 0.1411,
+      "step": 7512
+    },
+    {
+      "epoch": 0.120208,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.869032258064517e-05,
+      "loss": 0.1833,
+      "step": 7513
+    },
+    {
+      "epoch": 0.120224,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.868870967741936e-05,
+      "loss": 0.1597,
+      "step": 7514
+    },
+    {
+      "epoch": 0.12024,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.868709677419356e-05,
+      "loss": 0.1824,
+      "step": 7515
+    },
+    {
+      "epoch": 0.120256,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.868548387096774e-05,
+      "loss": 0.1795,
+      "step": 7516
+    },
+    {
+      "epoch": 0.120272,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.868387096774194e-05,
+      "loss": 0.1771,
+      "step": 7517
+    },
+    {
+      "epoch": 0.120288,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.868225806451613e-05,
+      "loss": 0.184,
+      "step": 7518
+    },
+    {
+      "epoch": 0.120304,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.868064516129033e-05,
+      "loss": 0.183,
+      "step": 7519
+    },
+    {
+      "epoch": 0.12032,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.867903225806452e-05,
+      "loss": 0.1749,
+      "step": 7520
+    },
+    {
+      "epoch": 0.120336,
+      "grad_norm": 1.3828125,
+      "learning_rate": 8.867741935483871e-05,
+      "loss": 0.2178,
+      "step": 7521
+    },
+    {
+      "epoch": 0.120352,
+      "grad_norm": 1.3125,
+      "learning_rate": 8.867580645161291e-05,
+      "loss": 0.1802,
+      "step": 7522
+    },
+    {
+      "epoch": 0.120368,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.86741935483871e-05,
+      "loss": 0.1771,
+      "step": 7523
+    },
+    {
+      "epoch": 0.120384,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.86725806451613e-05,
+      "loss": 0.1987,
+      "step": 7524
+    },
+    {
+      "epoch": 0.1204,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.867096774193549e-05,
+      "loss": 0.1573,
+      "step": 7525
+    },
+    {
+      "epoch": 0.120416,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.866935483870969e-05,
+      "loss": 0.1765,
+      "step": 7526
+    },
+    {
+      "epoch": 0.120432,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.866774193548387e-05,
+      "loss": 0.204,
+      "step": 7527
+    },
+    {
+      "epoch": 0.120448,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.866612903225807e-05,
+      "loss": 0.1609,
+      "step": 7528
+    },
+    {
+      "epoch": 0.120464,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.866451612903226e-05,
+      "loss": 0.1935,
+      "step": 7529
+    },
+    {
+      "epoch": 0.12048,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.866290322580646e-05,
+      "loss": 0.166,
+      "step": 7530
+    },
+    {
+      "epoch": 0.120496,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.866129032258064e-05,
+      "loss": 0.1607,
+      "step": 7531
+    },
+    {
+      "epoch": 0.120512,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.865967741935484e-05,
+      "loss": 0.181,
+      "step": 7532
+    },
+    {
+      "epoch": 0.120528,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.865806451612904e-05,
+      "loss": 0.1762,
+      "step": 7533
+    },
+    {
+      "epoch": 0.120544,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.865645161290323e-05,
+      "loss": 0.1509,
+      "step": 7534
+    },
+    {
+      "epoch": 0.12056,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.865483870967743e-05,
+      "loss": 0.2192,
+      "step": 7535
+    },
+    {
+      "epoch": 0.120576,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.865322580645161e-05,
+      "loss": 0.2181,
+      "step": 7536
+    },
+    {
+      "epoch": 0.120592,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.865161290322581e-05,
+      "loss": 0.167,
+      "step": 7537
+    },
+    {
+      "epoch": 0.120608,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.865e-05,
+      "loss": 0.149,
+      "step": 7538
+    },
+    {
+      "epoch": 0.120624,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.86483870967742e-05,
+      "loss": 0.1426,
+      "step": 7539
+    },
+    {
+      "epoch": 0.12064,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.864677419354839e-05,
+      "loss": 0.1724,
+      "step": 7540
+    },
+    {
+      "epoch": 0.120656,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.864516129032259e-05,
+      "loss": 0.184,
+      "step": 7541
+    },
+    {
+      "epoch": 0.120672,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.864354838709677e-05,
+      "loss": 0.1831,
+      "step": 7542
+    },
+    {
+      "epoch": 0.120688,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.864193548387097e-05,
+      "loss": 0.2306,
+      "step": 7543
+    },
+    {
+      "epoch": 0.120704,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.864032258064516e-05,
+      "loss": 0.1852,
+      "step": 7544
+    },
+    {
+      "epoch": 0.12072,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.863870967741936e-05,
+      "loss": 0.1613,
+      "step": 7545
+    },
+    {
+      "epoch": 0.120736,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.863709677419356e-05,
+      "loss": 0.2108,
+      "step": 7546
+    },
+    {
+      "epoch": 0.120752,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.863548387096776e-05,
+      "loss": 0.1623,
+      "step": 7547
+    },
+    {
+      "epoch": 0.120768,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.863387096774194e-05,
+      "loss": 0.1566,
+      "step": 7548
+    },
+    {
+      "epoch": 0.120784,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.863225806451614e-05,
+      "loss": 0.187,
+      "step": 7549
+    },
+    {
+      "epoch": 0.1208,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.863064516129033e-05,
+      "loss": 0.1672,
+      "step": 7550
+    },
+    {
+      "epoch": 0.120816,
+      "grad_norm": 0.95703125,
+      "learning_rate": 8.862903225806451e-05,
+      "loss": 0.1946,
+      "step": 7551
+    },
+    {
+      "epoch": 0.120832,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.862741935483871e-05,
+      "loss": 0.1613,
+      "step": 7552
+    },
+    {
+      "epoch": 0.120848,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.86258064516129e-05,
+      "loss": 0.1978,
+      "step": 7553
+    },
+    {
+      "epoch": 0.120864,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.86241935483871e-05,
+      "loss": 0.1844,
+      "step": 7554
+    },
+    {
+      "epoch": 0.12088,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.862258064516129e-05,
+      "loss": 0.1468,
+      "step": 7555
+    },
+    {
+      "epoch": 0.120896,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.862096774193548e-05,
+      "loss": 0.155,
+      "step": 7556
+    },
+    {
+      "epoch": 0.120912,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.861935483870968e-05,
+      "loss": 0.1135,
+      "step": 7557
+    },
+    {
+      "epoch": 0.120928,
+      "grad_norm": 0.486328125,
+      "learning_rate": 8.861774193548388e-05,
+      "loss": 0.1528,
+      "step": 7558
+    },
+    {
+      "epoch": 0.120944,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.861612903225807e-05,
+      "loss": 0.1491,
+      "step": 7559
+    },
+    {
+      "epoch": 0.12096,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.861451612903227e-05,
+      "loss": 0.1813,
+      "step": 7560
+    },
+    {
+      "epoch": 0.120976,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.861290322580646e-05,
+      "loss": 0.1674,
+      "step": 7561
+    },
+    {
+      "epoch": 0.120992,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.861129032258066e-05,
+      "loss": 0.1939,
+      "step": 7562
+    },
+    {
+      "epoch": 0.121008,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.860967741935484e-05,
+      "loss": 0.1998,
+      "step": 7563
+    },
+    {
+      "epoch": 0.121024,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.860806451612904e-05,
+      "loss": 0.1677,
+      "step": 7564
+    },
+    {
+      "epoch": 0.12104,
+      "grad_norm": 0.75,
+      "learning_rate": 8.860645161290323e-05,
+      "loss": 0.1843,
+      "step": 7565
+    },
+    {
+      "epoch": 0.121056,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.860483870967743e-05,
+      "loss": 0.1998,
+      "step": 7566
+    },
+    {
+      "epoch": 0.121072,
+      "grad_norm": 1.125,
+      "learning_rate": 8.860322580645161e-05,
+      "loss": 0.1428,
+      "step": 7567
+    },
+    {
+      "epoch": 0.121088,
+      "grad_norm": 1.296875,
+      "learning_rate": 8.86016129032258e-05,
+      "loss": 0.1878,
+      "step": 7568
+    },
+    {
+      "epoch": 0.121104,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.86e-05,
+      "loss": 0.1879,
+      "step": 7569
+    },
+    {
+      "epoch": 0.12112,
+      "grad_norm": 1.296875,
+      "learning_rate": 8.85983870967742e-05,
+      "loss": 0.1909,
+      "step": 7570
+    },
+    {
+      "epoch": 0.121136,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.85967741935484e-05,
+      "loss": 0.181,
+      "step": 7571
+    },
+    {
+      "epoch": 0.121152,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.859516129032258e-05,
+      "loss": 0.1358,
+      "step": 7572
+    },
+    {
+      "epoch": 0.121168,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.859354838709678e-05,
+      "loss": 0.1799,
+      "step": 7573
+    },
+    {
+      "epoch": 0.121184,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.859193548387097e-05,
+      "loss": 0.1713,
+      "step": 7574
+    },
+    {
+      "epoch": 0.1212,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.859032258064517e-05,
+      "loss": 0.1677,
+      "step": 7575
+    },
+    {
+      "epoch": 0.121216,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.858870967741936e-05,
+      "loss": 0.1853,
+      "step": 7576
+    },
+    {
+      "epoch": 0.121232,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.858709677419356e-05,
+      "loss": 0.1536,
+      "step": 7577
+    },
+    {
+      "epoch": 0.121248,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.858548387096774e-05,
+      "loss": 0.137,
+      "step": 7578
+    },
+    {
+      "epoch": 0.121264,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.858387096774194e-05,
+      "loss": 0.1635,
+      "step": 7579
+    },
+    {
+      "epoch": 0.12128,
+      "grad_norm": 0.75,
+      "learning_rate": 8.858225806451613e-05,
+      "loss": 0.1682,
+      "step": 7580
+    },
+    {
+      "epoch": 0.121296,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.858064516129033e-05,
+      "loss": 0.1439,
+      "step": 7581
+    },
+    {
+      "epoch": 0.121312,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.857903225806453e-05,
+      "loss": 0.1872,
+      "step": 7582
+    },
+    {
+      "epoch": 0.121328,
+      "grad_norm": 1.2578125,
+      "learning_rate": 8.857741935483871e-05,
+      "loss": 0.1875,
+      "step": 7583
+    },
+    {
+      "epoch": 0.121344,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.857580645161291e-05,
+      "loss": 0.1744,
+      "step": 7584
+    },
+    {
+      "epoch": 0.12136,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.85741935483871e-05,
+      "loss": 0.1662,
+      "step": 7585
+    },
+    {
+      "epoch": 0.121376,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.85725806451613e-05,
+      "loss": 0.1983,
+      "step": 7586
+    },
+    {
+      "epoch": 0.121392,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.857096774193548e-05,
+      "loss": 0.1958,
+      "step": 7587
+    },
+    {
+      "epoch": 0.121408,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.856935483870968e-05,
+      "loss": 0.1584,
+      "step": 7588
+    },
+    {
+      "epoch": 0.121424,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.856774193548387e-05,
+      "loss": 0.1674,
+      "step": 7589
+    },
+    {
+      "epoch": 0.12144,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.856612903225807e-05,
+      "loss": 0.1674,
+      "step": 7590
+    },
+    {
+      "epoch": 0.121456,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.856451612903226e-05,
+      "loss": 0.1642,
+      "step": 7591
+    },
+    {
+      "epoch": 0.121472,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.856290322580645e-05,
+      "loss": 0.1848,
+      "step": 7592
+    },
+    {
+      "epoch": 0.121488,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.856129032258065e-05,
+      "loss": 0.1826,
+      "step": 7593
+    },
+    {
+      "epoch": 0.121504,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.855967741935485e-05,
+      "loss": 0.2438,
+      "step": 7594
+    },
+    {
+      "epoch": 0.12152,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.855806451612904e-05,
+      "loss": 0.1416,
+      "step": 7595
+    },
+    {
+      "epoch": 0.121536,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.855645161290324e-05,
+      "loss": 0.2006,
+      "step": 7596
+    },
+    {
+      "epoch": 0.121552,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.855483870967743e-05,
+      "loss": 0.1749,
+      "step": 7597
+    },
+    {
+      "epoch": 0.121568,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.855322580645161e-05,
+      "loss": 0.1815,
+      "step": 7598
+    },
+    {
+      "epoch": 0.121584,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.855161290322581e-05,
+      "loss": 0.1884,
+      "step": 7599
+    },
+    {
+      "epoch": 0.1216,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.855e-05,
+      "loss": 0.1853,
+      "step": 7600
+    },
+    {
+      "epoch": 0.121616,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.85483870967742e-05,
+      "loss": 0.1902,
+      "step": 7601
+    },
+    {
+      "epoch": 0.121632,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.854677419354838e-05,
+      "loss": 0.1869,
+      "step": 7602
+    },
+    {
+      "epoch": 0.121648,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.854516129032258e-05,
+      "loss": 0.1808,
+      "step": 7603
+    },
+    {
+      "epoch": 0.121664,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.854354838709677e-05,
+      "loss": 0.1976,
+      "step": 7604
+    },
+    {
+      "epoch": 0.12168,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.854193548387097e-05,
+      "loss": 0.2293,
+      "step": 7605
+    },
+    {
+      "epoch": 0.121696,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.854032258064517e-05,
+      "loss": 0.2007,
+      "step": 7606
+    },
+    {
+      "epoch": 0.121712,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.853870967741937e-05,
+      "loss": 0.1468,
+      "step": 7607
+    },
+    {
+      "epoch": 0.121728,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.853709677419355e-05,
+      "loss": 0.1802,
+      "step": 7608
+    },
+    {
+      "epoch": 0.121744,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.853548387096775e-05,
+      "loss": 0.2131,
+      "step": 7609
+    },
+    {
+      "epoch": 0.12176,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.853387096774194e-05,
+      "loss": 0.1835,
+      "step": 7610
+    },
+    {
+      "epoch": 0.121776,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.853225806451614e-05,
+      "loss": 0.1813,
+      "step": 7611
+    },
+    {
+      "epoch": 0.121792,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.853064516129033e-05,
+      "loss": 0.1455,
+      "step": 7612
+    },
+    {
+      "epoch": 0.121808,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.852903225806451e-05,
+      "loss": 0.1749,
+      "step": 7613
+    },
+    {
+      "epoch": 0.121824,
+      "grad_norm": 1.4296875,
+      "learning_rate": 8.852741935483871e-05,
+      "loss": 0.1822,
+      "step": 7614
+    },
+    {
+      "epoch": 0.12184,
+      "grad_norm": 1.4140625,
+      "learning_rate": 8.85258064516129e-05,
+      "loss": 0.1878,
+      "step": 7615
+    },
+    {
+      "epoch": 0.121856,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.85241935483871e-05,
+      "loss": 0.1651,
+      "step": 7616
+    },
+    {
+      "epoch": 0.121872,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.85225806451613e-05,
+      "loss": 0.1794,
+      "step": 7617
+    },
+    {
+      "epoch": 0.121888,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.85209677419355e-05,
+      "loss": 0.1636,
+      "step": 7618
+    },
+    {
+      "epoch": 0.121904,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.851935483870968e-05,
+      "loss": 0.1763,
+      "step": 7619
+    },
+    {
+      "epoch": 0.12192,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.851774193548388e-05,
+      "loss": 0.1812,
+      "step": 7620
+    },
+    {
+      "epoch": 0.121936,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.851612903225807e-05,
+      "loss": 0.1804,
+      "step": 7621
+    },
+    {
+      "epoch": 0.121952,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.851451612903227e-05,
+      "loss": 0.1489,
+      "step": 7622
+    },
+    {
+      "epoch": 0.121968,
+      "grad_norm": 1.265625,
+      "learning_rate": 8.851290322580645e-05,
+      "loss": 0.1561,
+      "step": 7623
+    },
+    {
+      "epoch": 0.121984,
+      "grad_norm": 1.359375,
+      "learning_rate": 8.851129032258065e-05,
+      "loss": 0.1959,
+      "step": 7624
+    },
+    {
+      "epoch": 0.122,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.850967741935484e-05,
+      "loss": 0.1817,
+      "step": 7625
+    },
+    {
+      "epoch": 0.122016,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.850806451612904e-05,
+      "loss": 0.1703,
+      "step": 7626
+    },
+    {
+      "epoch": 0.122032,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.850645161290322e-05,
+      "loss": 0.2248,
+      "step": 7627
+    },
+    {
+      "epoch": 0.122048,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.850483870967742e-05,
+      "loss": 0.1721,
+      "step": 7628
+    },
+    {
+      "epoch": 0.122064,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.850322580645161e-05,
+      "loss": 0.1563,
+      "step": 7629
+    },
+    {
+      "epoch": 0.12208,
+      "grad_norm": 0.75,
+      "learning_rate": 8.850161290322581e-05,
+      "loss": 0.1895,
+      "step": 7630
+    },
+    {
+      "epoch": 0.122096,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.850000000000001e-05,
+      "loss": 0.1926,
+      "step": 7631
+    },
+    {
+      "epoch": 0.122112,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.84983870967742e-05,
+      "loss": 0.1743,
+      "step": 7632
+    },
+    {
+      "epoch": 0.122128,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.84967741935484e-05,
+      "loss": 0.1671,
+      "step": 7633
+    },
+    {
+      "epoch": 0.122144,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.849516129032258e-05,
+      "loss": 0.1625,
+      "step": 7634
+    },
+    {
+      "epoch": 0.12216,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.849354838709678e-05,
+      "loss": 0.1373,
+      "step": 7635
+    },
+    {
+      "epoch": 0.122176,
+      "grad_norm": 0.95703125,
+      "learning_rate": 8.849193548387097e-05,
+      "loss": 0.1667,
+      "step": 7636
+    },
+    {
+      "epoch": 0.122192,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.849032258064517e-05,
+      "loss": 0.1846,
+      "step": 7637
+    },
+    {
+      "epoch": 0.122208,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.848870967741935e-05,
+      "loss": 0.1751,
+      "step": 7638
+    },
+    {
+      "epoch": 0.122224,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.848709677419355e-05,
+      "loss": 0.1704,
+      "step": 7639
+    },
+    {
+      "epoch": 0.12224,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.848548387096774e-05,
+      "loss": 0.1939,
+      "step": 7640
+    },
+    {
+      "epoch": 0.122256,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.848387096774194e-05,
+      "loss": 0.172,
+      "step": 7641
+    },
+    {
+      "epoch": 0.122272,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.848225806451614e-05,
+      "loss": 0.1812,
+      "step": 7642
+    },
+    {
+      "epoch": 0.122288,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.848064516129034e-05,
+      "loss": 0.1655,
+      "step": 7643
+    },
+    {
+      "epoch": 0.122304,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.847903225806452e-05,
+      "loss": 0.1648,
+      "step": 7644
+    },
+    {
+      "epoch": 0.12232,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.847741935483871e-05,
+      "loss": 0.2066,
+      "step": 7645
+    },
+    {
+      "epoch": 0.122336,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.847580645161291e-05,
+      "loss": 0.1619,
+      "step": 7646
+    },
+    {
+      "epoch": 0.122352,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.84741935483871e-05,
+      "loss": 0.2023,
+      "step": 7647
+    },
+    {
+      "epoch": 0.122368,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.84725806451613e-05,
+      "loss": 0.135,
+      "step": 7648
+    },
+    {
+      "epoch": 0.122384,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.847096774193548e-05,
+      "loss": 0.1773,
+      "step": 7649
+    },
+    {
+      "epoch": 0.1224,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.846935483870968e-05,
+      "loss": 0.1853,
+      "step": 7650
+    },
+    {
+      "epoch": 0.122416,
+      "grad_norm": 2.65625,
+      "learning_rate": 8.846774193548387e-05,
+      "loss": 0.2015,
+      "step": 7651
+    },
+    {
+      "epoch": 0.122432,
+      "grad_norm": 1.859375,
+      "learning_rate": 8.846612903225807e-05,
+      "loss": 0.1422,
+      "step": 7652
+    },
+    {
+      "epoch": 0.122448,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.846451612903227e-05,
+      "loss": 0.2599,
+      "step": 7653
+    },
+    {
+      "epoch": 0.122464,
+      "grad_norm": 1.296875,
+      "learning_rate": 8.846290322580647e-05,
+      "loss": 0.1771,
+      "step": 7654
+    },
+    {
+      "epoch": 0.12248,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.846129032258065e-05,
+      "loss": 0.1681,
+      "step": 7655
+    },
+    {
+      "epoch": 0.122496,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.845967741935485e-05,
+      "loss": 0.1218,
+      "step": 7656
+    },
+    {
+      "epoch": 0.122512,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.845806451612904e-05,
+      "loss": 0.1878,
+      "step": 7657
+    },
+    {
+      "epoch": 0.122528,
+      "grad_norm": 1.4453125,
+      "learning_rate": 8.845645161290324e-05,
+      "loss": 0.2088,
+      "step": 7658
+    },
+    {
+      "epoch": 0.122544,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.845483870967742e-05,
+      "loss": 0.1788,
+      "step": 7659
+    },
+    {
+      "epoch": 0.12256,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.845322580645161e-05,
+      "loss": 0.1806,
+      "step": 7660
+    },
+    {
+      "epoch": 0.122576,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.845161290322581e-05,
+      "loss": 0.1779,
+      "step": 7661
+    },
+    {
+      "epoch": 0.122592,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.845e-05,
+      "loss": 0.1801,
+      "step": 7662
+    },
+    {
+      "epoch": 0.122608,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.84483870967742e-05,
+      "loss": 0.1998,
+      "step": 7663
+    },
+    {
+      "epoch": 0.122624,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.844677419354838e-05,
+      "loss": 0.1684,
+      "step": 7664
+    },
+    {
+      "epoch": 0.12264,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.844516129032258e-05,
+      "loss": 0.1336,
+      "step": 7665
+    },
+    {
+      "epoch": 0.122656,
+      "grad_norm": 1.0,
+      "learning_rate": 8.844354838709678e-05,
+      "loss": 0.1611,
+      "step": 7666
+    },
+    {
+      "epoch": 0.122672,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.844193548387098e-05,
+      "loss": 0.2143,
+      "step": 7667
+    },
+    {
+      "epoch": 0.122688,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.844032258064517e-05,
+      "loss": 0.1586,
+      "step": 7668
+    },
+    {
+      "epoch": 0.122704,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.843870967741937e-05,
+      "loss": 0.1569,
+      "step": 7669
+    },
+    {
+      "epoch": 0.12272,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.843709677419355e-05,
+      "loss": 0.2018,
+      "step": 7670
+    },
+    {
+      "epoch": 0.122736,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.843548387096775e-05,
+      "loss": 0.1892,
+      "step": 7671
+    },
+    {
+      "epoch": 0.122752,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.843387096774194e-05,
+      "loss": 0.1683,
+      "step": 7672
+    },
+    {
+      "epoch": 0.122768,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.843225806451614e-05,
+      "loss": 0.1872,
+      "step": 7673
+    },
+    {
+      "epoch": 0.122784,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.843064516129032e-05,
+      "loss": 0.1452,
+      "step": 7674
+    },
+    {
+      "epoch": 0.1228,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.842903225806452e-05,
+      "loss": 0.1833,
+      "step": 7675
+    },
+    {
+      "epoch": 0.122816,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.842741935483871e-05,
+      "loss": 0.1419,
+      "step": 7676
+    },
+    {
+      "epoch": 0.122832,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.842580645161291e-05,
+      "loss": 0.209,
+      "step": 7677
+    },
+    {
+      "epoch": 0.122848,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.842419354838711e-05,
+      "loss": 0.1726,
+      "step": 7678
+    },
+    {
+      "epoch": 0.122864,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.84225806451613e-05,
+      "loss": 0.2035,
+      "step": 7679
+    },
+    {
+      "epoch": 0.12288,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.84209677419355e-05,
+      "loss": 0.1818,
+      "step": 7680
+    },
+    {
+      "epoch": 0.122896,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.841935483870968e-05,
+      "loss": 0.1602,
+      "step": 7681
+    },
+    {
+      "epoch": 0.122912,
+      "grad_norm": 1.2109375,
+      "learning_rate": 8.841774193548388e-05,
+      "loss": 0.1712,
+      "step": 7682
+    },
+    {
+      "epoch": 0.122928,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.841612903225807e-05,
+      "loss": 0.1892,
+      "step": 7683
+    },
+    {
+      "epoch": 0.122944,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.841451612903226e-05,
+      "loss": 0.1542,
+      "step": 7684
+    },
+    {
+      "epoch": 0.12296,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.841290322580645e-05,
+      "loss": 0.1503,
+      "step": 7685
+    },
+    {
+      "epoch": 0.122976,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.841129032258065e-05,
+      "loss": 0.1737,
+      "step": 7686
+    },
+    {
+      "epoch": 0.122992,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.840967741935484e-05,
+      "loss": 0.2088,
+      "step": 7687
+    },
+    {
+      "epoch": 0.123008,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.840806451612904e-05,
+      "loss": 0.1793,
+      "step": 7688
+    },
+    {
+      "epoch": 0.123024,
+      "grad_norm": 0.625,
+      "learning_rate": 8.840645161290324e-05,
+      "loss": 0.1656,
+      "step": 7689
+    },
+    {
+      "epoch": 0.12304,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.840483870967742e-05,
+      "loss": 0.152,
+      "step": 7690
+    },
+    {
+      "epoch": 0.123056,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.840322580645162e-05,
+      "loss": 0.1951,
+      "step": 7691
+    },
+    {
+      "epoch": 0.123072,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.840161290322581e-05,
+      "loss": 0.2178,
+      "step": 7692
+    },
+    {
+      "epoch": 0.123088,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.840000000000001e-05,
+      "loss": 0.1813,
+      "step": 7693
+    },
+    {
+      "epoch": 0.123104,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.83983870967742e-05,
+      "loss": 0.1718,
+      "step": 7694
+    },
+    {
+      "epoch": 0.12312,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.839677419354839e-05,
+      "loss": 0.1321,
+      "step": 7695
+    },
+    {
+      "epoch": 0.123136,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.839516129032258e-05,
+      "loss": 0.1566,
+      "step": 7696
+    },
+    {
+      "epoch": 0.123152,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.839354838709678e-05,
+      "loss": 0.1526,
+      "step": 7697
+    },
+    {
+      "epoch": 0.123168,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.839193548387096e-05,
+      "loss": 0.1803,
+      "step": 7698
+    },
+    {
+      "epoch": 0.123184,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.839032258064516e-05,
+      "loss": 0.1521,
+      "step": 7699
+    },
+    {
+      "epoch": 0.1232,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.838870967741935e-05,
+      "loss": 0.1972,
+      "step": 7700
+    },
+    {
+      "epoch": 0.123216,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.838709677419355e-05,
+      "loss": 0.1502,
+      "step": 7701
+    },
+    {
+      "epoch": 0.123232,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.838548387096775e-05,
+      "loss": 0.1942,
+      "step": 7702
+    },
+    {
+      "epoch": 0.123248,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.838387096774195e-05,
+      "loss": 0.1892,
+      "step": 7703
+    },
+    {
+      "epoch": 0.123264,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.838225806451614e-05,
+      "loss": 0.2077,
+      "step": 7704
+    },
+    {
+      "epoch": 0.12328,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.838064516129034e-05,
+      "loss": 0.1545,
+      "step": 7705
+    },
+    {
+      "epoch": 0.123296,
+      "grad_norm": 1.234375,
+      "learning_rate": 8.837903225806452e-05,
+      "loss": 0.2078,
+      "step": 7706
+    },
+    {
+      "epoch": 0.123312,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.837741935483871e-05,
+      "loss": 0.1847,
+      "step": 7707
+    },
+    {
+      "epoch": 0.123328,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.837580645161291e-05,
+      "loss": 0.1742,
+      "step": 7708
+    },
+    {
+      "epoch": 0.123344,
+      "grad_norm": 1.25,
+      "learning_rate": 8.837419354838709e-05,
+      "loss": 0.19,
+      "step": 7709
+    },
+    {
+      "epoch": 0.12336,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.837258064516129e-05,
+      "loss": 0.2347,
+      "step": 7710
+    },
+    {
+      "epoch": 0.123376,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.837096774193548e-05,
+      "loss": 0.1822,
+      "step": 7711
+    },
+    {
+      "epoch": 0.123392,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.836935483870968e-05,
+      "loss": 0.1565,
+      "step": 7712
+    },
+    {
+      "epoch": 0.123408,
+      "grad_norm": 1.859375,
+      "learning_rate": 8.836774193548388e-05,
+      "loss": 0.2332,
+      "step": 7713
+    },
+    {
+      "epoch": 0.123424,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.836612903225808e-05,
+      "loss": 0.2039,
+      "step": 7714
+    },
+    {
+      "epoch": 0.12344,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.836451612903226e-05,
+      "loss": 0.1342,
+      "step": 7715
+    },
+    {
+      "epoch": 0.123456,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.836290322580646e-05,
+      "loss": 0.2026,
+      "step": 7716
+    },
+    {
+      "epoch": 0.123472,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.836129032258065e-05,
+      "loss": 0.1744,
+      "step": 7717
+    },
+    {
+      "epoch": 0.123488,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.835967741935485e-05,
+      "loss": 0.1468,
+      "step": 7718
+    },
+    {
+      "epoch": 0.123504,
+      "grad_norm": 1.921875,
+      "learning_rate": 8.835806451612904e-05,
+      "loss": 0.1836,
+      "step": 7719
+    },
+    {
+      "epoch": 0.12352,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.835645161290323e-05,
+      "loss": 0.1452,
+      "step": 7720
+    },
+    {
+      "epoch": 0.123536,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.835483870967742e-05,
+      "loss": 0.1344,
+      "step": 7721
+    },
+    {
+      "epoch": 0.123552,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.835322580645161e-05,
+      "loss": 0.1901,
+      "step": 7722
+    },
+    {
+      "epoch": 0.123568,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.83516129032258e-05,
+      "loss": 0.161,
+      "step": 7723
+    },
+    {
+      "epoch": 0.123584,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.834999999999999e-05,
+      "loss": 0.2083,
+      "step": 7724
+    },
+    {
+      "epoch": 0.1236,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.834838709677419e-05,
+      "loss": 0.2173,
+      "step": 7725
+    },
+    {
+      "epoch": 0.123616,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.834677419354839e-05,
+      "loss": 0.1659,
+      "step": 7726
+    },
+    {
+      "epoch": 0.123632,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.834516129032259e-05,
+      "loss": 0.1804,
+      "step": 7727
+    },
+    {
+      "epoch": 0.123648,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.834354838709678e-05,
+      "loss": 0.1736,
+      "step": 7728
+    },
+    {
+      "epoch": 0.123664,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.834193548387098e-05,
+      "loss": 0.2023,
+      "step": 7729
+    },
+    {
+      "epoch": 0.12368,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.834032258064516e-05,
+      "loss": 0.1645,
+      "step": 7730
+    },
+    {
+      "epoch": 0.123696,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.833870967741936e-05,
+      "loss": 0.204,
+      "step": 7731
+    },
+    {
+      "epoch": 0.123712,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.833709677419355e-05,
+      "loss": 0.1445,
+      "step": 7732
+    },
+    {
+      "epoch": 0.123728,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.833548387096775e-05,
+      "loss": 0.1786,
+      "step": 7733
+    },
+    {
+      "epoch": 0.123744,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.833387096774193e-05,
+      "loss": 0.1868,
+      "step": 7734
+    },
+    {
+      "epoch": 0.12376,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.833225806451613e-05,
+      "loss": 0.1541,
+      "step": 7735
+    },
+    {
+      "epoch": 0.123776,
+      "grad_norm": 1.40625,
+      "learning_rate": 8.833064516129032e-05,
+      "loss": 0.2079,
+      "step": 7736
+    },
+    {
+      "epoch": 0.123792,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.832903225806452e-05,
+      "loss": 0.1944,
+      "step": 7737
+    },
+    {
+      "epoch": 0.123808,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.832741935483872e-05,
+      "loss": 0.1876,
+      "step": 7738
+    },
+    {
+      "epoch": 0.123824,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.83258064516129e-05,
+      "loss": 0.2166,
+      "step": 7739
+    },
+    {
+      "epoch": 0.12384,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.83241935483871e-05,
+      "loss": 0.1735,
+      "step": 7740
+    },
+    {
+      "epoch": 0.123856,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.832258064516129e-05,
+      "loss": 0.2157,
+      "step": 7741
+    },
+    {
+      "epoch": 0.123872,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.832096774193549e-05,
+      "loss": 0.1542,
+      "step": 7742
+    },
+    {
+      "epoch": 0.123888,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.831935483870968e-05,
+      "loss": 0.1806,
+      "step": 7743
+    },
+    {
+      "epoch": 0.123904,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.831774193548388e-05,
+      "loss": 0.179,
+      "step": 7744
+    },
+    {
+      "epoch": 0.12392,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.831612903225806e-05,
+      "loss": 0.2021,
+      "step": 7745
+    },
+    {
+      "epoch": 0.123936,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.831451612903226e-05,
+      "loss": 0.1675,
+      "step": 7746
+    },
+    {
+      "epoch": 0.123952,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.831290322580645e-05,
+      "loss": 0.1853,
+      "step": 7747
+    },
+    {
+      "epoch": 0.123968,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.831129032258065e-05,
+      "loss": 0.1874,
+      "step": 7748
+    },
+    {
+      "epoch": 0.123984,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.830967741935485e-05,
+      "loss": 0.1465,
+      "step": 7749
+    },
+    {
+      "epoch": 0.124,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.830806451612905e-05,
+      "loss": 0.1928,
+      "step": 7750
+    },
+    {
+      "epoch": 0.124016,
+      "grad_norm": 1.234375,
+      "learning_rate": 8.830645161290323e-05,
+      "loss": 0.1771,
+      "step": 7751
+    },
+    {
+      "epoch": 0.124032,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.830483870967743e-05,
+      "loss": 0.1694,
+      "step": 7752
+    },
+    {
+      "epoch": 0.124048,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.830322580645162e-05,
+      "loss": 0.1747,
+      "step": 7753
+    },
+    {
+      "epoch": 0.124064,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.83016129032258e-05,
+      "loss": 0.1949,
+      "step": 7754
+    },
+    {
+      "epoch": 0.12408,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.83e-05,
+      "loss": 0.1592,
+      "step": 7755
+    },
+    {
+      "epoch": 0.124096,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.829838709677419e-05,
+      "loss": 0.17,
+      "step": 7756
+    },
+    {
+      "epoch": 0.124112,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.829677419354839e-05,
+      "loss": 0.1852,
+      "step": 7757
+    },
+    {
+      "epoch": 0.124128,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.829516129032258e-05,
+      "loss": 0.1874,
+      "step": 7758
+    },
+    {
+      "epoch": 0.124144,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.829354838709678e-05,
+      "loss": 0.1756,
+      "step": 7759
+    },
+    {
+      "epoch": 0.12416,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.829193548387096e-05,
+      "loss": 0.1581,
+      "step": 7760
+    },
+    {
+      "epoch": 0.124176,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.829032258064516e-05,
+      "loss": 0.1875,
+      "step": 7761
+    },
+    {
+      "epoch": 0.124192,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.828870967741936e-05,
+      "loss": 0.2003,
+      "step": 7762
+    },
+    {
+      "epoch": 0.124208,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.828709677419356e-05,
+      "loss": 0.1581,
+      "step": 7763
+    },
+    {
+      "epoch": 0.124224,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.828548387096775e-05,
+      "loss": 0.1932,
+      "step": 7764
+    },
+    {
+      "epoch": 0.12424,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.828387096774195e-05,
+      "loss": 0.1796,
+      "step": 7765
+    },
+    {
+      "epoch": 0.124256,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.828225806451613e-05,
+      "loss": 0.1569,
+      "step": 7766
+    },
+    {
+      "epoch": 0.124272,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.828064516129033e-05,
+      "loss": 0.1766,
+      "step": 7767
+    },
+    {
+      "epoch": 0.124288,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.827903225806452e-05,
+      "loss": 0.1507,
+      "step": 7768
+    },
+    {
+      "epoch": 0.124304,
+      "grad_norm": 0.625,
+      "learning_rate": 8.82774193548387e-05,
+      "loss": 0.1768,
+      "step": 7769
+    },
+    {
+      "epoch": 0.12432,
+      "grad_norm": 1.2890625,
+      "learning_rate": 8.82758064516129e-05,
+      "loss": 0.1988,
+      "step": 7770
+    },
+    {
+      "epoch": 0.124336,
+      "grad_norm": 0.875,
+      "learning_rate": 8.827419354838709e-05,
+      "loss": 0.2052,
+      "step": 7771
+    },
+    {
+      "epoch": 0.124352,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.827258064516129e-05,
+      "loss": 0.1792,
+      "step": 7772
+    },
+    {
+      "epoch": 0.124368,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.827096774193549e-05,
+      "loss": 0.2047,
+      "step": 7773
+    },
+    {
+      "epoch": 0.124384,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.826935483870969e-05,
+      "loss": 0.1655,
+      "step": 7774
+    },
+    {
+      "epoch": 0.1244,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.826774193548388e-05,
+      "loss": 0.1603,
+      "step": 7775
+    },
+    {
+      "epoch": 0.124416,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.826612903225808e-05,
+      "loss": 0.1564,
+      "step": 7776
+    },
+    {
+      "epoch": 0.124432,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.826451612903226e-05,
+      "loss": 0.1859,
+      "step": 7777
+    },
+    {
+      "epoch": 0.124448,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.826290322580646e-05,
+      "loss": 0.1595,
+      "step": 7778
+    },
+    {
+      "epoch": 0.124464,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.826129032258065e-05,
+      "loss": 0.1862,
+      "step": 7779
+    },
+    {
+      "epoch": 0.12448,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.825967741935485e-05,
+      "loss": 0.1615,
+      "step": 7780
+    },
+    {
+      "epoch": 0.124496,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.825806451612903e-05,
+      "loss": 0.1543,
+      "step": 7781
+    },
+    {
+      "epoch": 0.124512,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.825645161290323e-05,
+      "loss": 0.1975,
+      "step": 7782
+    },
+    {
+      "epoch": 0.124528,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.825483870967742e-05,
+      "loss": 0.1688,
+      "step": 7783
+    },
+    {
+      "epoch": 0.124544,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.825322580645162e-05,
+      "loss": 0.1613,
+      "step": 7784
+    },
+    {
+      "epoch": 0.12456,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.825161290322582e-05,
+      "loss": 0.1424,
+      "step": 7785
+    },
+    {
+      "epoch": 0.124576,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.825e-05,
+      "loss": 0.1888,
+      "step": 7786
+    },
+    {
+      "epoch": 0.124592,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.82483870967742e-05,
+      "loss": 0.2026,
+      "step": 7787
+    },
+    {
+      "epoch": 0.124608,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.824677419354839e-05,
+      "loss": 0.1546,
+      "step": 7788
+    },
+    {
+      "epoch": 0.124624,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.824516129032259e-05,
+      "loss": 0.1847,
+      "step": 7789
+    },
+    {
+      "epoch": 0.12464,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.824354838709678e-05,
+      "loss": 0.2003,
+      "step": 7790
+    },
+    {
+      "epoch": 0.124656,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.824193548387097e-05,
+      "loss": 0.1685,
+      "step": 7791
+    },
+    {
+      "epoch": 0.124672,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.824032258064516e-05,
+      "loss": 0.1468,
+      "step": 7792
+    },
+    {
+      "epoch": 0.124688,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.823870967741936e-05,
+      "loss": 0.204,
+      "step": 7793
+    },
+    {
+      "epoch": 0.124704,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.823709677419355e-05,
+      "loss": 0.1822,
+      "step": 7794
+    },
+    {
+      "epoch": 0.12472,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.823548387096775e-05,
+      "loss": 0.2051,
+      "step": 7795
+    },
+    {
+      "epoch": 0.124736,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.823387096774193e-05,
+      "loss": 0.1667,
+      "step": 7796
+    },
+    {
+      "epoch": 0.124752,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.823225806451613e-05,
+      "loss": 0.1919,
+      "step": 7797
+    },
+    {
+      "epoch": 0.124768,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.823064516129033e-05,
+      "loss": 0.2116,
+      "step": 7798
+    },
+    {
+      "epoch": 0.124784,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.822903225806453e-05,
+      "loss": 0.1559,
+      "step": 7799
+    },
+    {
+      "epoch": 0.1248,
+      "grad_norm": 1.265625,
+      "learning_rate": 8.822741935483872e-05,
+      "loss": 0.1997,
+      "step": 7800
+    },
+    {
+      "epoch": 0.124816,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.82258064516129e-05,
+      "loss": 0.1765,
+      "step": 7801
+    },
+    {
+      "epoch": 0.124832,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.82241935483871e-05,
+      "loss": 0.2147,
+      "step": 7802
+    },
+    {
+      "epoch": 0.124848,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.822258064516129e-05,
+      "loss": 0.2003,
+      "step": 7803
+    },
+    {
+      "epoch": 0.124864,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.822096774193549e-05,
+      "loss": 0.191,
+      "step": 7804
+    },
+    {
+      "epoch": 0.12488,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.821935483870967e-05,
+      "loss": 0.1907,
+      "step": 7805
+    },
+    {
+      "epoch": 0.124896,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.821774193548387e-05,
+      "loss": 0.1289,
+      "step": 7806
+    },
+    {
+      "epoch": 0.124912,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.821612903225806e-05,
+      "loss": 0.1807,
+      "step": 7807
+    },
+    {
+      "epoch": 0.124928,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.821451612903226e-05,
+      "loss": 0.1431,
+      "step": 7808
+    },
+    {
+      "epoch": 0.124944,
+      "grad_norm": 1.5078125,
+      "learning_rate": 8.821290322580646e-05,
+      "loss": 0.2092,
+      "step": 7809
+    },
+    {
+      "epoch": 0.12496,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.821129032258066e-05,
+      "loss": 0.1395,
+      "step": 7810
+    },
+    {
+      "epoch": 0.124976,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.820967741935485e-05,
+      "loss": 0.1677,
+      "step": 7811
+    },
+    {
+      "epoch": 0.124992,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.820806451612905e-05,
+      "loss": 0.2011,
+      "step": 7812
+    },
+    {
+      "epoch": 0.125008,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.820645161290323e-05,
+      "loss": 0.156,
+      "step": 7813
+    },
+    {
+      "epoch": 0.125024,
+      "grad_norm": 1.3046875,
+      "learning_rate": 8.820483870967743e-05,
+      "loss": 0.1995,
+      "step": 7814
+    },
+    {
+      "epoch": 0.12504,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.820322580645162e-05,
+      "loss": 0.1697,
+      "step": 7815
+    },
+    {
+      "epoch": 0.125056,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.82016129032258e-05,
+      "loss": 0.146,
+      "step": 7816
+    },
+    {
+      "epoch": 0.125072,
+      "grad_norm": 1.3359375,
+      "learning_rate": 8.82e-05,
+      "loss": 0.2289,
+      "step": 7817
+    },
+    {
+      "epoch": 0.125088,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.819838709677419e-05,
+      "loss": 0.1764,
+      "step": 7818
+    },
+    {
+      "epoch": 0.125104,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.819677419354839e-05,
+      "loss": 0.1673,
+      "step": 7819
+    },
+    {
+      "epoch": 0.12512,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.819516129032257e-05,
+      "loss": 0.1522,
+      "step": 7820
+    },
+    {
+      "epoch": 0.125136,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.819354838709677e-05,
+      "loss": 0.1635,
+      "step": 7821
+    },
+    {
+      "epoch": 0.125152,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.819193548387097e-05,
+      "loss": 0.144,
+      "step": 7822
+    },
+    {
+      "epoch": 0.125168,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.819032258064517e-05,
+      "loss": 0.1912,
+      "step": 7823
+    },
+    {
+      "epoch": 0.125184,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.818870967741936e-05,
+      "loss": 0.1593,
+      "step": 7824
+    },
+    {
+      "epoch": 0.1252,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.818709677419356e-05,
+      "loss": 0.2176,
+      "step": 7825
+    },
+    {
+      "epoch": 0.125216,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.818548387096775e-05,
+      "loss": 0.1833,
+      "step": 7826
+    },
+    {
+      "epoch": 0.125232,
+      "grad_norm": 0.625,
+      "learning_rate": 8.818387096774194e-05,
+      "loss": 0.1663,
+      "step": 7827
+    },
+    {
+      "epoch": 0.125248,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.818225806451613e-05,
+      "loss": 0.0971,
+      "step": 7828
+    },
+    {
+      "epoch": 0.125264,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.818064516129033e-05,
+      "loss": 0.1641,
+      "step": 7829
+    },
+    {
+      "epoch": 0.12528,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.817903225806452e-05,
+      "loss": 0.1659,
+      "step": 7830
+    },
+    {
+      "epoch": 0.125296,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.81774193548387e-05,
+      "loss": 0.1413,
+      "step": 7831
+    },
+    {
+      "epoch": 0.125312,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.81758064516129e-05,
+      "loss": 0.1716,
+      "step": 7832
+    },
+    {
+      "epoch": 0.125328,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.81741935483871e-05,
+      "loss": 0.1858,
+      "step": 7833
+    },
+    {
+      "epoch": 0.125344,
+      "grad_norm": 0.4921875,
+      "learning_rate": 8.81725806451613e-05,
+      "loss": 0.14,
+      "step": 7834
+    },
+    {
+      "epoch": 0.12536,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.817096774193549e-05,
+      "loss": 0.1633,
+      "step": 7835
+    },
+    {
+      "epoch": 0.125376,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.816935483870969e-05,
+      "loss": 0.175,
+      "step": 7836
+    },
+    {
+      "epoch": 0.125392,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.816774193548387e-05,
+      "loss": 0.1863,
+      "step": 7837
+    },
+    {
+      "epoch": 0.125408,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.816612903225807e-05,
+      "loss": 0.1518,
+      "step": 7838
+    },
+    {
+      "epoch": 0.125424,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.816451612903226e-05,
+      "loss": 0.1639,
+      "step": 7839
+    },
+    {
+      "epoch": 0.12544,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.816290322580646e-05,
+      "loss": 0.1887,
+      "step": 7840
+    },
+    {
+      "epoch": 0.125456,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.816129032258064e-05,
+      "loss": 0.2165,
+      "step": 7841
+    },
+    {
+      "epoch": 0.125472,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.815967741935484e-05,
+      "loss": 0.1672,
+      "step": 7842
+    },
+    {
+      "epoch": 0.125488,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.815806451612903e-05,
+      "loss": 0.1927,
+      "step": 7843
+    },
+    {
+      "epoch": 0.125504,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.815645161290323e-05,
+      "loss": 0.1809,
+      "step": 7844
+    },
+    {
+      "epoch": 0.12552,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.815483870967743e-05,
+      "loss": 0.134,
+      "step": 7845
+    },
+    {
+      "epoch": 0.125536,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.815322580645163e-05,
+      "loss": 0.175,
+      "step": 7846
+    },
+    {
+      "epoch": 0.125552,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.815161290322582e-05,
+      "loss": 0.1432,
+      "step": 7847
+    },
+    {
+      "epoch": 0.125568,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.815e-05,
+      "loss": 0.1788,
+      "step": 7848
+    },
+    {
+      "epoch": 0.125584,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.81483870967742e-05,
+      "loss": 0.1713,
+      "step": 7849
+    },
+    {
+      "epoch": 0.1256,
+      "grad_norm": 1.265625,
+      "learning_rate": 8.814677419354839e-05,
+      "loss": 0.1945,
+      "step": 7850
+    },
+    {
+      "epoch": 0.125616,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.814516129032259e-05,
+      "loss": 0.2105,
+      "step": 7851
+    },
+    {
+      "epoch": 0.125632,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.814354838709677e-05,
+      "loss": 0.1858,
+      "step": 7852
+    },
+    {
+      "epoch": 0.125648,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.814193548387097e-05,
+      "loss": 0.2026,
+      "step": 7853
+    },
+    {
+      "epoch": 0.125664,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.814032258064516e-05,
+      "loss": 0.17,
+      "step": 7854
+    },
+    {
+      "epoch": 0.12568,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.813870967741936e-05,
+      "loss": 0.2198,
+      "step": 7855
+    },
+    {
+      "epoch": 0.125696,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.813709677419354e-05,
+      "loss": 0.1863,
+      "step": 7856
+    },
+    {
+      "epoch": 0.125712,
+      "grad_norm": 1.375,
+      "learning_rate": 8.813548387096774e-05,
+      "loss": 0.1618,
+      "step": 7857
+    },
+    {
+      "epoch": 0.125728,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.813387096774194e-05,
+      "loss": 0.1722,
+      "step": 7858
+    },
+    {
+      "epoch": 0.125744,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.813225806451614e-05,
+      "loss": 0.1538,
+      "step": 7859
+    },
+    {
+      "epoch": 0.12576,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.813064516129033e-05,
+      "loss": 0.1768,
+      "step": 7860
+    },
+    {
+      "epoch": 0.125776,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.812903225806453e-05,
+      "loss": 0.1693,
+      "step": 7861
+    },
+    {
+      "epoch": 0.125792,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.812741935483871e-05,
+      "loss": 0.1733,
+      "step": 7862
+    },
+    {
+      "epoch": 0.125808,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.81258064516129e-05,
+      "loss": 0.1742,
+      "step": 7863
+    },
+    {
+      "epoch": 0.125824,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.81241935483871e-05,
+      "loss": 0.1739,
+      "step": 7864
+    },
+    {
+      "epoch": 0.12584,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.812258064516129e-05,
+      "loss": 0.1926,
+      "step": 7865
+    },
+    {
+      "epoch": 0.125856,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.812096774193549e-05,
+      "loss": 0.1647,
+      "step": 7866
+    },
+    {
+      "epoch": 0.125872,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.811935483870967e-05,
+      "loss": 0.2094,
+      "step": 7867
+    },
+    {
+      "epoch": 0.125888,
+      "grad_norm": 0.455078125,
+      "learning_rate": 8.811774193548387e-05,
+      "loss": 0.1454,
+      "step": 7868
+    },
+    {
+      "epoch": 0.125904,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.811612903225807e-05,
+      "loss": 0.1907,
+      "step": 7869
+    },
+    {
+      "epoch": 0.12592,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.811451612903227e-05,
+      "loss": 0.1722,
+      "step": 7870
+    },
+    {
+      "epoch": 0.125936,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.811290322580646e-05,
+      "loss": 0.2148,
+      "step": 7871
+    },
+    {
+      "epoch": 0.125952,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.811129032258066e-05,
+      "loss": 0.1841,
+      "step": 7872
+    },
+    {
+      "epoch": 0.125968,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.810967741935484e-05,
+      "loss": 0.1859,
+      "step": 7873
+    },
+    {
+      "epoch": 0.125984,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.810806451612904e-05,
+      "loss": 0.1685,
+      "step": 7874
+    },
+    {
+      "epoch": 0.126,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.810645161290323e-05,
+      "loss": 0.1762,
+      "step": 7875
+    },
+    {
+      "epoch": 0.126016,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.810483870967743e-05,
+      "loss": 0.1677,
+      "step": 7876
+    },
+    {
+      "epoch": 0.126032,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.810322580645161e-05,
+      "loss": 0.149,
+      "step": 7877
+    },
+    {
+      "epoch": 0.126048,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.81016129032258e-05,
+      "loss": 0.1874,
+      "step": 7878
+    },
+    {
+      "epoch": 0.126064,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.81e-05,
+      "loss": 0.1969,
+      "step": 7879
+    },
+    {
+      "epoch": 0.12608,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.80983870967742e-05,
+      "loss": 0.1683,
+      "step": 7880
+    },
+    {
+      "epoch": 0.126096,
+      "grad_norm": 1.4140625,
+      "learning_rate": 8.809677419354839e-05,
+      "loss": 0.1662,
+      "step": 7881
+    },
+    {
+      "epoch": 0.126112,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.809516129032259e-05,
+      "loss": 0.1819,
+      "step": 7882
+    },
+    {
+      "epoch": 0.126128,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.809354838709679e-05,
+      "loss": 0.1488,
+      "step": 7883
+    },
+    {
+      "epoch": 0.126144,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.809193548387097e-05,
+      "loss": 0.1696,
+      "step": 7884
+    },
+    {
+      "epoch": 0.12616,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.809032258064517e-05,
+      "loss": 0.1817,
+      "step": 7885
+    },
+    {
+      "epoch": 0.126176,
+      "grad_norm": 0.625,
+      "learning_rate": 8.808870967741936e-05,
+      "loss": 0.1568,
+      "step": 7886
+    },
+    {
+      "epoch": 0.126192,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.808709677419356e-05,
+      "loss": 0.2174,
+      "step": 7887
+    },
+    {
+      "epoch": 0.126208,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.808548387096774e-05,
+      "loss": 0.1638,
+      "step": 7888
+    },
+    {
+      "epoch": 0.126224,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.808387096774194e-05,
+      "loss": 0.1956,
+      "step": 7889
+    },
+    {
+      "epoch": 0.12624,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.808225806451613e-05,
+      "loss": 0.1947,
+      "step": 7890
+    },
+    {
+      "epoch": 0.126256,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.808064516129033e-05,
+      "loss": 0.1781,
+      "step": 7891
+    },
+    {
+      "epoch": 0.126272,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.807903225806451e-05,
+      "loss": 0.1577,
+      "step": 7892
+    },
+    {
+      "epoch": 0.126288,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.807741935483871e-05,
+      "loss": 0.138,
+      "step": 7893
+    },
+    {
+      "epoch": 0.126304,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.807580645161291e-05,
+      "loss": 0.1933,
+      "step": 7894
+    },
+    {
+      "epoch": 0.12632,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.80741935483871e-05,
+      "loss": 0.1961,
+      "step": 7895
+    },
+    {
+      "epoch": 0.126336,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.80725806451613e-05,
+      "loss": 0.1592,
+      "step": 7896
+    },
+    {
+      "epoch": 0.126352,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.807096774193549e-05,
+      "loss": 0.2101,
+      "step": 7897
+    },
+    {
+      "epoch": 0.126368,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.806935483870968e-05,
+      "loss": 0.1686,
+      "step": 7898
+    },
+    {
+      "epoch": 0.126384,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.806774193548387e-05,
+      "loss": 0.1745,
+      "step": 7899
+    },
+    {
+      "epoch": 0.1264,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.806612903225807e-05,
+      "loss": 0.1578,
+      "step": 7900
+    },
+    {
+      "epoch": 0.126416,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.806451612903226e-05,
+      "loss": 0.149,
+      "step": 7901
+    },
+    {
+      "epoch": 0.126432,
+      "grad_norm": 1.0,
+      "learning_rate": 8.806290322580646e-05,
+      "loss": 0.1902,
+      "step": 7902
+    },
+    {
+      "epoch": 0.126448,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.806129032258064e-05,
+      "loss": 0.2033,
+      "step": 7903
+    },
+    {
+      "epoch": 0.126464,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.805967741935484e-05,
+      "loss": 0.1512,
+      "step": 7904
+    },
+    {
+      "epoch": 0.12648,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.805806451612904e-05,
+      "loss": 0.1796,
+      "step": 7905
+    },
+    {
+      "epoch": 0.126496,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.805645161290324e-05,
+      "loss": 0.1841,
+      "step": 7906
+    },
+    {
+      "epoch": 0.126512,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.805483870967743e-05,
+      "loss": 0.1197,
+      "step": 7907
+    },
+    {
+      "epoch": 0.126528,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.805322580645163e-05,
+      "loss": 0.1772,
+      "step": 7908
+    },
+    {
+      "epoch": 0.126544,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.805161290322581e-05,
+      "loss": 0.1773,
+      "step": 7909
+    },
+    {
+      "epoch": 0.12656,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.805e-05,
+      "loss": 0.2046,
+      "step": 7910
+    },
+    {
+      "epoch": 0.126576,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.80483870967742e-05,
+      "loss": 0.2008,
+      "step": 7911
+    },
+    {
+      "epoch": 0.126592,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.804677419354838e-05,
+      "loss": 0.1708,
+      "step": 7912
+    },
+    {
+      "epoch": 0.126608,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.804516129032258e-05,
+      "loss": 0.1922,
+      "step": 7913
+    },
+    {
+      "epoch": 0.126624,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.804354838709677e-05,
+      "loss": 0.1375,
+      "step": 7914
+    },
+    {
+      "epoch": 0.12664,
+      "grad_norm": 1.2578125,
+      "learning_rate": 8.804193548387097e-05,
+      "loss": 0.1652,
+      "step": 7915
+    },
+    {
+      "epoch": 0.126656,
+      "grad_norm": 1.5703125,
+      "learning_rate": 8.804032258064516e-05,
+      "loss": 0.1751,
+      "step": 7916
+    },
+    {
+      "epoch": 0.126672,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.803870967741936e-05,
+      "loss": 0.1634,
+      "step": 7917
+    },
+    {
+      "epoch": 0.126688,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.803709677419356e-05,
+      "loss": 0.1612,
+      "step": 7918
+    },
+    {
+      "epoch": 0.126704,
+      "grad_norm": 1.3828125,
+      "learning_rate": 8.803548387096775e-05,
+      "loss": 0.1723,
+      "step": 7919
+    },
+    {
+      "epoch": 0.12672,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.803387096774194e-05,
+      "loss": 0.1702,
+      "step": 7920
+    },
+    {
+      "epoch": 0.126736,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.803225806451614e-05,
+      "loss": 0.1537,
+      "step": 7921
+    },
+    {
+      "epoch": 0.126752,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.803064516129033e-05,
+      "loss": 0.1355,
+      "step": 7922
+    },
+    {
+      "epoch": 0.126768,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.802903225806453e-05,
+      "loss": 0.1619,
+      "step": 7923
+    },
+    {
+      "epoch": 0.126784,
+      "grad_norm": 0.625,
+      "learning_rate": 8.802741935483871e-05,
+      "loss": 0.1628,
+      "step": 7924
+    },
+    {
+      "epoch": 0.1268,
+      "grad_norm": 1.125,
+      "learning_rate": 8.80258064516129e-05,
+      "loss": 0.1775,
+      "step": 7925
+    },
+    {
+      "epoch": 0.126816,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.80241935483871e-05,
+      "loss": 0.1413,
+      "step": 7926
+    },
+    {
+      "epoch": 0.126832,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.802258064516128e-05,
+      "loss": 0.2017,
+      "step": 7927
+    },
+    {
+      "epoch": 0.126848,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.802096774193548e-05,
+      "loss": 0.1523,
+      "step": 7928
+    },
+    {
+      "epoch": 0.126864,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.801935483870968e-05,
+      "loss": 0.2202,
+      "step": 7929
+    },
+    {
+      "epoch": 0.12688,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.801774193548388e-05,
+      "loss": 0.1899,
+      "step": 7930
+    },
+    {
+      "epoch": 0.126896,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.801612903225807e-05,
+      "loss": 0.1835,
+      "step": 7931
+    },
+    {
+      "epoch": 0.126912,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.801451612903227e-05,
+      "loss": 0.1542,
+      "step": 7932
+    },
+    {
+      "epoch": 0.126928,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.801290322580645e-05,
+      "loss": 0.146,
+      "step": 7933
+    },
+    {
+      "epoch": 0.126944,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.801129032258065e-05,
+      "loss": 0.2057,
+      "step": 7934
+    },
+    {
+      "epoch": 0.12696,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.800967741935484e-05,
+      "loss": 0.1943,
+      "step": 7935
+    },
+    {
+      "epoch": 0.126976,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.800806451612904e-05,
+      "loss": 0.1943,
+      "step": 7936
+    },
+    {
+      "epoch": 0.126992,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.800645161290323e-05,
+      "loss": 0.1825,
+      "step": 7937
+    },
+    {
+      "epoch": 0.127008,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.800483870967743e-05,
+      "loss": 0.181,
+      "step": 7938
+    },
+    {
+      "epoch": 0.127024,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.800322580645161e-05,
+      "loss": 0.184,
+      "step": 7939
+    },
+    {
+      "epoch": 0.12704,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.800161290322581e-05,
+      "loss": 0.1494,
+      "step": 7940
+    },
+    {
+      "epoch": 0.127056,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.800000000000001e-05,
+      "loss": 0.1719,
+      "step": 7941
+    },
+    {
+      "epoch": 0.127072,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.79983870967742e-05,
+      "loss": 0.1447,
+      "step": 7942
+    },
+    {
+      "epoch": 0.127088,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.79967741935484e-05,
+      "loss": 0.1506,
+      "step": 7943
+    },
+    {
+      "epoch": 0.127104,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.799516129032258e-05,
+      "loss": 0.1515,
+      "step": 7944
+    },
+    {
+      "epoch": 0.12712,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.799354838709678e-05,
+      "loss": 0.1731,
+      "step": 7945
+    },
+    {
+      "epoch": 0.127136,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.799193548387097e-05,
+      "loss": 0.1857,
+      "step": 7946
+    },
+    {
+      "epoch": 0.127152,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.799032258064517e-05,
+      "loss": 0.2238,
+      "step": 7947
+    },
+    {
+      "epoch": 0.127168,
+      "grad_norm": 1.3046875,
+      "learning_rate": 8.798870967741935e-05,
+      "loss": 0.2362,
+      "step": 7948
+    },
+    {
+      "epoch": 0.127184,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.798709677419355e-05,
+      "loss": 0.1705,
+      "step": 7949
+    },
+    {
+      "epoch": 0.1272,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.798548387096774e-05,
+      "loss": 0.1501,
+      "step": 7950
+    },
+    {
+      "epoch": 0.127216,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.798387096774194e-05,
+      "loss": 0.1344,
+      "step": 7951
+    },
+    {
+      "epoch": 0.127232,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.798225806451613e-05,
+      "loss": 0.182,
+      "step": 7952
+    },
+    {
+      "epoch": 0.127248,
+      "grad_norm": 0.875,
+      "learning_rate": 8.798064516129033e-05,
+      "loss": 0.1564,
+      "step": 7953
+    },
+    {
+      "epoch": 0.127264,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.797903225806453e-05,
+      "loss": 0.1654,
+      "step": 7954
+    },
+    {
+      "epoch": 0.12728,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.797741935483872e-05,
+      "loss": 0.1656,
+      "step": 7955
+    },
+    {
+      "epoch": 0.127296,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.797580645161291e-05,
+      "loss": 0.1889,
+      "step": 7956
+    },
+    {
+      "epoch": 0.127312,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.79741935483871e-05,
+      "loss": 0.1878,
+      "step": 7957
+    },
+    {
+      "epoch": 0.127328,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.79725806451613e-05,
+      "loss": 0.1794,
+      "step": 7958
+    },
+    {
+      "epoch": 0.127344,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.797096774193548e-05,
+      "loss": 0.1808,
+      "step": 7959
+    },
+    {
+      "epoch": 0.12736,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.796935483870968e-05,
+      "loss": 0.1854,
+      "step": 7960
+    },
+    {
+      "epoch": 0.127376,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.796774193548387e-05,
+      "loss": 0.1534,
+      "step": 7961
+    },
+    {
+      "epoch": 0.127392,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.796612903225807e-05,
+      "loss": 0.2009,
+      "step": 7962
+    },
+    {
+      "epoch": 0.127408,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.796451612903225e-05,
+      "loss": 0.1589,
+      "step": 7963
+    },
+    {
+      "epoch": 0.127424,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.796290322580645e-05,
+      "loss": 0.1752,
+      "step": 7964
+    },
+    {
+      "epoch": 0.12744,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.796129032258065e-05,
+      "loss": 0.1557,
+      "step": 7965
+    },
+    {
+      "epoch": 0.127456,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.795967741935485e-05,
+      "loss": 0.1709,
+      "step": 7966
+    },
+    {
+      "epoch": 0.127472,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.795806451612904e-05,
+      "loss": 0.1816,
+      "step": 7967
+    },
+    {
+      "epoch": 0.127488,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.795645161290324e-05,
+      "loss": 0.1943,
+      "step": 7968
+    },
+    {
+      "epoch": 0.127504,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.795483870967742e-05,
+      "loss": 0.1599,
+      "step": 7969
+    },
+    {
+      "epoch": 0.12752,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.795322580645162e-05,
+      "loss": 0.1272,
+      "step": 7970
+    },
+    {
+      "epoch": 0.127536,
+      "grad_norm": 1.46875,
+      "learning_rate": 8.795161290322581e-05,
+      "loss": 0.193,
+      "step": 7971
+    },
+    {
+      "epoch": 0.127552,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.795e-05,
+      "loss": 0.1386,
+      "step": 7972
+    },
+    {
+      "epoch": 0.127568,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.79483870967742e-05,
+      "loss": 0.1893,
+      "step": 7973
+    },
+    {
+      "epoch": 0.127584,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.794677419354838e-05,
+      "loss": 0.1931,
+      "step": 7974
+    },
+    {
+      "epoch": 0.1276,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.794516129032258e-05,
+      "loss": 0.1722,
+      "step": 7975
+    },
+    {
+      "epoch": 0.127616,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.794354838709677e-05,
+      "loss": 0.2186,
+      "step": 7976
+    },
+    {
+      "epoch": 0.127632,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.794193548387097e-05,
+      "loss": 0.1582,
+      "step": 7977
+    },
+    {
+      "epoch": 0.127648,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.794032258064517e-05,
+      "loss": 0.1628,
+      "step": 7978
+    },
+    {
+      "epoch": 0.127664,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.793870967741937e-05,
+      "loss": 0.1406,
+      "step": 7979
+    },
+    {
+      "epoch": 0.12768,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.793709677419355e-05,
+      "loss": 0.196,
+      "step": 7980
+    },
+    {
+      "epoch": 0.127696,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.793548387096775e-05,
+      "loss": 0.2088,
+      "step": 7981
+    },
+    {
+      "epoch": 0.127712,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.793387096774194e-05,
+      "loss": 0.2178,
+      "step": 7982
+    },
+    {
+      "epoch": 0.127728,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.793225806451614e-05,
+      "loss": 0.1903,
+      "step": 7983
+    },
+    {
+      "epoch": 0.127744,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.793064516129032e-05,
+      "loss": 0.235,
+      "step": 7984
+    },
+    {
+      "epoch": 0.12776,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.792903225806452e-05,
+      "loss": 0.1725,
+      "step": 7985
+    },
+    {
+      "epoch": 0.127776,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.792741935483871e-05,
+      "loss": 0.2348,
+      "step": 7986
+    },
+    {
+      "epoch": 0.127792,
+      "grad_norm": 1.46875,
+      "learning_rate": 8.79258064516129e-05,
+      "loss": 0.1729,
+      "step": 7987
+    },
+    {
+      "epoch": 0.127808,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.79241935483871e-05,
+      "loss": 0.1496,
+      "step": 7988
+    },
+    {
+      "epoch": 0.127824,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.79225806451613e-05,
+      "loss": 0.1995,
+      "step": 7989
+    },
+    {
+      "epoch": 0.12784,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.79209677419355e-05,
+      "loss": 0.154,
+      "step": 7990
+    },
+    {
+      "epoch": 0.127856,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.791935483870968e-05,
+      "loss": 0.17,
+      "step": 7991
+    },
+    {
+      "epoch": 0.127872,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.791774193548388e-05,
+      "loss": 0.1933,
+      "step": 7992
+    },
+    {
+      "epoch": 0.127888,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.791612903225807e-05,
+      "loss": 0.1652,
+      "step": 7993
+    },
+    {
+      "epoch": 0.127904,
+      "grad_norm": 1.4296875,
+      "learning_rate": 8.791451612903227e-05,
+      "loss": 0.1709,
+      "step": 7994
+    },
+    {
+      "epoch": 0.12792,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.791290322580645e-05,
+      "loss": 0.1381,
+      "step": 7995
+    },
+    {
+      "epoch": 0.127936,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.791129032258065e-05,
+      "loss": 0.1497,
+      "step": 7996
+    },
+    {
+      "epoch": 0.127952,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.790967741935484e-05,
+      "loss": 0.1478,
+      "step": 7997
+    },
+    {
+      "epoch": 0.127968,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.790806451612904e-05,
+      "loss": 0.1802,
+      "step": 7998
+    },
+    {
+      "epoch": 0.127984,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.790645161290322e-05,
+      "loss": 0.1966,
+      "step": 7999
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 1.5703125,
+      "learning_rate": 8.790483870967742e-05,
+      "loss": 0.1691,
+      "step": 8000
+    },
+    {
+      "epoch": 0.128016,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.790322580645162e-05,
+      "loss": 0.1321,
+      "step": 8001
+    },
+    {
+      "epoch": 0.128032,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.790161290322582e-05,
+      "loss": 0.1484,
+      "step": 8002
+    },
+    {
+      "epoch": 0.128048,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.790000000000001e-05,
+      "loss": 0.1851,
+      "step": 8003
+    },
+    {
+      "epoch": 0.128064,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.78983870967742e-05,
+      "loss": 0.1817,
+      "step": 8004
+    },
+    {
+      "epoch": 0.12808,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.78967741935484e-05,
+      "loss": 0.1629,
+      "step": 8005
+    },
+    {
+      "epoch": 0.128096,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.789516129032258e-05,
+      "loss": 0.2017,
+      "step": 8006
+    },
+    {
+      "epoch": 0.128112,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.789354838709678e-05,
+      "loss": 0.21,
+      "step": 8007
+    },
+    {
+      "epoch": 0.128128,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.789193548387097e-05,
+      "loss": 0.1907,
+      "step": 8008
+    },
+    {
+      "epoch": 0.128144,
+      "grad_norm": 1.421875,
+      "learning_rate": 8.789032258064517e-05,
+      "loss": 0.1981,
+      "step": 8009
+    },
+    {
+      "epoch": 0.12816,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.788870967741935e-05,
+      "loss": 0.2068,
+      "step": 8010
+    },
+    {
+      "epoch": 0.128176,
+      "grad_norm": 0.75,
+      "learning_rate": 8.788709677419355e-05,
+      "loss": 0.1727,
+      "step": 8011
+    },
+    {
+      "epoch": 0.128192,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.788548387096774e-05,
+      "loss": 0.1842,
+      "step": 8012
+    },
+    {
+      "epoch": 0.128208,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.788387096774194e-05,
+      "loss": 0.1486,
+      "step": 8013
+    },
+    {
+      "epoch": 0.128224,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.788225806451614e-05,
+      "loss": 0.1792,
+      "step": 8014
+    },
+    {
+      "epoch": 0.12824,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.788064516129034e-05,
+      "loss": 0.1658,
+      "step": 8015
+    },
+    {
+      "epoch": 0.128256,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.787903225806452e-05,
+      "loss": 0.1525,
+      "step": 8016
+    },
+    {
+      "epoch": 0.128272,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.787741935483872e-05,
+      "loss": 0.201,
+      "step": 8017
+    },
+    {
+      "epoch": 0.128288,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.787580645161291e-05,
+      "loss": 0.1691,
+      "step": 8018
+    },
+    {
+      "epoch": 0.128304,
+      "grad_norm": 0.625,
+      "learning_rate": 8.78741935483871e-05,
+      "loss": 0.1416,
+      "step": 8019
+    },
+    {
+      "epoch": 0.12832,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.78725806451613e-05,
+      "loss": 0.2343,
+      "step": 8020
+    },
+    {
+      "epoch": 0.128336,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.787096774193548e-05,
+      "loss": 0.1478,
+      "step": 8021
+    },
+    {
+      "epoch": 0.128352,
+      "grad_norm": 1.234375,
+      "learning_rate": 8.786935483870968e-05,
+      "loss": 0.1812,
+      "step": 8022
+    },
+    {
+      "epoch": 0.128368,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.786774193548387e-05,
+      "loss": 0.1821,
+      "step": 8023
+    },
+    {
+      "epoch": 0.128384,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.786612903225807e-05,
+      "loss": 0.1911,
+      "step": 8024
+    },
+    {
+      "epoch": 0.1284,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.786451612903227e-05,
+      "loss": 0.1698,
+      "step": 8025
+    },
+    {
+      "epoch": 0.128416,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.786290322580646e-05,
+      "loss": 0.209,
+      "step": 8026
+    },
+    {
+      "epoch": 0.128432,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.786129032258065e-05,
+      "loss": 0.171,
+      "step": 8027
+    },
+    {
+      "epoch": 0.128448,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.785967741935485e-05,
+      "loss": 0.1476,
+      "step": 8028
+    },
+    {
+      "epoch": 0.128464,
+      "grad_norm": 1.40625,
+      "learning_rate": 8.785806451612904e-05,
+      "loss": 0.1954,
+      "step": 8029
+    },
+    {
+      "epoch": 0.12848,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.785645161290324e-05,
+      "loss": 0.1965,
+      "step": 8030
+    },
+    {
+      "epoch": 0.128496,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.785483870967742e-05,
+      "loss": 0.1546,
+      "step": 8031
+    },
+    {
+      "epoch": 0.128512,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.785322580645162e-05,
+      "loss": 0.2037,
+      "step": 8032
+    },
+    {
+      "epoch": 0.128528,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.785161290322581e-05,
+      "loss": 0.1454,
+      "step": 8033
+    },
+    {
+      "epoch": 0.128544,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.785e-05,
+      "loss": 0.165,
+      "step": 8034
+    },
+    {
+      "epoch": 0.12856,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.78483870967742e-05,
+      "loss": 0.1378,
+      "step": 8035
+    },
+    {
+      "epoch": 0.128576,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.78467741935484e-05,
+      "loss": 0.1525,
+      "step": 8036
+    },
+    {
+      "epoch": 0.128592,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.784516129032259e-05,
+      "loss": 0.1307,
+      "step": 8037
+    },
+    {
+      "epoch": 0.128608,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.784354838709678e-05,
+      "loss": 0.1621,
+      "step": 8038
+    },
+    {
+      "epoch": 0.128624,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.784193548387098e-05,
+      "loss": 0.1783,
+      "step": 8039
+    },
+    {
+      "epoch": 0.12864,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.784032258064516e-05,
+      "loss": 0.1713,
+      "step": 8040
+    },
+    {
+      "epoch": 0.128656,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.783870967741936e-05,
+      "loss": 0.1743,
+      "step": 8041
+    },
+    {
+      "epoch": 0.128672,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.783709677419355e-05,
+      "loss": 0.1783,
+      "step": 8042
+    },
+    {
+      "epoch": 0.128688,
+      "grad_norm": 1.28125,
+      "learning_rate": 8.783548387096775e-05,
+      "loss": 0.2195,
+      "step": 8043
+    },
+    {
+      "epoch": 0.128704,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.783387096774194e-05,
+      "loss": 0.1425,
+      "step": 8044
+    },
+    {
+      "epoch": 0.12872,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.783225806451614e-05,
+      "loss": 0.1907,
+      "step": 8045
+    },
+    {
+      "epoch": 0.128736,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.783064516129032e-05,
+      "loss": 0.1722,
+      "step": 8046
+    },
+    {
+      "epoch": 0.128752,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.782903225806452e-05,
+      "loss": 0.1901,
+      "step": 8047
+    },
+    {
+      "epoch": 0.128768,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.782741935483871e-05,
+      "loss": 0.1709,
+      "step": 8048
+    },
+    {
+      "epoch": 0.128784,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.782580645161291e-05,
+      "loss": 0.1798,
+      "step": 8049
+    },
+    {
+      "epoch": 0.1288,
+      "grad_norm": 1.2578125,
+      "learning_rate": 8.782419354838711e-05,
+      "loss": 0.1887,
+      "step": 8050
+    },
+    {
+      "epoch": 0.128816,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.782258064516129e-05,
+      "loss": 0.2193,
+      "step": 8051
+    },
+    {
+      "epoch": 0.128832,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.782096774193549e-05,
+      "loss": 0.1407,
+      "step": 8052
+    },
+    {
+      "epoch": 0.128848,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.781935483870968e-05,
+      "loss": 0.1803,
+      "step": 8053
+    },
+    {
+      "epoch": 0.128864,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.781774193548388e-05,
+      "loss": 0.1606,
+      "step": 8054
+    },
+    {
+      "epoch": 0.12888,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.781612903225806e-05,
+      "loss": 0.1904,
+      "step": 8055
+    },
+    {
+      "epoch": 0.128896,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.781451612903226e-05,
+      "loss": 0.1851,
+      "step": 8056
+    },
+    {
+      "epoch": 0.128912,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.781290322580645e-05,
+      "loss": 0.2105,
+      "step": 8057
+    },
+    {
+      "epoch": 0.128928,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.781129032258065e-05,
+      "loss": 0.1649,
+      "step": 8058
+    },
+    {
+      "epoch": 0.128944,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.780967741935484e-05,
+      "loss": 0.2085,
+      "step": 8059
+    },
+    {
+      "epoch": 0.12896,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.780806451612904e-05,
+      "loss": 0.17,
+      "step": 8060
+    },
+    {
+      "epoch": 0.128976,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.780645161290323e-05,
+      "loss": 0.1899,
+      "step": 8061
+    },
+    {
+      "epoch": 0.128992,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.780483870967743e-05,
+      "loss": 0.2164,
+      "step": 8062
+    },
+    {
+      "epoch": 0.129008,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.780322580645162e-05,
+      "loss": 0.2045,
+      "step": 8063
+    },
+    {
+      "epoch": 0.129024,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.780161290322582e-05,
+      "loss": 0.2167,
+      "step": 8064
+    },
+    {
+      "epoch": 0.12904,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.78e-05,
+      "loss": 0.1698,
+      "step": 8065
+    },
+    {
+      "epoch": 0.129056,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.779838709677419e-05,
+      "loss": 0.1783,
+      "step": 8066
+    },
+    {
+      "epoch": 0.129072,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.779677419354839e-05,
+      "loss": 0.1729,
+      "step": 8067
+    },
+    {
+      "epoch": 0.129088,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.779516129032258e-05,
+      "loss": 0.1915,
+      "step": 8068
+    },
+    {
+      "epoch": 0.129104,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.779354838709678e-05,
+      "loss": 0.1689,
+      "step": 8069
+    },
+    {
+      "epoch": 0.12912,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.779193548387096e-05,
+      "loss": 0.164,
+      "step": 8070
+    },
+    {
+      "epoch": 0.129136,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.779032258064516e-05,
+      "loss": 0.1278,
+      "step": 8071
+    },
+    {
+      "epoch": 0.129152,
+      "grad_norm": 1.0,
+      "learning_rate": 8.778870967741935e-05,
+      "loss": 0.1418,
+      "step": 8072
+    },
+    {
+      "epoch": 0.129168,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.778709677419355e-05,
+      "loss": 0.218,
+      "step": 8073
+    },
+    {
+      "epoch": 0.129184,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.778548387096775e-05,
+      "loss": 0.1524,
+      "step": 8074
+    },
+    {
+      "epoch": 0.1292,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.778387096774195e-05,
+      "loss": 0.1783,
+      "step": 8075
+    },
+    {
+      "epoch": 0.129216,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.778225806451613e-05,
+      "loss": 0.198,
+      "step": 8076
+    },
+    {
+      "epoch": 0.129232,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.778064516129033e-05,
+      "loss": 0.216,
+      "step": 8077
+    },
+    {
+      "epoch": 0.129248,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.777903225806452e-05,
+      "loss": 0.1765,
+      "step": 8078
+    },
+    {
+      "epoch": 0.129264,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.777741935483872e-05,
+      "loss": 0.2083,
+      "step": 8079
+    },
+    {
+      "epoch": 0.12928,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.77758064516129e-05,
+      "loss": 0.1571,
+      "step": 8080
+    },
+    {
+      "epoch": 0.129296,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.777419354838709e-05,
+      "loss": 0.1638,
+      "step": 8081
+    },
+    {
+      "epoch": 0.129312,
+      "grad_norm": 1.3046875,
+      "learning_rate": 8.777258064516129e-05,
+      "loss": 0.1701,
+      "step": 8082
+    },
+    {
+      "epoch": 0.129328,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.777096774193548e-05,
+      "loss": 0.201,
+      "step": 8083
+    },
+    {
+      "epoch": 0.129344,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.776935483870968e-05,
+      "loss": 0.1365,
+      "step": 8084
+    },
+    {
+      "epoch": 0.12936,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.776774193548388e-05,
+      "loss": 0.1413,
+      "step": 8085
+    },
+    {
+      "epoch": 0.129376,
+      "grad_norm": 1.234375,
+      "learning_rate": 8.776612903225808e-05,
+      "loss": 0.1515,
+      "step": 8086
+    },
+    {
+      "epoch": 0.129392,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.776451612903226e-05,
+      "loss": 0.1678,
+      "step": 8087
+    },
+    {
+      "epoch": 0.129408,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.776290322580646e-05,
+      "loss": 0.186,
+      "step": 8088
+    },
+    {
+      "epoch": 0.129424,
+      "grad_norm": 1.296875,
+      "learning_rate": 8.776129032258065e-05,
+      "loss": 0.211,
+      "step": 8089
+    },
+    {
+      "epoch": 0.12944,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.775967741935485e-05,
+      "loss": 0.1914,
+      "step": 8090
+    },
+    {
+      "epoch": 0.129456,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.775806451612903e-05,
+      "loss": 0.1392,
+      "step": 8091
+    },
+    {
+      "epoch": 0.129472,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.775645161290323e-05,
+      "loss": 0.1489,
+      "step": 8092
+    },
+    {
+      "epoch": 0.129488,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.775483870967742e-05,
+      "loss": 0.1662,
+      "step": 8093
+    },
+    {
+      "epoch": 0.129504,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.775322580645162e-05,
+      "loss": 0.1455,
+      "step": 8094
+    },
+    {
+      "epoch": 0.12952,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.77516129032258e-05,
+      "loss": 0.1633,
+      "step": 8095
+    },
+    {
+      "epoch": 0.129536,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.775e-05,
+      "loss": 0.2179,
+      "step": 8096
+    },
+    {
+      "epoch": 0.129552,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.77483870967742e-05,
+      "loss": 0.1725,
+      "step": 8097
+    },
+    {
+      "epoch": 0.129568,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.774677419354839e-05,
+      "loss": 0.1704,
+      "step": 8098
+    },
+    {
+      "epoch": 0.129584,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.774516129032259e-05,
+      "loss": 0.2,
+      "step": 8099
+    },
+    {
+      "epoch": 0.1296,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.774354838709678e-05,
+      "loss": 0.1464,
+      "step": 8100
+    },
+    {
+      "epoch": 0.129616,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.774193548387098e-05,
+      "loss": 0.1647,
+      "step": 8101
+    },
+    {
+      "epoch": 0.129632,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.774032258064516e-05,
+      "loss": 0.1585,
+      "step": 8102
+    },
+    {
+      "epoch": 0.129648,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.773870967741936e-05,
+      "loss": 0.1368,
+      "step": 8103
+    },
+    {
+      "epoch": 0.129664,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.773709677419355e-05,
+      "loss": 0.1594,
+      "step": 8104
+    },
+    {
+      "epoch": 0.12968,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.773548387096775e-05,
+      "loss": 0.1684,
+      "step": 8105
+    },
+    {
+      "epoch": 0.129696,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.773387096774193e-05,
+      "loss": 0.1961,
+      "step": 8106
+    },
+    {
+      "epoch": 0.129712,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.773225806451613e-05,
+      "loss": 0.1831,
+      "step": 8107
+    },
+    {
+      "epoch": 0.129728,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.773064516129032e-05,
+      "loss": 0.1904,
+      "step": 8108
+    },
+    {
+      "epoch": 0.129744,
+      "grad_norm": 0.75,
+      "learning_rate": 8.772903225806452e-05,
+      "loss": 0.1809,
+      "step": 8109
+    },
+    {
+      "epoch": 0.12976,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.772741935483872e-05,
+      "loss": 0.1666,
+      "step": 8110
+    },
+    {
+      "epoch": 0.129776,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.77258064516129e-05,
+      "loss": 0.1779,
+      "step": 8111
+    },
+    {
+      "epoch": 0.129792,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.77241935483871e-05,
+      "loss": 0.1809,
+      "step": 8112
+    },
+    {
+      "epoch": 0.129808,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.772258064516129e-05,
+      "loss": 0.1574,
+      "step": 8113
+    },
+    {
+      "epoch": 0.129824,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.772096774193549e-05,
+      "loss": 0.1686,
+      "step": 8114
+    },
+    {
+      "epoch": 0.12984,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.771935483870968e-05,
+      "loss": 0.1521,
+      "step": 8115
+    },
+    {
+      "epoch": 0.129856,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.771774193548388e-05,
+      "loss": 0.1618,
+      "step": 8116
+    },
+    {
+      "epoch": 0.129872,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.771612903225806e-05,
+      "loss": 0.1592,
+      "step": 8117
+    },
+    {
+      "epoch": 0.129888,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.771451612903226e-05,
+      "loss": 0.2323,
+      "step": 8118
+    },
+    {
+      "epoch": 0.129904,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.771290322580645e-05,
+      "loss": 0.1739,
+      "step": 8119
+    },
+    {
+      "epoch": 0.12992,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.771129032258065e-05,
+      "loss": 0.1557,
+      "step": 8120
+    },
+    {
+      "epoch": 0.129936,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.770967741935485e-05,
+      "loss": 0.1909,
+      "step": 8121
+    },
+    {
+      "epoch": 0.129952,
+      "grad_norm": 1.125,
+      "learning_rate": 8.770806451612905e-05,
+      "loss": 0.1734,
+      "step": 8122
+    },
+    {
+      "epoch": 0.129968,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.770645161290323e-05,
+      "loss": 0.1922,
+      "step": 8123
+    },
+    {
+      "epoch": 0.129984,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.770483870967743e-05,
+      "loss": 0.2005,
+      "step": 8124
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.770322580645162e-05,
+      "loss": 0.16,
+      "step": 8125
+    },
+    {
+      "epoch": 0.130016,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.770161290322582e-05,
+      "loss": 0.1491,
+      "step": 8126
+    },
+    {
+      "epoch": 0.130032,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.77e-05,
+      "loss": 0.2153,
+      "step": 8127
+    },
+    {
+      "epoch": 0.130048,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.769838709677419e-05,
+      "loss": 0.2147,
+      "step": 8128
+    },
+    {
+      "epoch": 0.130064,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.769677419354839e-05,
+      "loss": 0.1602,
+      "step": 8129
+    },
+    {
+      "epoch": 0.13008,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.769516129032258e-05,
+      "loss": 0.1627,
+      "step": 8130
+    },
+    {
+      "epoch": 0.130096,
+      "grad_norm": 1.25,
+      "learning_rate": 8.769354838709678e-05,
+      "loss": 0.174,
+      "step": 8131
+    },
+    {
+      "epoch": 0.130112,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.769193548387098e-05,
+      "loss": 0.2542,
+      "step": 8132
+    },
+    {
+      "epoch": 0.130128,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.769032258064516e-05,
+      "loss": 0.1831,
+      "step": 8133
+    },
+    {
+      "epoch": 0.130144,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.768870967741936e-05,
+      "loss": 0.1278,
+      "step": 8134
+    },
+    {
+      "epoch": 0.13016,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.768709677419356e-05,
+      "loss": 0.1497,
+      "step": 8135
+    },
+    {
+      "epoch": 0.130176,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.768548387096775e-05,
+      "loss": 0.148,
+      "step": 8136
+    },
+    {
+      "epoch": 0.130192,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.768387096774195e-05,
+      "loss": 0.14,
+      "step": 8137
+    },
+    {
+      "epoch": 0.130208,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.768225806451613e-05,
+      "loss": 0.1804,
+      "step": 8138
+    },
+    {
+      "epoch": 0.130224,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.768064516129033e-05,
+      "loss": 0.1496,
+      "step": 8139
+    },
+    {
+      "epoch": 0.13024,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.767903225806452e-05,
+      "loss": 0.1518,
+      "step": 8140
+    },
+    {
+      "epoch": 0.130256,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.767741935483872e-05,
+      "loss": 0.1829,
+      "step": 8141
+    },
+    {
+      "epoch": 0.130272,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.76758064516129e-05,
+      "loss": 0.1647,
+      "step": 8142
+    },
+    {
+      "epoch": 0.130288,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.767419354838709e-05,
+      "loss": 0.1957,
+      "step": 8143
+    },
+    {
+      "epoch": 0.130304,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.767258064516129e-05,
+      "loss": 0.1652,
+      "step": 8144
+    },
+    {
+      "epoch": 0.13032,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.767096774193549e-05,
+      "loss": 0.1576,
+      "step": 8145
+    },
+    {
+      "epoch": 0.130336,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.766935483870969e-05,
+      "loss": 0.1854,
+      "step": 8146
+    },
+    {
+      "epoch": 0.130352,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.766774193548387e-05,
+      "loss": 0.1785,
+      "step": 8147
+    },
+    {
+      "epoch": 0.130368,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.766612903225807e-05,
+      "loss": 0.1536,
+      "step": 8148
+    },
+    {
+      "epoch": 0.130384,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.766451612903226e-05,
+      "loss": 0.162,
+      "step": 8149
+    },
+    {
+      "epoch": 0.1304,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.766290322580646e-05,
+      "loss": 0.1927,
+      "step": 8150
+    },
+    {
+      "epoch": 0.130416,
+      "grad_norm": 1.2890625,
+      "learning_rate": 8.766129032258065e-05,
+      "loss": 0.2025,
+      "step": 8151
+    },
+    {
+      "epoch": 0.130432,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.765967741935485e-05,
+      "loss": 0.1388,
+      "step": 8152
+    },
+    {
+      "epoch": 0.130448,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.765806451612903e-05,
+      "loss": 0.2194,
+      "step": 8153
+    },
+    {
+      "epoch": 0.130464,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.765645161290323e-05,
+      "loss": 0.1679,
+      "step": 8154
+    },
+    {
+      "epoch": 0.13048,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.765483870967742e-05,
+      "loss": 0.1856,
+      "step": 8155
+    },
+    {
+      "epoch": 0.130496,
+      "grad_norm": 1.3046875,
+      "learning_rate": 8.765322580645162e-05,
+      "loss": 0.1813,
+      "step": 8156
+    },
+    {
+      "epoch": 0.130512,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.765161290322582e-05,
+      "loss": 0.2055,
+      "step": 8157
+    },
+    {
+      "epoch": 0.130528,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.765e-05,
+      "loss": 0.2086,
+      "step": 8158
+    },
+    {
+      "epoch": 0.130544,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.76483870967742e-05,
+      "loss": 0.2269,
+      "step": 8159
+    },
+    {
+      "epoch": 0.13056,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.764677419354839e-05,
+      "loss": 0.1449,
+      "step": 8160
+    },
+    {
+      "epoch": 0.130576,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.764516129032259e-05,
+      "loss": 0.1275,
+      "step": 8161
+    },
+    {
+      "epoch": 0.130592,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.764354838709677e-05,
+      "loss": 0.1853,
+      "step": 8162
+    },
+    {
+      "epoch": 0.130608,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.764193548387097e-05,
+      "loss": 0.1553,
+      "step": 8163
+    },
+    {
+      "epoch": 0.130624,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.764032258064516e-05,
+      "loss": 0.2091,
+      "step": 8164
+    },
+    {
+      "epoch": 0.13064,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.763870967741936e-05,
+      "loss": 0.1558,
+      "step": 8165
+    },
+    {
+      "epoch": 0.130656,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.763709677419355e-05,
+      "loss": 0.178,
+      "step": 8166
+    },
+    {
+      "epoch": 0.130672,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.763548387096775e-05,
+      "loss": 0.1456,
+      "step": 8167
+    },
+    {
+      "epoch": 0.130688,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.763387096774193e-05,
+      "loss": 0.222,
+      "step": 8168
+    },
+    {
+      "epoch": 0.130704,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.763225806451613e-05,
+      "loss": 0.1887,
+      "step": 8169
+    },
+    {
+      "epoch": 0.13072,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.763064516129033e-05,
+      "loss": 0.1531,
+      "step": 8170
+    },
+    {
+      "epoch": 0.130736,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.762903225806453e-05,
+      "loss": 0.1703,
+      "step": 8171
+    },
+    {
+      "epoch": 0.130752,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.762741935483872e-05,
+      "loss": 0.1712,
+      "step": 8172
+    },
+    {
+      "epoch": 0.130768,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.762580645161292e-05,
+      "loss": 0.1894,
+      "step": 8173
+    },
+    {
+      "epoch": 0.130784,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.76241935483871e-05,
+      "loss": 0.173,
+      "step": 8174
+    },
+    {
+      "epoch": 0.1308,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.762258064516129e-05,
+      "loss": 0.1886,
+      "step": 8175
+    },
+    {
+      "epoch": 0.130816,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.762096774193549e-05,
+      "loss": 0.1756,
+      "step": 8176
+    },
+    {
+      "epoch": 0.130832,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.761935483870967e-05,
+      "loss": 0.1932,
+      "step": 8177
+    },
+    {
+      "epoch": 0.130848,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.761774193548387e-05,
+      "loss": 0.2184,
+      "step": 8178
+    },
+    {
+      "epoch": 0.130864,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.761612903225806e-05,
+      "loss": 0.2252,
+      "step": 8179
+    },
+    {
+      "epoch": 0.13088,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.761451612903226e-05,
+      "loss": 0.1837,
+      "step": 8180
+    },
+    {
+      "epoch": 0.130896,
+      "grad_norm": 1.25,
+      "learning_rate": 8.761290322580646e-05,
+      "loss": 0.1555,
+      "step": 8181
+    },
+    {
+      "epoch": 0.130912,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.761129032258066e-05,
+      "loss": 0.1809,
+      "step": 8182
+    },
+    {
+      "epoch": 0.130928,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.760967741935484e-05,
+      "loss": 0.143,
+      "step": 8183
+    },
+    {
+      "epoch": 0.130944,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.760806451612904e-05,
+      "loss": 0.1922,
+      "step": 8184
+    },
+    {
+      "epoch": 0.13096,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.760645161290323e-05,
+      "loss": 0.1547,
+      "step": 8185
+    },
+    {
+      "epoch": 0.130976,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.760483870967743e-05,
+      "loss": 0.1884,
+      "step": 8186
+    },
+    {
+      "epoch": 0.130992,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.760322580645162e-05,
+      "loss": 0.1722,
+      "step": 8187
+    },
+    {
+      "epoch": 0.131008,
+      "grad_norm": 1.125,
+      "learning_rate": 8.760161290322582e-05,
+      "loss": 0.1569,
+      "step": 8188
+    },
+    {
+      "epoch": 0.131024,
+      "grad_norm": 1.125,
+      "learning_rate": 8.76e-05,
+      "loss": 0.1892,
+      "step": 8189
+    },
+    {
+      "epoch": 0.13104,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.759838709677419e-05,
+      "loss": 0.1714,
+      "step": 8190
+    },
+    {
+      "epoch": 0.131056,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.759677419354839e-05,
+      "loss": 0.186,
+      "step": 8191
+    },
+    {
+      "epoch": 0.131072,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.759516129032259e-05,
+      "loss": 0.1484,
+      "step": 8192
+    },
+    {
+      "epoch": 0.131088,
+      "grad_norm": 1.2734375,
+      "learning_rate": 8.759354838709679e-05,
+      "loss": 0.1922,
+      "step": 8193
+    },
+    {
+      "epoch": 0.131104,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.759193548387097e-05,
+      "loss": 0.2156,
+      "step": 8194
+    },
+    {
+      "epoch": 0.13112,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.759032258064517e-05,
+      "loss": 0.1543,
+      "step": 8195
+    },
+    {
+      "epoch": 0.131136,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.758870967741936e-05,
+      "loss": 0.2074,
+      "step": 8196
+    },
+    {
+      "epoch": 0.131152,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.758709677419356e-05,
+      "loss": 0.2473,
+      "step": 8197
+    },
+    {
+      "epoch": 0.131168,
+      "grad_norm": 1.390625,
+      "learning_rate": 8.758548387096774e-05,
+      "loss": 0.2091,
+      "step": 8198
+    },
+    {
+      "epoch": 0.131184,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.758387096774194e-05,
+      "loss": 0.1534,
+      "step": 8199
+    },
+    {
+      "epoch": 0.1312,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.758225806451613e-05,
+      "loss": 0.1547,
+      "step": 8200
+    },
+    {
+      "epoch": 0.131216,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.758064516129033e-05,
+      "loss": 0.1585,
+      "step": 8201
+    },
+    {
+      "epoch": 0.131232,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.757903225806452e-05,
+      "loss": 0.2075,
+      "step": 8202
+    },
+    {
+      "epoch": 0.131248,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.757741935483872e-05,
+      "loss": 0.1585,
+      "step": 8203
+    },
+    {
+      "epoch": 0.131264,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.75758064516129e-05,
+      "loss": 0.1648,
+      "step": 8204
+    },
+    {
+      "epoch": 0.13128,
+      "grad_norm": 0.75,
+      "learning_rate": 8.75741935483871e-05,
+      "loss": 0.1841,
+      "step": 8205
+    },
+    {
+      "epoch": 0.131296,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.75725806451613e-05,
+      "loss": 0.1681,
+      "step": 8206
+    },
+    {
+      "epoch": 0.131312,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.757096774193549e-05,
+      "loss": 0.1688,
+      "step": 8207
+    },
+    {
+      "epoch": 0.131328,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.756935483870969e-05,
+      "loss": 0.1749,
+      "step": 8208
+    },
+    {
+      "epoch": 0.131344,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.756774193548387e-05,
+      "loss": 0.1712,
+      "step": 8209
+    },
+    {
+      "epoch": 0.13136,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.756612903225807e-05,
+      "loss": 0.1356,
+      "step": 8210
+    },
+    {
+      "epoch": 0.131376,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.756451612903226e-05,
+      "loss": 0.1598,
+      "step": 8211
+    },
+    {
+      "epoch": 0.131392,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.756290322580646e-05,
+      "loss": 0.1373,
+      "step": 8212
+    },
+    {
+      "epoch": 0.131408,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.756129032258064e-05,
+      "loss": 0.1721,
+      "step": 8213
+    },
+    {
+      "epoch": 0.131424,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.755967741935484e-05,
+      "loss": 0.1806,
+      "step": 8214
+    },
+    {
+      "epoch": 0.13144,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.755806451612903e-05,
+      "loss": 0.1545,
+      "step": 8215
+    },
+    {
+      "epoch": 0.131456,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.755645161290323e-05,
+      "loss": 0.1858,
+      "step": 8216
+    },
+    {
+      "epoch": 0.131472,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.755483870967743e-05,
+      "loss": 0.1536,
+      "step": 8217
+    },
+    {
+      "epoch": 0.131488,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.755322580645163e-05,
+      "loss": 0.1493,
+      "step": 8218
+    },
+    {
+      "epoch": 0.131504,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.755161290322581e-05,
+      "loss": 0.1799,
+      "step": 8219
+    },
+    {
+      "epoch": 0.13152,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.755e-05,
+      "loss": 0.1616,
+      "step": 8220
+    },
+    {
+      "epoch": 0.131536,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.75483870967742e-05,
+      "loss": 0.1571,
+      "step": 8221
+    },
+    {
+      "epoch": 0.131552,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.754677419354839e-05,
+      "loss": 0.167,
+      "step": 8222
+    },
+    {
+      "epoch": 0.131568,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.754516129032259e-05,
+      "loss": 0.1456,
+      "step": 8223
+    },
+    {
+      "epoch": 0.131584,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.754354838709677e-05,
+      "loss": 0.1997,
+      "step": 8224
+    },
+    {
+      "epoch": 0.1316,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.754193548387097e-05,
+      "loss": 0.1814,
+      "step": 8225
+    },
+    {
+      "epoch": 0.131616,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.754032258064516e-05,
+      "loss": 0.1709,
+      "step": 8226
+    },
+    {
+      "epoch": 0.131632,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.753870967741936e-05,
+      "loss": 0.167,
+      "step": 8227
+    },
+    {
+      "epoch": 0.131648,
+      "grad_norm": 1.3671875,
+      "learning_rate": 8.753709677419354e-05,
+      "loss": 0.1867,
+      "step": 8228
+    },
+    {
+      "epoch": 0.131664,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.753548387096774e-05,
+      "loss": 0.1609,
+      "step": 8229
+    },
+    {
+      "epoch": 0.13168,
+      "grad_norm": 1.5546875,
+      "learning_rate": 8.753387096774194e-05,
+      "loss": 0.1685,
+      "step": 8230
+    },
+    {
+      "epoch": 0.131696,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.753225806451614e-05,
+      "loss": 0.1399,
+      "step": 8231
+    },
+    {
+      "epoch": 0.131712,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.753064516129033e-05,
+      "loss": 0.1885,
+      "step": 8232
+    },
+    {
+      "epoch": 0.131728,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.752903225806453e-05,
+      "loss": 0.1869,
+      "step": 8233
+    },
+    {
+      "epoch": 0.131744,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.752741935483871e-05,
+      "loss": 0.1798,
+      "step": 8234
+    },
+    {
+      "epoch": 0.13176,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.752580645161291e-05,
+      "loss": 0.1894,
+      "step": 8235
+    },
+    {
+      "epoch": 0.131776,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.75241935483871e-05,
+      "loss": 0.1459,
+      "step": 8236
+    },
+    {
+      "epoch": 0.131792,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.752258064516129e-05,
+      "loss": 0.1525,
+      "step": 8237
+    },
+    {
+      "epoch": 0.131808,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.752096774193549e-05,
+      "loss": 0.1737,
+      "step": 8238
+    },
+    {
+      "epoch": 0.131824,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.751935483870967e-05,
+      "loss": 0.1478,
+      "step": 8239
+    },
+    {
+      "epoch": 0.13184,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.751774193548387e-05,
+      "loss": 0.1425,
+      "step": 8240
+    },
+    {
+      "epoch": 0.131856,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.751612903225807e-05,
+      "loss": 0.1615,
+      "step": 8241
+    },
+    {
+      "epoch": 0.131872,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.751451612903227e-05,
+      "loss": 0.1828,
+      "step": 8242
+    },
+    {
+      "epoch": 0.131888,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.751290322580646e-05,
+      "loss": 0.1809,
+      "step": 8243
+    },
+    {
+      "epoch": 0.131904,
+      "grad_norm": 1.484375,
+      "learning_rate": 8.751129032258066e-05,
+      "loss": 0.233,
+      "step": 8244
+    },
+    {
+      "epoch": 0.13192,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.750967741935484e-05,
+      "loss": 0.1935,
+      "step": 8245
+    },
+    {
+      "epoch": 0.131936,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.750806451612904e-05,
+      "loss": 0.1693,
+      "step": 8246
+    },
+    {
+      "epoch": 0.131952,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.750645161290323e-05,
+      "loss": 0.1727,
+      "step": 8247
+    },
+    {
+      "epoch": 0.131968,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.750483870967743e-05,
+      "loss": 0.2091,
+      "step": 8248
+    },
+    {
+      "epoch": 0.131984,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.750322580645161e-05,
+      "loss": 0.1789,
+      "step": 8249
+    },
+    {
+      "epoch": 0.132,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.750161290322581e-05,
+      "loss": 0.1746,
+      "step": 8250
+    },
+    {
+      "epoch": 0.132016,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.75e-05,
+      "loss": 0.2042,
+      "step": 8251
+    },
+    {
+      "epoch": 0.132032,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.74983870967742e-05,
+      "loss": 0.1751,
+      "step": 8252
+    },
+    {
+      "epoch": 0.132048,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.74967741935484e-05,
+      "loss": 0.1811,
+      "step": 8253
+    },
+    {
+      "epoch": 0.132064,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.749516129032258e-05,
+      "loss": 0.1904,
+      "step": 8254
+    },
+    {
+      "epoch": 0.13208,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.749354838709678e-05,
+      "loss": 0.1495,
+      "step": 8255
+    },
+    {
+      "epoch": 0.132096,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.749193548387097e-05,
+      "loss": 0.15,
+      "step": 8256
+    },
+    {
+      "epoch": 0.132112,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.749032258064517e-05,
+      "loss": 0.1957,
+      "step": 8257
+    },
+    {
+      "epoch": 0.132128,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.748870967741936e-05,
+      "loss": 0.1752,
+      "step": 8258
+    },
+    {
+      "epoch": 0.132144,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.748709677419356e-05,
+      "loss": 0.2227,
+      "step": 8259
+    },
+    {
+      "epoch": 0.13216,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.748548387096774e-05,
+      "loss": 0.1666,
+      "step": 8260
+    },
+    {
+      "epoch": 0.132176,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.748387096774194e-05,
+      "loss": 0.2036,
+      "step": 8261
+    },
+    {
+      "epoch": 0.132192,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.748225806451613e-05,
+      "loss": 0.1534,
+      "step": 8262
+    },
+    {
+      "epoch": 0.132208,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.748064516129033e-05,
+      "loss": 0.1615,
+      "step": 8263
+    },
+    {
+      "epoch": 0.132224,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.747903225806451e-05,
+      "loss": 0.2067,
+      "step": 8264
+    },
+    {
+      "epoch": 0.13224,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.747741935483871e-05,
+      "loss": 0.137,
+      "step": 8265
+    },
+    {
+      "epoch": 0.132256,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.747580645161291e-05,
+      "loss": 0.1707,
+      "step": 8266
+    },
+    {
+      "epoch": 0.132272,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.74741935483871e-05,
+      "loss": 0.1692,
+      "step": 8267
+    },
+    {
+      "epoch": 0.132288,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.74725806451613e-05,
+      "loss": 0.1953,
+      "step": 8268
+    },
+    {
+      "epoch": 0.132304,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.747096774193548e-05,
+      "loss": 0.2016,
+      "step": 8269
+    },
+    {
+      "epoch": 0.13232,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.746935483870968e-05,
+      "loss": 0.1848,
+      "step": 8270
+    },
+    {
+      "epoch": 0.132336,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.746774193548387e-05,
+      "loss": 0.1406,
+      "step": 8271
+    },
+    {
+      "epoch": 0.132352,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.746612903225807e-05,
+      "loss": 0.2084,
+      "step": 8272
+    },
+    {
+      "epoch": 0.132368,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.746451612903226e-05,
+      "loss": 0.1603,
+      "step": 8273
+    },
+    {
+      "epoch": 0.132384,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.746290322580646e-05,
+      "loss": 0.2118,
+      "step": 8274
+    },
+    {
+      "epoch": 0.1324,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.746129032258064e-05,
+      "loss": 0.1514,
+      "step": 8275
+    },
+    {
+      "epoch": 0.132416,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.745967741935484e-05,
+      "loss": 0.1342,
+      "step": 8276
+    },
+    {
+      "epoch": 0.132432,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.745806451612904e-05,
+      "loss": 0.1578,
+      "step": 8277
+    },
+    {
+      "epoch": 0.132448,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.745645161290324e-05,
+      "loss": 0.182,
+      "step": 8278
+    },
+    {
+      "epoch": 0.132464,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.745483870967743e-05,
+      "loss": 0.1965,
+      "step": 8279
+    },
+    {
+      "epoch": 0.13248,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.745322580645163e-05,
+      "loss": 0.204,
+      "step": 8280
+    },
+    {
+      "epoch": 0.132496,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.745161290322581e-05,
+      "loss": 0.1476,
+      "step": 8281
+    },
+    {
+      "epoch": 0.132512,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.745000000000001e-05,
+      "loss": 0.1524,
+      "step": 8282
+    },
+    {
+      "epoch": 0.132528,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.74483870967742e-05,
+      "loss": 0.1839,
+      "step": 8283
+    },
+    {
+      "epoch": 0.132544,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.744677419354838e-05,
+      "loss": 0.2016,
+      "step": 8284
+    },
+    {
+      "epoch": 0.13256,
+      "grad_norm": 1.84375,
+      "learning_rate": 8.744516129032258e-05,
+      "loss": 0.2009,
+      "step": 8285
+    },
+    {
+      "epoch": 0.132576,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.744354838709677e-05,
+      "loss": 0.1841,
+      "step": 8286
+    },
+    {
+      "epoch": 0.132592,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.744193548387097e-05,
+      "loss": 0.1771,
+      "step": 8287
+    },
+    {
+      "epoch": 0.132608,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.744032258064517e-05,
+      "loss": 0.1708,
+      "step": 8288
+    },
+    {
+      "epoch": 0.132624,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.743870967741937e-05,
+      "loss": 0.1409,
+      "step": 8289
+    },
+    {
+      "epoch": 0.13264,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.743709677419355e-05,
+      "loss": 0.1722,
+      "step": 8290
+    },
+    {
+      "epoch": 0.132656,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.743548387096775e-05,
+      "loss": 0.1817,
+      "step": 8291
+    },
+    {
+      "epoch": 0.132672,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.743387096774194e-05,
+      "loss": 0.1675,
+      "step": 8292
+    },
+    {
+      "epoch": 0.132688,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.743225806451614e-05,
+      "loss": 0.172,
+      "step": 8293
+    },
+    {
+      "epoch": 0.132704,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.743064516129033e-05,
+      "loss": 0.1691,
+      "step": 8294
+    },
+    {
+      "epoch": 0.13272,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.742903225806453e-05,
+      "loss": 0.1771,
+      "step": 8295
+    },
+    {
+      "epoch": 0.132736,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.742741935483871e-05,
+      "loss": 0.1764,
+      "step": 8296
+    },
+    {
+      "epoch": 0.132752,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.742580645161291e-05,
+      "loss": 0.1808,
+      "step": 8297
+    },
+    {
+      "epoch": 0.132768,
+      "grad_norm": 0.625,
+      "learning_rate": 8.74241935483871e-05,
+      "loss": 0.1869,
+      "step": 8298
+    },
+    {
+      "epoch": 0.132784,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.742258064516128e-05,
+      "loss": 0.1449,
+      "step": 8299
+    },
+    {
+      "epoch": 0.1328,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.742096774193548e-05,
+      "loss": 0.1725,
+      "step": 8300
+    },
+    {
+      "epoch": 0.132816,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.741935483870968e-05,
+      "loss": 0.1733,
+      "step": 8301
+    },
+    {
+      "epoch": 0.132832,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.741774193548388e-05,
+      "loss": 0.1948,
+      "step": 8302
+    },
+    {
+      "epoch": 0.132848,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.741612903225807e-05,
+      "loss": 0.1757,
+      "step": 8303
+    },
+    {
+      "epoch": 0.132864,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.741451612903227e-05,
+      "loss": 0.2057,
+      "step": 8304
+    },
+    {
+      "epoch": 0.13288,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.741290322580645e-05,
+      "loss": 0.1886,
+      "step": 8305
+    },
+    {
+      "epoch": 0.132896,
+      "grad_norm": 1.3828125,
+      "learning_rate": 8.741129032258065e-05,
+      "loss": 0.1875,
+      "step": 8306
+    },
+    {
+      "epoch": 0.132912,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.740967741935484e-05,
+      "loss": 0.1887,
+      "step": 8307
+    },
+    {
+      "epoch": 0.132928,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.740806451612904e-05,
+      "loss": 0.1903,
+      "step": 8308
+    },
+    {
+      "epoch": 0.132944,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.740645161290323e-05,
+      "loss": 0.1844,
+      "step": 8309
+    },
+    {
+      "epoch": 0.13296,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.740483870967742e-05,
+      "loss": 0.1617,
+      "step": 8310
+    },
+    {
+      "epoch": 0.132976,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.740322580645161e-05,
+      "loss": 0.193,
+      "step": 8311
+    },
+    {
+      "epoch": 0.132992,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.740161290322581e-05,
+      "loss": 0.1625,
+      "step": 8312
+    },
+    {
+      "epoch": 0.133008,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.740000000000001e-05,
+      "loss": 0.2024,
+      "step": 8313
+    },
+    {
+      "epoch": 0.133024,
+      "grad_norm": 1.5078125,
+      "learning_rate": 8.73983870967742e-05,
+      "loss": 0.1815,
+      "step": 8314
+    },
+    {
+      "epoch": 0.13304,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.73967741935484e-05,
+      "loss": 0.2176,
+      "step": 8315
+    },
+    {
+      "epoch": 0.133056,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.739516129032258e-05,
+      "loss": 0.1626,
+      "step": 8316
+    },
+    {
+      "epoch": 0.133072,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.739354838709678e-05,
+      "loss": 0.1697,
+      "step": 8317
+    },
+    {
+      "epoch": 0.133088,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.739193548387097e-05,
+      "loss": 0.1567,
+      "step": 8318
+    },
+    {
+      "epoch": 0.133104,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.739032258064517e-05,
+      "loss": 0.1621,
+      "step": 8319
+    },
+    {
+      "epoch": 0.13312,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.738870967741935e-05,
+      "loss": 0.1672,
+      "step": 8320
+    },
+    {
+      "epoch": 0.133136,
+      "grad_norm": 1.2734375,
+      "learning_rate": 8.738709677419355e-05,
+      "loss": 0.1962,
+      "step": 8321
+    },
+    {
+      "epoch": 0.133152,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.738548387096774e-05,
+      "loss": 0.1633,
+      "step": 8322
+    },
+    {
+      "epoch": 0.133168,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.738387096774194e-05,
+      "loss": 0.1811,
+      "step": 8323
+    },
+    {
+      "epoch": 0.133184,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.738225806451612e-05,
+      "loss": 0.199,
+      "step": 8324
+    },
+    {
+      "epoch": 0.1332,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.738064516129032e-05,
+      "loss": 0.1945,
+      "step": 8325
+    },
+    {
+      "epoch": 0.133216,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.737903225806452e-05,
+      "loss": 0.1699,
+      "step": 8326
+    },
+    {
+      "epoch": 0.133232,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.737741935483872e-05,
+      "loss": 0.1928,
+      "step": 8327
+    },
+    {
+      "epoch": 0.133248,
+      "grad_norm": 1.125,
+      "learning_rate": 8.737580645161291e-05,
+      "loss": 0.199,
+      "step": 8328
+    },
+    {
+      "epoch": 0.133264,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.73741935483871e-05,
+      "loss": 0.158,
+      "step": 8329
+    },
+    {
+      "epoch": 0.13328,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.73725806451613e-05,
+      "loss": 0.1529,
+      "step": 8330
+    },
+    {
+      "epoch": 0.133296,
+      "grad_norm": 1.4296875,
+      "learning_rate": 8.737096774193548e-05,
+      "loss": 0.1782,
+      "step": 8331
+    },
+    {
+      "epoch": 0.133312,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.736935483870968e-05,
+      "loss": 0.167,
+      "step": 8332
+    },
+    {
+      "epoch": 0.133328,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.736774193548387e-05,
+      "loss": 0.1696,
+      "step": 8333
+    },
+    {
+      "epoch": 0.133344,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.736612903225807e-05,
+      "loss": 0.1617,
+      "step": 8334
+    },
+    {
+      "epoch": 0.13336,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.736451612903225e-05,
+      "loss": 0.1903,
+      "step": 8335
+    },
+    {
+      "epoch": 0.133376,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.736290322580645e-05,
+      "loss": 0.151,
+      "step": 8336
+    },
+    {
+      "epoch": 0.133392,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.736129032258065e-05,
+      "loss": 0.1871,
+      "step": 8337
+    },
+    {
+      "epoch": 0.133408,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.735967741935485e-05,
+      "loss": 0.2013,
+      "step": 8338
+    },
+    {
+      "epoch": 0.133424,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.735806451612904e-05,
+      "loss": 0.1763,
+      "step": 8339
+    },
+    {
+      "epoch": 0.13344,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.735645161290324e-05,
+      "loss": 0.1727,
+      "step": 8340
+    },
+    {
+      "epoch": 0.133456,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.735483870967742e-05,
+      "loss": 0.1711,
+      "step": 8341
+    },
+    {
+      "epoch": 0.133472,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.735322580645162e-05,
+      "loss": 0.188,
+      "step": 8342
+    },
+    {
+      "epoch": 0.133488,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.735161290322581e-05,
+      "loss": 0.1651,
+      "step": 8343
+    },
+    {
+      "epoch": 0.133504,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.735000000000001e-05,
+      "loss": 0.155,
+      "step": 8344
+    },
+    {
+      "epoch": 0.13352,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.73483870967742e-05,
+      "loss": 0.1914,
+      "step": 8345
+    },
+    {
+      "epoch": 0.133536,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.734677419354838e-05,
+      "loss": 0.1748,
+      "step": 8346
+    },
+    {
+      "epoch": 0.133552,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.734516129032258e-05,
+      "loss": 0.1939,
+      "step": 8347
+    },
+    {
+      "epoch": 0.133568,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.734354838709678e-05,
+      "loss": 0.1902,
+      "step": 8348
+    },
+    {
+      "epoch": 0.133584,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.734193548387098e-05,
+      "loss": 0.1787,
+      "step": 8349
+    },
+    {
+      "epoch": 0.1336,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.734032258064517e-05,
+      "loss": 0.1628,
+      "step": 8350
+    },
+    {
+      "epoch": 0.133616,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.733870967741937e-05,
+      "loss": 0.164,
+      "step": 8351
+    },
+    {
+      "epoch": 0.133632,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.733709677419355e-05,
+      "loss": 0.1756,
+      "step": 8352
+    },
+    {
+      "epoch": 0.133648,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.733548387096775e-05,
+      "loss": 0.1618,
+      "step": 8353
+    },
+    {
+      "epoch": 0.133664,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.733387096774194e-05,
+      "loss": 0.2049,
+      "step": 8354
+    },
+    {
+      "epoch": 0.13368,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.733225806451614e-05,
+      "loss": 0.1724,
+      "step": 8355
+    },
+    {
+      "epoch": 0.133696,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.733064516129032e-05,
+      "loss": 0.1477,
+      "step": 8356
+    },
+    {
+      "epoch": 0.133712,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.732903225806452e-05,
+      "loss": 0.1908,
+      "step": 8357
+    },
+    {
+      "epoch": 0.133728,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.732741935483871e-05,
+      "loss": 0.1786,
+      "step": 8358
+    },
+    {
+      "epoch": 0.133744,
+      "grad_norm": 1.3515625,
+      "learning_rate": 8.732580645161291e-05,
+      "loss": 0.1821,
+      "step": 8359
+    },
+    {
+      "epoch": 0.13376,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.73241935483871e-05,
+      "loss": 0.183,
+      "step": 8360
+    },
+    {
+      "epoch": 0.133776,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.73225806451613e-05,
+      "loss": 0.1429,
+      "step": 8361
+    },
+    {
+      "epoch": 0.133792,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.73209677419355e-05,
+      "loss": 0.1413,
+      "step": 8362
+    },
+    {
+      "epoch": 0.133808,
+      "grad_norm": 0.625,
+      "learning_rate": 8.731935483870968e-05,
+      "loss": 0.17,
+      "step": 8363
+    },
+    {
+      "epoch": 0.133824,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.731774193548388e-05,
+      "loss": 0.1962,
+      "step": 8364
+    },
+    {
+      "epoch": 0.13384,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.731612903225807e-05,
+      "loss": 0.1923,
+      "step": 8365
+    },
+    {
+      "epoch": 0.133856,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.731451612903227e-05,
+      "loss": 0.1636,
+      "step": 8366
+    },
+    {
+      "epoch": 0.133872,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.731290322580645e-05,
+      "loss": 0.2533,
+      "step": 8367
+    },
+    {
+      "epoch": 0.133888,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.731129032258065e-05,
+      "loss": 0.1825,
+      "step": 8368
+    },
+    {
+      "epoch": 0.133904,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.730967741935484e-05,
+      "loss": 0.1705,
+      "step": 8369
+    },
+    {
+      "epoch": 0.13392,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.730806451612904e-05,
+      "loss": 0.2069,
+      "step": 8370
+    },
+    {
+      "epoch": 0.133936,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.730645161290322e-05,
+      "loss": 0.1881,
+      "step": 8371
+    },
+    {
+      "epoch": 0.133952,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.730483870967742e-05,
+      "loss": 0.1782,
+      "step": 8372
+    },
+    {
+      "epoch": 0.133968,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.730322580645162e-05,
+      "loss": 0.1758,
+      "step": 8373
+    },
+    {
+      "epoch": 0.133984,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.730161290322582e-05,
+      "loss": 0.161,
+      "step": 8374
+    },
+    {
+      "epoch": 0.134,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.730000000000001e-05,
+      "loss": 0.212,
+      "step": 8375
+    },
+    {
+      "epoch": 0.134016,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.72983870967742e-05,
+      "loss": 0.1864,
+      "step": 8376
+    },
+    {
+      "epoch": 0.134032,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.72967741935484e-05,
+      "loss": 0.2079,
+      "step": 8377
+    },
+    {
+      "epoch": 0.134048,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.729516129032258e-05,
+      "loss": 0.1711,
+      "step": 8378
+    },
+    {
+      "epoch": 0.134064,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.729354838709678e-05,
+      "loss": 0.1988,
+      "step": 8379
+    },
+    {
+      "epoch": 0.13408,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.729193548387097e-05,
+      "loss": 0.1771,
+      "step": 8380
+    },
+    {
+      "epoch": 0.134096,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.729032258064516e-05,
+      "loss": 0.1686,
+      "step": 8381
+    },
+    {
+      "epoch": 0.134112,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.728870967741935e-05,
+      "loss": 0.2098,
+      "step": 8382
+    },
+    {
+      "epoch": 0.134128,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.728709677419355e-05,
+      "loss": 0.1829,
+      "step": 8383
+    },
+    {
+      "epoch": 0.134144,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.728548387096775e-05,
+      "loss": 0.1533,
+      "step": 8384
+    },
+    {
+      "epoch": 0.13416,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.728387096774194e-05,
+      "loss": 0.1936,
+      "step": 8385
+    },
+    {
+      "epoch": 0.134176,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.728225806451614e-05,
+      "loss": 0.1638,
+      "step": 8386
+    },
+    {
+      "epoch": 0.134192,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.728064516129034e-05,
+      "loss": 0.1573,
+      "step": 8387
+    },
+    {
+      "epoch": 0.134208,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.727903225806452e-05,
+      "loss": 0.2008,
+      "step": 8388
+    },
+    {
+      "epoch": 0.134224,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.727741935483872e-05,
+      "loss": 0.217,
+      "step": 8389
+    },
+    {
+      "epoch": 0.13424,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.727580645161291e-05,
+      "loss": 0.1505,
+      "step": 8390
+    },
+    {
+      "epoch": 0.134256,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.727419354838711e-05,
+      "loss": 0.1774,
+      "step": 8391
+    },
+    {
+      "epoch": 0.134272,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.727258064516129e-05,
+      "loss": 0.1793,
+      "step": 8392
+    },
+    {
+      "epoch": 0.134288,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.727096774193548e-05,
+      "loss": 0.2101,
+      "step": 8393
+    },
+    {
+      "epoch": 0.134304,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.726935483870968e-05,
+      "loss": 0.2338,
+      "step": 8394
+    },
+    {
+      "epoch": 0.13432,
+      "grad_norm": 1.3984375,
+      "learning_rate": 8.726774193548386e-05,
+      "loss": 0.165,
+      "step": 8395
+    },
+    {
+      "epoch": 0.134336,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.726612903225806e-05,
+      "loss": 0.1881,
+      "step": 8396
+    },
+    {
+      "epoch": 0.134352,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.726451612903226e-05,
+      "loss": 0.1476,
+      "step": 8397
+    },
+    {
+      "epoch": 0.134368,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.726290322580646e-05,
+      "loss": 0.1398,
+      "step": 8398
+    },
+    {
+      "epoch": 0.134384,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.726129032258065e-05,
+      "loss": 0.2025,
+      "step": 8399
+    },
+    {
+      "epoch": 0.1344,
+      "grad_norm": 1.0,
+      "learning_rate": 8.725967741935485e-05,
+      "loss": 0.1639,
+      "step": 8400
+    },
+    {
+      "epoch": 0.134416,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.725806451612904e-05,
+      "loss": 0.1466,
+      "step": 8401
+    },
+    {
+      "epoch": 0.134432,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.725645161290324e-05,
+      "loss": 0.1807,
+      "step": 8402
+    },
+    {
+      "epoch": 0.134448,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.725483870967742e-05,
+      "loss": 0.2408,
+      "step": 8403
+    },
+    {
+      "epoch": 0.134464,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.725322580645162e-05,
+      "loss": 0.1691,
+      "step": 8404
+    },
+    {
+      "epoch": 0.13448,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.725161290322581e-05,
+      "loss": 0.1733,
+      "step": 8405
+    },
+    {
+      "epoch": 0.134496,
+      "grad_norm": 1.375,
+      "learning_rate": 8.725e-05,
+      "loss": 0.138,
+      "step": 8406
+    },
+    {
+      "epoch": 0.134512,
+      "grad_norm": 1.3828125,
+      "learning_rate": 8.724838709677419e-05,
+      "loss": 0.2173,
+      "step": 8407
+    },
+    {
+      "epoch": 0.134528,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.724677419354839e-05,
+      "loss": 0.1856,
+      "step": 8408
+    },
+    {
+      "epoch": 0.134544,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.724516129032259e-05,
+      "loss": 0.1683,
+      "step": 8409
+    },
+    {
+      "epoch": 0.13456,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.724354838709678e-05,
+      "loss": 0.176,
+      "step": 8410
+    },
+    {
+      "epoch": 0.134576,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.724193548387098e-05,
+      "loss": 0.1888,
+      "step": 8411
+    },
+    {
+      "epoch": 0.134592,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.724032258064516e-05,
+      "loss": 0.1995,
+      "step": 8412
+    },
+    {
+      "epoch": 0.134608,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.723870967741936e-05,
+      "loss": 0.1232,
+      "step": 8413
+    },
+    {
+      "epoch": 0.134624,
+      "grad_norm": 1.34375,
+      "learning_rate": 8.723709677419355e-05,
+      "loss": 0.176,
+      "step": 8414
+    },
+    {
+      "epoch": 0.13464,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.723548387096775e-05,
+      "loss": 0.1755,
+      "step": 8415
+    },
+    {
+      "epoch": 0.134656,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.723387096774194e-05,
+      "loss": 0.1641,
+      "step": 8416
+    },
+    {
+      "epoch": 0.134672,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.723225806451613e-05,
+      "loss": 0.1786,
+      "step": 8417
+    },
+    {
+      "epoch": 0.134688,
+      "grad_norm": 0.75,
+      "learning_rate": 8.723064516129032e-05,
+      "loss": 0.1801,
+      "step": 8418
+    },
+    {
+      "epoch": 0.134704,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.722903225806452e-05,
+      "loss": 0.1813,
+      "step": 8419
+    },
+    {
+      "epoch": 0.13472,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.72274193548387e-05,
+      "loss": 0.1978,
+      "step": 8420
+    },
+    {
+      "epoch": 0.134736,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.72258064516129e-05,
+      "loss": 0.2548,
+      "step": 8421
+    },
+    {
+      "epoch": 0.134752,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.72241935483871e-05,
+      "loss": 0.197,
+      "step": 8422
+    },
+    {
+      "epoch": 0.134768,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.722258064516129e-05,
+      "loss": 0.1482,
+      "step": 8423
+    },
+    {
+      "epoch": 0.134784,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.722096774193549e-05,
+      "loss": 0.1535,
+      "step": 8424
+    },
+    {
+      "epoch": 0.1348,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.721935483870968e-05,
+      "loss": 0.1667,
+      "step": 8425
+    },
+    {
+      "epoch": 0.134816,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.721774193548388e-05,
+      "loss": 0.139,
+      "step": 8426
+    },
+    {
+      "epoch": 0.134832,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.721612903225806e-05,
+      "loss": 0.1912,
+      "step": 8427
+    },
+    {
+      "epoch": 0.134848,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.721451612903226e-05,
+      "loss": 0.1537,
+      "step": 8428
+    },
+    {
+      "epoch": 0.134864,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.721290322580645e-05,
+      "loss": 0.163,
+      "step": 8429
+    },
+    {
+      "epoch": 0.13488,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.721129032258065e-05,
+      "loss": 0.2029,
+      "step": 8430
+    },
+    {
+      "epoch": 0.134896,
+      "grad_norm": 0.447265625,
+      "learning_rate": 8.720967741935483e-05,
+      "loss": 0.1482,
+      "step": 8431
+    },
+    {
+      "epoch": 0.134912,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.720806451612903e-05,
+      "loss": 0.2038,
+      "step": 8432
+    },
+    {
+      "epoch": 0.134928,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.720645161290323e-05,
+      "loss": 0.1368,
+      "step": 8433
+    },
+    {
+      "epoch": 0.134944,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.720483870967743e-05,
+      "loss": 0.1958,
+      "step": 8434
+    },
+    {
+      "epoch": 0.13496,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.720322580645162e-05,
+      "loss": 0.2156,
+      "step": 8435
+    },
+    {
+      "epoch": 0.134976,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.720161290322582e-05,
+      "loss": 0.1864,
+      "step": 8436
+    },
+    {
+      "epoch": 0.134992,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.72e-05,
+      "loss": 0.1651,
+      "step": 8437
+    },
+    {
+      "epoch": 0.135008,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.719838709677419e-05,
+      "loss": 0.1907,
+      "step": 8438
+    },
+    {
+      "epoch": 0.135024,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.719677419354839e-05,
+      "loss": 0.2103,
+      "step": 8439
+    },
+    {
+      "epoch": 0.13504,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.719516129032258e-05,
+      "loss": 0.2096,
+      "step": 8440
+    },
+    {
+      "epoch": 0.135056,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.719354838709678e-05,
+      "loss": 0.1856,
+      "step": 8441
+    },
+    {
+      "epoch": 0.135072,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.719193548387096e-05,
+      "loss": 0.1855,
+      "step": 8442
+    },
+    {
+      "epoch": 0.135088,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.719032258064516e-05,
+      "loss": 0.164,
+      "step": 8443
+    },
+    {
+      "epoch": 0.135104,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.718870967741936e-05,
+      "loss": 0.1321,
+      "step": 8444
+    },
+    {
+      "epoch": 0.13512,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.718709677419356e-05,
+      "loss": 0.1569,
+      "step": 8445
+    },
+    {
+      "epoch": 0.135136,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.718548387096775e-05,
+      "loss": 0.1452,
+      "step": 8446
+    },
+    {
+      "epoch": 0.135152,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.718387096774195e-05,
+      "loss": 0.1724,
+      "step": 8447
+    },
+    {
+      "epoch": 0.135168,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.718225806451613e-05,
+      "loss": 0.1672,
+      "step": 8448
+    },
+    {
+      "epoch": 0.135184,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.718064516129033e-05,
+      "loss": 0.1491,
+      "step": 8449
+    },
+    {
+      "epoch": 0.1352,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.717903225806452e-05,
+      "loss": 0.1695,
+      "step": 8450
+    },
+    {
+      "epoch": 0.135216,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.717741935483872e-05,
+      "loss": 0.1588,
+      "step": 8451
+    },
+    {
+      "epoch": 0.135232,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.71758064516129e-05,
+      "loss": 0.1741,
+      "step": 8452
+    },
+    {
+      "epoch": 0.135248,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.71741935483871e-05,
+      "loss": 0.1207,
+      "step": 8453
+    },
+    {
+      "epoch": 0.135264,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.717258064516129e-05,
+      "loss": 0.1599,
+      "step": 8454
+    },
+    {
+      "epoch": 0.13528,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.717096774193548e-05,
+      "loss": 0.1489,
+      "step": 8455
+    },
+    {
+      "epoch": 0.135296,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.716935483870968e-05,
+      "loss": 0.1952,
+      "step": 8456
+    },
+    {
+      "epoch": 0.135312,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.716774193548388e-05,
+      "loss": 0.1412,
+      "step": 8457
+    },
+    {
+      "epoch": 0.135328,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.716612903225808e-05,
+      "loss": 0.165,
+      "step": 8458
+    },
+    {
+      "epoch": 0.135344,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.716451612903226e-05,
+      "loss": 0.1987,
+      "step": 8459
+    },
+    {
+      "epoch": 0.13536,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.716290322580646e-05,
+      "loss": 0.1725,
+      "step": 8460
+    },
+    {
+      "epoch": 0.135376,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.716129032258065e-05,
+      "loss": 0.2352,
+      "step": 8461
+    },
+    {
+      "epoch": 0.135392,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.715967741935485e-05,
+      "loss": 0.1905,
+      "step": 8462
+    },
+    {
+      "epoch": 0.135408,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.715806451612903e-05,
+      "loss": 0.1735,
+      "step": 8463
+    },
+    {
+      "epoch": 0.135424,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.715645161290323e-05,
+      "loss": 0.1667,
+      "step": 8464
+    },
+    {
+      "epoch": 0.13544,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.715483870967742e-05,
+      "loss": 0.1364,
+      "step": 8465
+    },
+    {
+      "epoch": 0.135456,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.715322580645162e-05,
+      "loss": 0.1797,
+      "step": 8466
+    },
+    {
+      "epoch": 0.135472,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.71516129032258e-05,
+      "loss": 0.1572,
+      "step": 8467
+    },
+    {
+      "epoch": 0.135488,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.715e-05,
+      "loss": 0.1822,
+      "step": 8468
+    },
+    {
+      "epoch": 0.135504,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.71483870967742e-05,
+      "loss": 0.1469,
+      "step": 8469
+    },
+    {
+      "epoch": 0.13552,
+      "grad_norm": 1.28125,
+      "learning_rate": 8.714677419354839e-05,
+      "loss": 0.1773,
+      "step": 8470
+    },
+    {
+      "epoch": 0.135536,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.714516129032259e-05,
+      "loss": 0.1546,
+      "step": 8471
+    },
+    {
+      "epoch": 0.135552,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.714354838709678e-05,
+      "loss": 0.1653,
+      "step": 8472
+    },
+    {
+      "epoch": 0.135568,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.714193548387098e-05,
+      "loss": 0.1839,
+      "step": 8473
+    },
+    {
+      "epoch": 0.135584,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.714032258064516e-05,
+      "loss": 0.2019,
+      "step": 8474
+    },
+    {
+      "epoch": 0.1356,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.713870967741936e-05,
+      "loss": 0.1738,
+      "step": 8475
+    },
+    {
+      "epoch": 0.135616,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.713709677419355e-05,
+      "loss": 0.17,
+      "step": 8476
+    },
+    {
+      "epoch": 0.135632,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.713548387096775e-05,
+      "loss": 0.1729,
+      "step": 8477
+    },
+    {
+      "epoch": 0.135648,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.713387096774193e-05,
+      "loss": 0.1862,
+      "step": 8478
+    },
+    {
+      "epoch": 0.135664,
+      "grad_norm": 1.25,
+      "learning_rate": 8.713225806451613e-05,
+      "loss": 0.1715,
+      "step": 8479
+    },
+    {
+      "epoch": 0.13568,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.713064516129033e-05,
+      "loss": 0.202,
+      "step": 8480
+    },
+    {
+      "epoch": 0.135696,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.712903225806452e-05,
+      "loss": 0.1698,
+      "step": 8481
+    },
+    {
+      "epoch": 0.135712,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.712741935483872e-05,
+      "loss": 0.1745,
+      "step": 8482
+    },
+    {
+      "epoch": 0.135728,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.712580645161292e-05,
+      "loss": 0.1683,
+      "step": 8483
+    },
+    {
+      "epoch": 0.135744,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.71241935483871e-05,
+      "loss": 0.1731,
+      "step": 8484
+    },
+    {
+      "epoch": 0.13576,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.712258064516129e-05,
+      "loss": 0.1523,
+      "step": 8485
+    },
+    {
+      "epoch": 0.135776,
+      "grad_norm": 1.4765625,
+      "learning_rate": 8.712096774193549e-05,
+      "loss": 0.2475,
+      "step": 8486
+    },
+    {
+      "epoch": 0.135792,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.711935483870968e-05,
+      "loss": 0.1915,
+      "step": 8487
+    },
+    {
+      "epoch": 0.135808,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.711774193548387e-05,
+      "loss": 0.2,
+      "step": 8488
+    },
+    {
+      "epoch": 0.135824,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.711612903225806e-05,
+      "loss": 0.2043,
+      "step": 8489
+    },
+    {
+      "epoch": 0.13584,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.711451612903226e-05,
+      "loss": 0.2108,
+      "step": 8490
+    },
+    {
+      "epoch": 0.135856,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.711290322580645e-05,
+      "loss": 0.1563,
+      "step": 8491
+    },
+    {
+      "epoch": 0.135872,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.711129032258065e-05,
+      "loss": 0.1549,
+      "step": 8492
+    },
+    {
+      "epoch": 0.135888,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.710967741935485e-05,
+      "loss": 0.1994,
+      "step": 8493
+    },
+    {
+      "epoch": 0.135904,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.710806451612905e-05,
+      "loss": 0.1597,
+      "step": 8494
+    },
+    {
+      "epoch": 0.13592,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.710645161290323e-05,
+      "loss": 0.1637,
+      "step": 8495
+    },
+    {
+      "epoch": 0.135936,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.710483870967743e-05,
+      "loss": 0.1582,
+      "step": 8496
+    },
+    {
+      "epoch": 0.135952,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.710322580645162e-05,
+      "loss": 0.1577,
+      "step": 8497
+    },
+    {
+      "epoch": 0.135968,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.710161290322582e-05,
+      "loss": 0.161,
+      "step": 8498
+    },
+    {
+      "epoch": 0.135984,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.71e-05,
+      "loss": 0.1663,
+      "step": 8499
+    },
+    {
+      "epoch": 0.136,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.709838709677419e-05,
+      "loss": 0.1792,
+      "step": 8500
+    },
+    {
+      "epoch": 0.136016,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.709677419354839e-05,
+      "loss": 0.1762,
+      "step": 8501
+    },
+    {
+      "epoch": 0.136032,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.709516129032257e-05,
+      "loss": 0.1848,
+      "step": 8502
+    },
+    {
+      "epoch": 0.136048,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.709354838709677e-05,
+      "loss": 0.1749,
+      "step": 8503
+    },
+    {
+      "epoch": 0.136064,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.709193548387097e-05,
+      "loss": 0.215,
+      "step": 8504
+    },
+    {
+      "epoch": 0.13608,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.709032258064517e-05,
+      "loss": 0.1746,
+      "step": 8505
+    },
+    {
+      "epoch": 0.136096,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.708870967741936e-05,
+      "loss": 0.1973,
+      "step": 8506
+    },
+    {
+      "epoch": 0.136112,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.708709677419356e-05,
+      "loss": 0.1573,
+      "step": 8507
+    },
+    {
+      "epoch": 0.136128,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.708548387096775e-05,
+      "loss": 0.1606,
+      "step": 8508
+    },
+    {
+      "epoch": 0.136144,
+      "grad_norm": 1.4296875,
+      "learning_rate": 8.708387096774195e-05,
+      "loss": 0.202,
+      "step": 8509
+    },
+    {
+      "epoch": 0.13616,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.708225806451613e-05,
+      "loss": 0.1737,
+      "step": 8510
+    },
+    {
+      "epoch": 0.136176,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.708064516129033e-05,
+      "loss": 0.1789,
+      "step": 8511
+    },
+    {
+      "epoch": 0.136192,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.707903225806452e-05,
+      "loss": 0.1485,
+      "step": 8512
+    },
+    {
+      "epoch": 0.136208,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.707741935483872e-05,
+      "loss": 0.2047,
+      "step": 8513
+    },
+    {
+      "epoch": 0.136224,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.70758064516129e-05,
+      "loss": 0.1745,
+      "step": 8514
+    },
+    {
+      "epoch": 0.13624,
+      "grad_norm": 2.625,
+      "learning_rate": 8.70741935483871e-05,
+      "loss": 0.2464,
+      "step": 8515
+    },
+    {
+      "epoch": 0.136256,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.707258064516129e-05,
+      "loss": 0.1538,
+      "step": 8516
+    },
+    {
+      "epoch": 0.136272,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.707096774193549e-05,
+      "loss": 0.1457,
+      "step": 8517
+    },
+    {
+      "epoch": 0.136288,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.706935483870969e-05,
+      "loss": 0.1759,
+      "step": 8518
+    },
+    {
+      "epoch": 0.136304,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.706774193548387e-05,
+      "loss": 0.166,
+      "step": 8519
+    },
+    {
+      "epoch": 0.13632,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.706612903225807e-05,
+      "loss": 0.1952,
+      "step": 8520
+    },
+    {
+      "epoch": 0.136336,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.706451612903226e-05,
+      "loss": 0.2024,
+      "step": 8521
+    },
+    {
+      "epoch": 0.136352,
+      "grad_norm": 1.125,
+      "learning_rate": 8.706290322580646e-05,
+      "loss": 0.1884,
+      "step": 8522
+    },
+    {
+      "epoch": 0.136368,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.706129032258064e-05,
+      "loss": 0.2081,
+      "step": 8523
+    },
+    {
+      "epoch": 0.136384,
+      "grad_norm": 1.4375,
+      "learning_rate": 8.705967741935484e-05,
+      "loss": 0.2138,
+      "step": 8524
+    },
+    {
+      "epoch": 0.1364,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.705806451612903e-05,
+      "loss": 0.1563,
+      "step": 8525
+    },
+    {
+      "epoch": 0.136416,
+      "grad_norm": 1.3203125,
+      "learning_rate": 8.705645161290323e-05,
+      "loss": 0.1661,
+      "step": 8526
+    },
+    {
+      "epoch": 0.136432,
+      "grad_norm": 0.625,
+      "learning_rate": 8.705483870967742e-05,
+      "loss": 0.1866,
+      "step": 8527
+    },
+    {
+      "epoch": 0.136448,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.705322580645162e-05,
+      "loss": 0.1825,
+      "step": 8528
+    },
+    {
+      "epoch": 0.136464,
+      "grad_norm": 0.75,
+      "learning_rate": 8.705161290322582e-05,
+      "loss": 0.1597,
+      "step": 8529
+    },
+    {
+      "epoch": 0.13648,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.705000000000002e-05,
+      "loss": 0.1596,
+      "step": 8530
+    },
+    {
+      "epoch": 0.136496,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.70483870967742e-05,
+      "loss": 0.1835,
+      "step": 8531
+    },
+    {
+      "epoch": 0.136512,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.704677419354839e-05,
+      "loss": 0.1691,
+      "step": 8532
+    },
+    {
+      "epoch": 0.136528,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.704516129032259e-05,
+      "loss": 0.1688,
+      "step": 8533
+    },
+    {
+      "epoch": 0.136544,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.704354838709677e-05,
+      "loss": 0.1784,
+      "step": 8534
+    },
+    {
+      "epoch": 0.13656,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.704193548387097e-05,
+      "loss": 0.156,
+      "step": 8535
+    },
+    {
+      "epoch": 0.136576,
+      "grad_norm": 0.625,
+      "learning_rate": 8.704032258064516e-05,
+      "loss": 0.1659,
+      "step": 8536
+    },
+    {
+      "epoch": 0.136592,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.703870967741936e-05,
+      "loss": 0.1718,
+      "step": 8537
+    },
+    {
+      "epoch": 0.136608,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.703709677419354e-05,
+      "loss": 0.1831,
+      "step": 8538
+    },
+    {
+      "epoch": 0.136624,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.703548387096774e-05,
+      "loss": 0.1601,
+      "step": 8539
+    },
+    {
+      "epoch": 0.13664,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.703387096774194e-05,
+      "loss": 0.1773,
+      "step": 8540
+    },
+    {
+      "epoch": 0.136656,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.703225806451614e-05,
+      "loss": 0.1595,
+      "step": 8541
+    },
+    {
+      "epoch": 0.136672,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.703064516129033e-05,
+      "loss": 0.1435,
+      "step": 8542
+    },
+    {
+      "epoch": 0.136688,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.702903225806453e-05,
+      "loss": 0.1622,
+      "step": 8543
+    },
+    {
+      "epoch": 0.136704,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.702741935483872e-05,
+      "loss": 0.187,
+      "step": 8544
+    },
+    {
+      "epoch": 0.13672,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.702580645161291e-05,
+      "loss": 0.1431,
+      "step": 8545
+    },
+    {
+      "epoch": 0.136736,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.70241935483871e-05,
+      "loss": 0.1894,
+      "step": 8546
+    },
+    {
+      "epoch": 0.136752,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.702258064516129e-05,
+      "loss": 0.2029,
+      "step": 8547
+    },
+    {
+      "epoch": 0.136768,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.702096774193549e-05,
+      "loss": 0.123,
+      "step": 8548
+    },
+    {
+      "epoch": 0.136784,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.701935483870967e-05,
+      "loss": 0.1913,
+      "step": 8549
+    },
+    {
+      "epoch": 0.1368,
+      "grad_norm": 0.48046875,
+      "learning_rate": 8.701774193548387e-05,
+      "loss": 0.1316,
+      "step": 8550
+    },
+    {
+      "epoch": 0.136816,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.701612903225806e-05,
+      "loss": 0.15,
+      "step": 8551
+    },
+    {
+      "epoch": 0.136832,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.701451612903226e-05,
+      "loss": 0.1938,
+      "step": 8552
+    },
+    {
+      "epoch": 0.136848,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.701290322580646e-05,
+      "loss": 0.1856,
+      "step": 8553
+    },
+    {
+      "epoch": 0.136864,
+      "grad_norm": 1.125,
+      "learning_rate": 8.701129032258066e-05,
+      "loss": 0.1752,
+      "step": 8554
+    },
+    {
+      "epoch": 0.13688,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.700967741935484e-05,
+      "loss": 0.1568,
+      "step": 8555
+    },
+    {
+      "epoch": 0.136896,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.700806451612904e-05,
+      "loss": 0.1652,
+      "step": 8556
+    },
+    {
+      "epoch": 0.136912,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.700645161290323e-05,
+      "loss": 0.1809,
+      "step": 8557
+    },
+    {
+      "epoch": 0.136928,
+      "grad_norm": 0.95703125,
+      "learning_rate": 8.700483870967743e-05,
+      "loss": 0.1769,
+      "step": 8558
+    },
+    {
+      "epoch": 0.136944,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.700322580645161e-05,
+      "loss": 0.1936,
+      "step": 8559
+    },
+    {
+      "epoch": 0.13696,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.700161290322581e-05,
+      "loss": 0.2299,
+      "step": 8560
+    },
+    {
+      "epoch": 0.136976,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.7e-05,
+      "loss": 0.1778,
+      "step": 8561
+    },
+    {
+      "epoch": 0.136992,
+      "grad_norm": 1.3125,
+      "learning_rate": 8.69983870967742e-05,
+      "loss": 0.1994,
+      "step": 8562
+    },
+    {
+      "epoch": 0.137008,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.699677419354839e-05,
+      "loss": 0.1672,
+      "step": 8563
+    },
+    {
+      "epoch": 0.137024,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.699516129032259e-05,
+      "loss": 0.2266,
+      "step": 8564
+    },
+    {
+      "epoch": 0.13704,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.699354838709679e-05,
+      "loss": 0.1584,
+      "step": 8565
+    },
+    {
+      "epoch": 0.137056,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.699193548387097e-05,
+      "loss": 0.1523,
+      "step": 8566
+    },
+    {
+      "epoch": 0.137072,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.699032258064517e-05,
+      "loss": 0.1954,
+      "step": 8567
+    },
+    {
+      "epoch": 0.137088,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.698870967741936e-05,
+      "loss": 0.2008,
+      "step": 8568
+    },
+    {
+      "epoch": 0.137104,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.698709677419356e-05,
+      "loss": 0.1464,
+      "step": 8569
+    },
+    {
+      "epoch": 0.13712,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.698548387096774e-05,
+      "loss": 0.1553,
+      "step": 8570
+    },
+    {
+      "epoch": 0.137136,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.698387096774194e-05,
+      "loss": 0.1572,
+      "step": 8571
+    },
+    {
+      "epoch": 0.137152,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.698225806451613e-05,
+      "loss": 0.1563,
+      "step": 8572
+    },
+    {
+      "epoch": 0.137168,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.698064516129033e-05,
+      "loss": 0.1756,
+      "step": 8573
+    },
+    {
+      "epoch": 0.137184,
+      "grad_norm": 1.234375,
+      "learning_rate": 8.697903225806451e-05,
+      "loss": 0.2124,
+      "step": 8574
+    },
+    {
+      "epoch": 0.1372,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.697741935483871e-05,
+      "loss": 0.1709,
+      "step": 8575
+    },
+    {
+      "epoch": 0.137216,
+      "grad_norm": 1.328125,
+      "learning_rate": 8.69758064516129e-05,
+      "loss": 0.1777,
+      "step": 8576
+    },
+    {
+      "epoch": 0.137232,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.69741935483871e-05,
+      "loss": 0.154,
+      "step": 8577
+    },
+    {
+      "epoch": 0.137248,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.69725806451613e-05,
+      "loss": 0.1861,
+      "step": 8578
+    },
+    {
+      "epoch": 0.137264,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.697096774193549e-05,
+      "loss": 0.1829,
+      "step": 8579
+    },
+    {
+      "epoch": 0.13728,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.696935483870969e-05,
+      "loss": 0.1468,
+      "step": 8580
+    },
+    {
+      "epoch": 0.137296,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.696774193548387e-05,
+      "loss": 0.1994,
+      "step": 8581
+    },
+    {
+      "epoch": 0.137312,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.696612903225807e-05,
+      "loss": 0.1923,
+      "step": 8582
+    },
+    {
+      "epoch": 0.137328,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.696451612903226e-05,
+      "loss": 0.1793,
+      "step": 8583
+    },
+    {
+      "epoch": 0.137344,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.696290322580646e-05,
+      "loss": 0.1702,
+      "step": 8584
+    },
+    {
+      "epoch": 0.13736,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.696129032258064e-05,
+      "loss": 0.1617,
+      "step": 8585
+    },
+    {
+      "epoch": 0.137376,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.695967741935484e-05,
+      "loss": 0.1439,
+      "step": 8586
+    },
+    {
+      "epoch": 0.137392,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.695806451612903e-05,
+      "loss": 0.2048,
+      "step": 8587
+    },
+    {
+      "epoch": 0.137408,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.695645161290323e-05,
+      "loss": 0.1805,
+      "step": 8588
+    },
+    {
+      "epoch": 0.137424,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.695483870967743e-05,
+      "loss": 0.1615,
+      "step": 8589
+    },
+    {
+      "epoch": 0.13744,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.695322580645163e-05,
+      "loss": 0.204,
+      "step": 8590
+    },
+    {
+      "epoch": 0.137456,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.695161290322581e-05,
+      "loss": 0.1456,
+      "step": 8591
+    },
+    {
+      "epoch": 0.137472,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.695000000000001e-05,
+      "loss": 0.1825,
+      "step": 8592
+    },
+    {
+      "epoch": 0.137488,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.69483870967742e-05,
+      "loss": 0.2245,
+      "step": 8593
+    },
+    {
+      "epoch": 0.137504,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.694677419354839e-05,
+      "loss": 0.1883,
+      "step": 8594
+    },
+    {
+      "epoch": 0.13752,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.694516129032258e-05,
+      "loss": 0.1654,
+      "step": 8595
+    },
+    {
+      "epoch": 0.137536,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.694354838709677e-05,
+      "loss": 0.1578,
+      "step": 8596
+    },
+    {
+      "epoch": 0.137552,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.694193548387097e-05,
+      "loss": 0.1673,
+      "step": 8597
+    },
+    {
+      "epoch": 0.137568,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.694032258064516e-05,
+      "loss": 0.1641,
+      "step": 8598
+    },
+    {
+      "epoch": 0.137584,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.693870967741936e-05,
+      "loss": 0.1587,
+      "step": 8599
+    },
+    {
+      "epoch": 0.1376,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.693709677419356e-05,
+      "loss": 0.1696,
+      "step": 8600
+    },
+    {
+      "epoch": 0.137616,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.693548387096776e-05,
+      "loss": 0.1695,
+      "step": 8601
+    },
+    {
+      "epoch": 0.137632,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.693387096774194e-05,
+      "loss": 0.1303,
+      "step": 8602
+    },
+    {
+      "epoch": 0.137648,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.693225806451614e-05,
+      "loss": 0.2117,
+      "step": 8603
+    },
+    {
+      "epoch": 0.137664,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.693064516129033e-05,
+      "loss": 0.1873,
+      "step": 8604
+    },
+    {
+      "epoch": 0.13768,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.692903225806453e-05,
+      "loss": 0.1689,
+      "step": 8605
+    },
+    {
+      "epoch": 0.137696,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.692741935483871e-05,
+      "loss": 0.1638,
+      "step": 8606
+    },
+    {
+      "epoch": 0.137712,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.692580645161291e-05,
+      "loss": 0.1719,
+      "step": 8607
+    },
+    {
+      "epoch": 0.137728,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.69241935483871e-05,
+      "loss": 0.1996,
+      "step": 8608
+    },
+    {
+      "epoch": 0.137744,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.692258064516128e-05,
+      "loss": 0.1709,
+      "step": 8609
+    },
+    {
+      "epoch": 0.13776,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.692096774193548e-05,
+      "loss": 0.1656,
+      "step": 8610
+    },
+    {
+      "epoch": 0.137776,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.691935483870967e-05,
+      "loss": 0.1337,
+      "step": 8611
+    },
+    {
+      "epoch": 0.137792,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.691774193548387e-05,
+      "loss": 0.2081,
+      "step": 8612
+    },
+    {
+      "epoch": 0.137808,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.691612903225807e-05,
+      "loss": 0.1801,
+      "step": 8613
+    },
+    {
+      "epoch": 0.137824,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.691451612903227e-05,
+      "loss": 0.1559,
+      "step": 8614
+    },
+    {
+      "epoch": 0.13784,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.691290322580646e-05,
+      "loss": 0.1214,
+      "step": 8615
+    },
+    {
+      "epoch": 0.137856,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.691129032258065e-05,
+      "loss": 0.1448,
+      "step": 8616
+    },
+    {
+      "epoch": 0.137872,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.690967741935484e-05,
+      "loss": 0.1885,
+      "step": 8617
+    },
+    {
+      "epoch": 0.137888,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.690806451612904e-05,
+      "loss": 0.1813,
+      "step": 8618
+    },
+    {
+      "epoch": 0.137904,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.690645161290323e-05,
+      "loss": 0.1352,
+      "step": 8619
+    },
+    {
+      "epoch": 0.13792,
+      "grad_norm": 0.478515625,
+      "learning_rate": 8.690483870967743e-05,
+      "loss": 0.1447,
+      "step": 8620
+    },
+    {
+      "epoch": 0.137936,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.690322580645161e-05,
+      "loss": 0.1472,
+      "step": 8621
+    },
+    {
+      "epoch": 0.137952,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.690161290322581e-05,
+      "loss": 0.207,
+      "step": 8622
+    },
+    {
+      "epoch": 0.137968,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.69e-05,
+      "loss": 0.2033,
+      "step": 8623
+    },
+    {
+      "epoch": 0.137984,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.68983870967742e-05,
+      "loss": 0.1297,
+      "step": 8624
+    },
+    {
+      "epoch": 0.138,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.68967741935484e-05,
+      "loss": 0.1643,
+      "step": 8625
+    },
+    {
+      "epoch": 0.138016,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.689516129032258e-05,
+      "loss": 0.1735,
+      "step": 8626
+    },
+    {
+      "epoch": 0.138032,
+      "grad_norm": 1.125,
+      "learning_rate": 8.689354838709678e-05,
+      "loss": 0.1861,
+      "step": 8627
+    },
+    {
+      "epoch": 0.138048,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.689193548387097e-05,
+      "loss": 0.1997,
+      "step": 8628
+    },
+    {
+      "epoch": 0.138064,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.689032258064517e-05,
+      "loss": 0.1388,
+      "step": 8629
+    },
+    {
+      "epoch": 0.13808,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.688870967741935e-05,
+      "loss": 0.1469,
+      "step": 8630
+    },
+    {
+      "epoch": 0.138096,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.688709677419355e-05,
+      "loss": 0.1726,
+      "step": 8631
+    },
+    {
+      "epoch": 0.138112,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.688548387096774e-05,
+      "loss": 0.1959,
+      "step": 8632
+    },
+    {
+      "epoch": 0.138128,
+      "grad_norm": 1.296875,
+      "learning_rate": 8.688387096774194e-05,
+      "loss": 0.2022,
+      "step": 8633
+    },
+    {
+      "epoch": 0.138144,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.688225806451613e-05,
+      "loss": 0.1612,
+      "step": 8634
+    },
+    {
+      "epoch": 0.13816,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.688064516129033e-05,
+      "loss": 0.1574,
+      "step": 8635
+    },
+    {
+      "epoch": 0.138176,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.687903225806453e-05,
+      "loss": 0.1646,
+      "step": 8636
+    },
+    {
+      "epoch": 0.138192,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.687741935483871e-05,
+      "loss": 0.1367,
+      "step": 8637
+    },
+    {
+      "epoch": 0.138208,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.687580645161291e-05,
+      "loss": 0.1693,
+      "step": 8638
+    },
+    {
+      "epoch": 0.138224,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.687419354838711e-05,
+      "loss": 0.1449,
+      "step": 8639
+    },
+    {
+      "epoch": 0.13824,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.68725806451613e-05,
+      "loss": 0.1701,
+      "step": 8640
+    },
+    {
+      "epoch": 0.138256,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.687096774193548e-05,
+      "loss": 0.1751,
+      "step": 8641
+    },
+    {
+      "epoch": 0.138272,
+      "grad_norm": 1.4375,
+      "learning_rate": 8.686935483870968e-05,
+      "loss": 0.2088,
+      "step": 8642
+    },
+    {
+      "epoch": 0.138288,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.686774193548387e-05,
+      "loss": 0.1875,
+      "step": 8643
+    },
+    {
+      "epoch": 0.138304,
+      "grad_norm": 1.328125,
+      "learning_rate": 8.686612903225807e-05,
+      "loss": 0.1897,
+      "step": 8644
+    },
+    {
+      "epoch": 0.13832,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.686451612903225e-05,
+      "loss": 0.1694,
+      "step": 8645
+    },
+    {
+      "epoch": 0.138336,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.686290322580645e-05,
+      "loss": 0.1642,
+      "step": 8646
+    },
+    {
+      "epoch": 0.138352,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.686129032258064e-05,
+      "loss": 0.1469,
+      "step": 8647
+    },
+    {
+      "epoch": 0.138368,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.685967741935484e-05,
+      "loss": 0.1411,
+      "step": 8648
+    },
+    {
+      "epoch": 0.138384,
+      "grad_norm": 0.625,
+      "learning_rate": 8.685806451612904e-05,
+      "loss": 0.1763,
+      "step": 8649
+    },
+    {
+      "epoch": 0.1384,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.685645161290324e-05,
+      "loss": 0.163,
+      "step": 8650
+    },
+    {
+      "epoch": 0.138416,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.685483870967743e-05,
+      "loss": 0.1819,
+      "step": 8651
+    },
+    {
+      "epoch": 0.138432,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.685322580645162e-05,
+      "loss": 0.1754,
+      "step": 8652
+    },
+    {
+      "epoch": 0.138448,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.685161290322581e-05,
+      "loss": 0.1636,
+      "step": 8653
+    },
+    {
+      "epoch": 0.138464,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.685000000000001e-05,
+      "loss": 0.1368,
+      "step": 8654
+    },
+    {
+      "epoch": 0.13848,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.68483870967742e-05,
+      "loss": 0.1331,
+      "step": 8655
+    },
+    {
+      "epoch": 0.138496,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.684677419354838e-05,
+      "loss": 0.1649,
+      "step": 8656
+    },
+    {
+      "epoch": 0.138512,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.684516129032258e-05,
+      "loss": 0.1739,
+      "step": 8657
+    },
+    {
+      "epoch": 0.138528,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.684354838709677e-05,
+      "loss": 0.1589,
+      "step": 8658
+    },
+    {
+      "epoch": 0.138544,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.684193548387097e-05,
+      "loss": 0.1364,
+      "step": 8659
+    },
+    {
+      "epoch": 0.13856,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.684032258064517e-05,
+      "loss": 0.1661,
+      "step": 8660
+    },
+    {
+      "epoch": 0.138576,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.683870967741937e-05,
+      "loss": 0.1354,
+      "step": 8661
+    },
+    {
+      "epoch": 0.138592,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.683709677419355e-05,
+      "loss": 0.1891,
+      "step": 8662
+    },
+    {
+      "epoch": 0.138608,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.683548387096775e-05,
+      "loss": 0.1817,
+      "step": 8663
+    },
+    {
+      "epoch": 0.138624,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.683387096774194e-05,
+      "loss": 0.1435,
+      "step": 8664
+    },
+    {
+      "epoch": 0.13864,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.683225806451614e-05,
+      "loss": 0.152,
+      "step": 8665
+    },
+    {
+      "epoch": 0.138656,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.683064516129032e-05,
+      "loss": 0.2128,
+      "step": 8666
+    },
+    {
+      "epoch": 0.138672,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.682903225806452e-05,
+      "loss": 0.1915,
+      "step": 8667
+    },
+    {
+      "epoch": 0.138688,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.682741935483871e-05,
+      "loss": 0.1652,
+      "step": 8668
+    },
+    {
+      "epoch": 0.138704,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.682580645161291e-05,
+      "loss": 0.1888,
+      "step": 8669
+    },
+    {
+      "epoch": 0.13872,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.68241935483871e-05,
+      "loss": 0.1493,
+      "step": 8670
+    },
+    {
+      "epoch": 0.138736,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.68225806451613e-05,
+      "loss": 0.1588,
+      "step": 8671
+    },
+    {
+      "epoch": 0.138752,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.682096774193548e-05,
+      "loss": 0.2025,
+      "step": 8672
+    },
+    {
+      "epoch": 0.138768,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.681935483870968e-05,
+      "loss": 0.1489,
+      "step": 8673
+    },
+    {
+      "epoch": 0.138784,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.681774193548388e-05,
+      "loss": 0.1731,
+      "step": 8674
+    },
+    {
+      "epoch": 0.1388,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.681612903225807e-05,
+      "loss": 0.204,
+      "step": 8675
+    },
+    {
+      "epoch": 0.138816,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.681451612903227e-05,
+      "loss": 0.157,
+      "step": 8676
+    },
+    {
+      "epoch": 0.138832,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.681290322580645e-05,
+      "loss": 0.1861,
+      "step": 8677
+    },
+    {
+      "epoch": 0.138848,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.681129032258065e-05,
+      "loss": 0.1533,
+      "step": 8678
+    },
+    {
+      "epoch": 0.138864,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.680967741935484e-05,
+      "loss": 0.2042,
+      "step": 8679
+    },
+    {
+      "epoch": 0.13888,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.680806451612904e-05,
+      "loss": 0.1806,
+      "step": 8680
+    },
+    {
+      "epoch": 0.138896,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.680645161290322e-05,
+      "loss": 0.1757,
+      "step": 8681
+    },
+    {
+      "epoch": 0.138912,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.680483870967742e-05,
+      "loss": 0.1966,
+      "step": 8682
+    },
+    {
+      "epoch": 0.138928,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.680322580645161e-05,
+      "loss": 0.14,
+      "step": 8683
+    },
+    {
+      "epoch": 0.138944,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.680161290322581e-05,
+      "loss": 0.1383,
+      "step": 8684
+    },
+    {
+      "epoch": 0.13896,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.680000000000001e-05,
+      "loss": 0.2251,
+      "step": 8685
+    },
+    {
+      "epoch": 0.138976,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.679838709677421e-05,
+      "loss": 0.2191,
+      "step": 8686
+    },
+    {
+      "epoch": 0.138992,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.67967741935484e-05,
+      "loss": 0.1817,
+      "step": 8687
+    },
+    {
+      "epoch": 0.139008,
+      "grad_norm": 1.5546875,
+      "learning_rate": 8.679516129032258e-05,
+      "loss": 0.2383,
+      "step": 8688
+    },
+    {
+      "epoch": 0.139024,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.679354838709678e-05,
+      "loss": 0.1675,
+      "step": 8689
+    },
+    {
+      "epoch": 0.13904,
+      "grad_norm": 1.25,
+      "learning_rate": 8.679193548387097e-05,
+      "loss": 0.2285,
+      "step": 8690
+    },
+    {
+      "epoch": 0.139056,
+      "grad_norm": 1.515625,
+      "learning_rate": 8.679032258064517e-05,
+      "loss": 0.1908,
+      "step": 8691
+    },
+    {
+      "epoch": 0.139072,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.678870967741935e-05,
+      "loss": 0.1831,
+      "step": 8692
+    },
+    {
+      "epoch": 0.139088,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.678709677419355e-05,
+      "loss": 0.1865,
+      "step": 8693
+    },
+    {
+      "epoch": 0.139104,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.678548387096774e-05,
+      "loss": 0.1678,
+      "step": 8694
+    },
+    {
+      "epoch": 0.13912,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.678387096774194e-05,
+      "loss": 0.1976,
+      "step": 8695
+    },
+    {
+      "epoch": 0.139136,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.678225806451614e-05,
+      "loss": 0.1713,
+      "step": 8696
+    },
+    {
+      "epoch": 0.139152,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.678064516129034e-05,
+      "loss": 0.1393,
+      "step": 8697
+    },
+    {
+      "epoch": 0.139168,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.677903225806452e-05,
+      "loss": 0.1906,
+      "step": 8698
+    },
+    {
+      "epoch": 0.139184,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.677741935483872e-05,
+      "loss": 0.1516,
+      "step": 8699
+    },
+    {
+      "epoch": 0.1392,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.677580645161291e-05,
+      "loss": 0.1813,
+      "step": 8700
+    },
+    {
+      "epoch": 0.139216,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.677419354838711e-05,
+      "loss": 0.1982,
+      "step": 8701
+    },
+    {
+      "epoch": 0.139232,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.67725806451613e-05,
+      "loss": 0.1729,
+      "step": 8702
+    },
+    {
+      "epoch": 0.139248,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.677096774193548e-05,
+      "loss": 0.1405,
+      "step": 8703
+    },
+    {
+      "epoch": 0.139264,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.676935483870968e-05,
+      "loss": 0.2066,
+      "step": 8704
+    },
+    {
+      "epoch": 0.13928,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.676774193548387e-05,
+      "loss": 0.1973,
+      "step": 8705
+    },
+    {
+      "epoch": 0.139296,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.676612903225807e-05,
+      "loss": 0.1455,
+      "step": 8706
+    },
+    {
+      "epoch": 0.139312,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.676451612903225e-05,
+      "loss": 0.1778,
+      "step": 8707
+    },
+    {
+      "epoch": 0.139328,
+      "grad_norm": 0.453125,
+      "learning_rate": 8.676290322580645e-05,
+      "loss": 0.1264,
+      "step": 8708
+    },
+    {
+      "epoch": 0.139344,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.676129032258065e-05,
+      "loss": 0.2068,
+      "step": 8709
+    },
+    {
+      "epoch": 0.13936,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.675967741935485e-05,
+      "loss": 0.2002,
+      "step": 8710
+    },
+    {
+      "epoch": 0.139376,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.675806451612904e-05,
+      "loss": 0.1926,
+      "step": 8711
+    },
+    {
+      "epoch": 0.139392,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.675645161290324e-05,
+      "loss": 0.1653,
+      "step": 8712
+    },
+    {
+      "epoch": 0.139408,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.675483870967742e-05,
+      "loss": 0.1845,
+      "step": 8713
+    },
+    {
+      "epoch": 0.139424,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.675322580645162e-05,
+      "loss": 0.168,
+      "step": 8714
+    },
+    {
+      "epoch": 0.13944,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.675161290322581e-05,
+      "loss": 0.1532,
+      "step": 8715
+    },
+    {
+      "epoch": 0.139456,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.675000000000001e-05,
+      "loss": 0.1703,
+      "step": 8716
+    },
+    {
+      "epoch": 0.139472,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.67483870967742e-05,
+      "loss": 0.192,
+      "step": 8717
+    },
+    {
+      "epoch": 0.139488,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.674677419354838e-05,
+      "loss": 0.1478,
+      "step": 8718
+    },
+    {
+      "epoch": 0.139504,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.674516129032258e-05,
+      "loss": 0.192,
+      "step": 8719
+    },
+    {
+      "epoch": 0.13952,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.674354838709678e-05,
+      "loss": 0.1422,
+      "step": 8720
+    },
+    {
+      "epoch": 0.139536,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.674193548387098e-05,
+      "loss": 0.1796,
+      "step": 8721
+    },
+    {
+      "epoch": 0.139552,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.674032258064517e-05,
+      "loss": 0.1576,
+      "step": 8722
+    },
+    {
+      "epoch": 0.139568,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.673870967741936e-05,
+      "loss": 0.2285,
+      "step": 8723
+    },
+    {
+      "epoch": 0.139584,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.673709677419355e-05,
+      "loss": 0.1648,
+      "step": 8724
+    },
+    {
+      "epoch": 0.1396,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.673548387096775e-05,
+      "loss": 0.1837,
+      "step": 8725
+    },
+    {
+      "epoch": 0.139616,
+      "grad_norm": 0.75,
+      "learning_rate": 8.673387096774194e-05,
+      "loss": 0.2049,
+      "step": 8726
+    },
+    {
+      "epoch": 0.139632,
+      "grad_norm": 0.435546875,
+      "learning_rate": 8.673225806451614e-05,
+      "loss": 0.144,
+      "step": 8727
+    },
+    {
+      "epoch": 0.139648,
+      "grad_norm": 1.265625,
+      "learning_rate": 8.673064516129032e-05,
+      "loss": 0.1624,
+      "step": 8728
+    },
+    {
+      "epoch": 0.139664,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.672903225806452e-05,
+      "loss": 0.1719,
+      "step": 8729
+    },
+    {
+      "epoch": 0.13968,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.672741935483871e-05,
+      "loss": 0.2195,
+      "step": 8730
+    },
+    {
+      "epoch": 0.139696,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.672580645161291e-05,
+      "loss": 0.1628,
+      "step": 8731
+    },
+    {
+      "epoch": 0.139712,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.672419354838711e-05,
+      "loss": 0.1945,
+      "step": 8732
+    },
+    {
+      "epoch": 0.139728,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.67225806451613e-05,
+      "loss": 0.1693,
+      "step": 8733
+    },
+    {
+      "epoch": 0.139744,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.672096774193549e-05,
+      "loss": 0.1835,
+      "step": 8734
+    },
+    {
+      "epoch": 0.13976,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.671935483870968e-05,
+      "loss": 0.2075,
+      "step": 8735
+    },
+    {
+      "epoch": 0.139776,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.671774193548388e-05,
+      "loss": 0.1581,
+      "step": 8736
+    },
+    {
+      "epoch": 0.139792,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.671612903225806e-05,
+      "loss": 0.1942,
+      "step": 8737
+    },
+    {
+      "epoch": 0.139808,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.671451612903226e-05,
+      "loss": 0.1587,
+      "step": 8738
+    },
+    {
+      "epoch": 0.139824,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.671290322580645e-05,
+      "loss": 0.1572,
+      "step": 8739
+    },
+    {
+      "epoch": 0.13984,
+      "grad_norm": 0.484375,
+      "learning_rate": 8.671129032258065e-05,
+      "loss": 0.1668,
+      "step": 8740
+    },
+    {
+      "epoch": 0.139856,
+      "grad_norm": 1.4921875,
+      "learning_rate": 8.670967741935484e-05,
+      "loss": 0.183,
+      "step": 8741
+    },
+    {
+      "epoch": 0.139872,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.670806451612904e-05,
+      "loss": 0.1739,
+      "step": 8742
+    },
+    {
+      "epoch": 0.139888,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.670645161290322e-05,
+      "loss": 0.1447,
+      "step": 8743
+    },
+    {
+      "epoch": 0.139904,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.670483870967742e-05,
+      "loss": 0.1462,
+      "step": 8744
+    },
+    {
+      "epoch": 0.13992,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.670322580645162e-05,
+      "loss": 0.2006,
+      "step": 8745
+    },
+    {
+      "epoch": 0.139936,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.670161290322582e-05,
+      "loss": 0.1969,
+      "step": 8746
+    },
+    {
+      "epoch": 0.139952,
+      "grad_norm": 0.75,
+      "learning_rate": 8.67e-05,
+      "loss": 0.1867,
+      "step": 8747
+    },
+    {
+      "epoch": 0.139968,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.66983870967742e-05,
+      "loss": 0.1756,
+      "step": 8748
+    },
+    {
+      "epoch": 0.139984,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.669677419354839e-05,
+      "loss": 0.1742,
+      "step": 8749
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.669516129032258e-05,
+      "loss": 0.1548,
+      "step": 8750
+    },
+    {
+      "epoch": 0.140016,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.669354838709678e-05,
+      "loss": 0.1984,
+      "step": 8751
+    },
+    {
+      "epoch": 0.140032,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.669193548387096e-05,
+      "loss": 0.2087,
+      "step": 8752
+    },
+    {
+      "epoch": 0.140048,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.669032258064516e-05,
+      "loss": 0.1726,
+      "step": 8753
+    },
+    {
+      "epoch": 0.140064,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.668870967741935e-05,
+      "loss": 0.1446,
+      "step": 8754
+    },
+    {
+      "epoch": 0.14008,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.668709677419355e-05,
+      "loss": 0.1555,
+      "step": 8755
+    },
+    {
+      "epoch": 0.140096,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.668548387096775e-05,
+      "loss": 0.1553,
+      "step": 8756
+    },
+    {
+      "epoch": 0.140112,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.668387096774195e-05,
+      "loss": 0.219,
+      "step": 8757
+    },
+    {
+      "epoch": 0.140128,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.668225806451613e-05,
+      "loss": 0.2073,
+      "step": 8758
+    },
+    {
+      "epoch": 0.140144,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.668064516129033e-05,
+      "loss": 0.1656,
+      "step": 8759
+    },
+    {
+      "epoch": 0.14016,
+      "grad_norm": 1.2734375,
+      "learning_rate": 8.667903225806452e-05,
+      "loss": 0.2119,
+      "step": 8760
+    },
+    {
+      "epoch": 0.140176,
+      "grad_norm": 0.75,
+      "learning_rate": 8.667741935483872e-05,
+      "loss": 0.1783,
+      "step": 8761
+    },
+    {
+      "epoch": 0.140192,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.66758064516129e-05,
+      "loss": 0.1895,
+      "step": 8762
+    },
+    {
+      "epoch": 0.140208,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.66741935483871e-05,
+      "loss": 0.1589,
+      "step": 8763
+    },
+    {
+      "epoch": 0.140224,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.667258064516129e-05,
+      "loss": 0.1621,
+      "step": 8764
+    },
+    {
+      "epoch": 0.14024,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.667096774193548e-05,
+      "loss": 0.1682,
+      "step": 8765
+    },
+    {
+      "epoch": 0.140256,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.666935483870968e-05,
+      "loss": 0.1583,
+      "step": 8766
+    },
+    {
+      "epoch": 0.140272,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.666774193548386e-05,
+      "loss": 0.1978,
+      "step": 8767
+    },
+    {
+      "epoch": 0.140288,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.666612903225806e-05,
+      "loss": 0.1533,
+      "step": 8768
+    },
+    {
+      "epoch": 0.140304,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.666451612903226e-05,
+      "loss": 0.1764,
+      "step": 8769
+    },
+    {
+      "epoch": 0.14032,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.666290322580646e-05,
+      "loss": 0.1476,
+      "step": 8770
+    },
+    {
+      "epoch": 0.140336,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.666129032258065e-05,
+      "loss": 0.1782,
+      "step": 8771
+    },
+    {
+      "epoch": 0.140352,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.665967741935485e-05,
+      "loss": 0.1957,
+      "step": 8772
+    },
+    {
+      "epoch": 0.140368,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.665806451612903e-05,
+      "loss": 0.1238,
+      "step": 8773
+    },
+    {
+      "epoch": 0.140384,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.665645161290323e-05,
+      "loss": 0.1654,
+      "step": 8774
+    },
+    {
+      "epoch": 0.1404,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.665483870967742e-05,
+      "loss": 0.1547,
+      "step": 8775
+    },
+    {
+      "epoch": 0.140416,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.665322580645162e-05,
+      "loss": 0.1911,
+      "step": 8776
+    },
+    {
+      "epoch": 0.140432,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.66516129032258e-05,
+      "loss": 0.1914,
+      "step": 8777
+    },
+    {
+      "epoch": 0.140448,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.665e-05,
+      "loss": 0.1907,
+      "step": 8778
+    },
+    {
+      "epoch": 0.140464,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.664838709677419e-05,
+      "loss": 0.1849,
+      "step": 8779
+    },
+    {
+      "epoch": 0.14048,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.664677419354839e-05,
+      "loss": 0.1366,
+      "step": 8780
+    },
+    {
+      "epoch": 0.140496,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.664516129032259e-05,
+      "loss": 0.1524,
+      "step": 8781
+    },
+    {
+      "epoch": 0.140512,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.664354838709678e-05,
+      "loss": 0.1781,
+      "step": 8782
+    },
+    {
+      "epoch": 0.140528,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.664193548387098e-05,
+      "loss": 0.1816,
+      "step": 8783
+    },
+    {
+      "epoch": 0.140544,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.664032258064516e-05,
+      "loss": 0.168,
+      "step": 8784
+    },
+    {
+      "epoch": 0.14056,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.663870967741936e-05,
+      "loss": 0.169,
+      "step": 8785
+    },
+    {
+      "epoch": 0.140576,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.663709677419355e-05,
+      "loss": 0.1611,
+      "step": 8786
+    },
+    {
+      "epoch": 0.140592,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.663548387096775e-05,
+      "loss": 0.2421,
+      "step": 8787
+    },
+    {
+      "epoch": 0.140608,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.663387096774193e-05,
+      "loss": 0.1773,
+      "step": 8788
+    },
+    {
+      "epoch": 0.140624,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.663225806451613e-05,
+      "loss": 0.1761,
+      "step": 8789
+    },
+    {
+      "epoch": 0.14064,
+      "grad_norm": 1.3125,
+      "learning_rate": 8.663064516129032e-05,
+      "loss": 0.1778,
+      "step": 8790
+    },
+    {
+      "epoch": 0.140656,
+      "grad_norm": 1.3984375,
+      "learning_rate": 8.662903225806452e-05,
+      "loss": 0.1829,
+      "step": 8791
+    },
+    {
+      "epoch": 0.140672,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.662741935483872e-05,
+      "loss": 0.168,
+      "step": 8792
+    },
+    {
+      "epoch": 0.140688,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.662580645161292e-05,
+      "loss": 0.2003,
+      "step": 8793
+    },
+    {
+      "epoch": 0.140704,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.66241935483871e-05,
+      "loss": 0.1744,
+      "step": 8794
+    },
+    {
+      "epoch": 0.14072,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.66225806451613e-05,
+      "loss": 0.1671,
+      "step": 8795
+    },
+    {
+      "epoch": 0.140736,
+      "grad_norm": 1.4921875,
+      "learning_rate": 8.662096774193549e-05,
+      "loss": 0.166,
+      "step": 8796
+    },
+    {
+      "epoch": 0.140752,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.661935483870968e-05,
+      "loss": 0.1703,
+      "step": 8797
+    },
+    {
+      "epoch": 0.140768,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.661774193548388e-05,
+      "loss": 0.2124,
+      "step": 8798
+    },
+    {
+      "epoch": 0.140784,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.661612903225806e-05,
+      "loss": 0.1787,
+      "step": 8799
+    },
+    {
+      "epoch": 0.1408,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.661451612903226e-05,
+      "loss": 0.187,
+      "step": 8800
+    },
+    {
+      "epoch": 0.140816,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.661290322580645e-05,
+      "loss": 0.1976,
+      "step": 8801
+    },
+    {
+      "epoch": 0.140832,
+      "grad_norm": 1.734375,
+      "learning_rate": 8.661129032258065e-05,
+      "loss": 0.19,
+      "step": 8802
+    },
+    {
+      "epoch": 0.140848,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.660967741935483e-05,
+      "loss": 0.1745,
+      "step": 8803
+    },
+    {
+      "epoch": 0.140864,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.660806451612903e-05,
+      "loss": 0.1838,
+      "step": 8804
+    },
+    {
+      "epoch": 0.14088,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.660645161290323e-05,
+      "loss": 0.1449,
+      "step": 8805
+    },
+    {
+      "epoch": 0.140896,
+      "grad_norm": 1.296875,
+      "learning_rate": 8.660483870967743e-05,
+      "loss": 0.1752,
+      "step": 8806
+    },
+    {
+      "epoch": 0.140912,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.660322580645162e-05,
+      "loss": 0.1627,
+      "step": 8807
+    },
+    {
+      "epoch": 0.140928,
+      "grad_norm": 0.75,
+      "learning_rate": 8.660161290322582e-05,
+      "loss": 0.1971,
+      "step": 8808
+    },
+    {
+      "epoch": 0.140944,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.66e-05,
+      "loss": 0.2178,
+      "step": 8809
+    },
+    {
+      "epoch": 0.14096,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.65983870967742e-05,
+      "loss": 0.1406,
+      "step": 8810
+    },
+    {
+      "epoch": 0.140976,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.659677419354839e-05,
+      "loss": 0.1498,
+      "step": 8811
+    },
+    {
+      "epoch": 0.140992,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.659516129032258e-05,
+      "loss": 0.1789,
+      "step": 8812
+    },
+    {
+      "epoch": 0.141008,
+      "grad_norm": 1.4140625,
+      "learning_rate": 8.659354838709678e-05,
+      "loss": 0.1511,
+      "step": 8813
+    },
+    {
+      "epoch": 0.141024,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.659193548387096e-05,
+      "loss": 0.1848,
+      "step": 8814
+    },
+    {
+      "epoch": 0.14104,
+      "grad_norm": 1.515625,
+      "learning_rate": 8.659032258064516e-05,
+      "loss": 0.2342,
+      "step": 8815
+    },
+    {
+      "epoch": 0.141056,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.658870967741936e-05,
+      "loss": 0.2001,
+      "step": 8816
+    },
+    {
+      "epoch": 0.141072,
+      "grad_norm": 1.25,
+      "learning_rate": 8.658709677419356e-05,
+      "loss": 0.1667,
+      "step": 8817
+    },
+    {
+      "epoch": 0.141088,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.658548387096775e-05,
+      "loss": 0.2103,
+      "step": 8818
+    },
+    {
+      "epoch": 0.141104,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.658387096774195e-05,
+      "loss": 0.1499,
+      "step": 8819
+    },
+    {
+      "epoch": 0.14112,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.658225806451613e-05,
+      "loss": 0.1693,
+      "step": 8820
+    },
+    {
+      "epoch": 0.141136,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.658064516129033e-05,
+      "loss": 0.1831,
+      "step": 8821
+    },
+    {
+      "epoch": 0.141152,
+      "grad_norm": 1.453125,
+      "learning_rate": 8.657903225806452e-05,
+      "loss": 0.1687,
+      "step": 8822
+    },
+    {
+      "epoch": 0.141168,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.657741935483872e-05,
+      "loss": 0.1341,
+      "step": 8823
+    },
+    {
+      "epoch": 0.141184,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.65758064516129e-05,
+      "loss": 0.1753,
+      "step": 8824
+    },
+    {
+      "epoch": 0.1412,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.65741935483871e-05,
+      "loss": 0.172,
+      "step": 8825
+    },
+    {
+      "epoch": 0.141216,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.657258064516129e-05,
+      "loss": 0.1389,
+      "step": 8826
+    },
+    {
+      "epoch": 0.141232,
+      "grad_norm": 1.296875,
+      "learning_rate": 8.657096774193549e-05,
+      "loss": 0.2189,
+      "step": 8827
+    },
+    {
+      "epoch": 0.141248,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.656935483870968e-05,
+      "loss": 0.182,
+      "step": 8828
+    },
+    {
+      "epoch": 0.141264,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.656774193548387e-05,
+      "loss": 0.175,
+      "step": 8829
+    },
+    {
+      "epoch": 0.14128,
+      "grad_norm": 1.625,
+      "learning_rate": 8.656612903225807e-05,
+      "loss": 0.2024,
+      "step": 8830
+    },
+    {
+      "epoch": 0.141296,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.656451612903226e-05,
+      "loss": 0.1843,
+      "step": 8831
+    },
+    {
+      "epoch": 0.141312,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.656290322580646e-05,
+      "loss": 0.1725,
+      "step": 8832
+    },
+    {
+      "epoch": 0.141328,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.656129032258065e-05,
+      "loss": 0.1765,
+      "step": 8833
+    },
+    {
+      "epoch": 0.141344,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.655967741935485e-05,
+      "loss": 0.1918,
+      "step": 8834
+    },
+    {
+      "epoch": 0.14136,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.655806451612903e-05,
+      "loss": 0.1732,
+      "step": 8835
+    },
+    {
+      "epoch": 0.141376,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.655645161290323e-05,
+      "loss": 0.186,
+      "step": 8836
+    },
+    {
+      "epoch": 0.141392,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.655483870967742e-05,
+      "loss": 0.2286,
+      "step": 8837
+    },
+    {
+      "epoch": 0.141408,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.655322580645162e-05,
+      "loss": 0.1999,
+      "step": 8838
+    },
+    {
+      "epoch": 0.141424,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.65516129032258e-05,
+      "loss": 0.1334,
+      "step": 8839
+    },
+    {
+      "epoch": 0.14144,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.655e-05,
+      "loss": 0.1369,
+      "step": 8840
+    },
+    {
+      "epoch": 0.141456,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.65483870967742e-05,
+      "loss": 0.2098,
+      "step": 8841
+    },
+    {
+      "epoch": 0.141472,
+      "grad_norm": 0.75,
+      "learning_rate": 8.65467741935484e-05,
+      "loss": 0.1397,
+      "step": 8842
+    },
+    {
+      "epoch": 0.141488,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.654516129032259e-05,
+      "loss": 0.1846,
+      "step": 8843
+    },
+    {
+      "epoch": 0.141504,
+      "grad_norm": 0.75,
+      "learning_rate": 8.654354838709677e-05,
+      "loss": 0.1943,
+      "step": 8844
+    },
+    {
+      "epoch": 0.14152,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.654193548387097e-05,
+      "loss": 0.1824,
+      "step": 8845
+    },
+    {
+      "epoch": 0.141536,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.654032258064516e-05,
+      "loss": 0.2136,
+      "step": 8846
+    },
+    {
+      "epoch": 0.141552,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.653870967741936e-05,
+      "loss": 0.2081,
+      "step": 8847
+    },
+    {
+      "epoch": 0.141568,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.653709677419355e-05,
+      "loss": 0.1618,
+      "step": 8848
+    },
+    {
+      "epoch": 0.141584,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.653548387096775e-05,
+      "loss": 0.1595,
+      "step": 8849
+    },
+    {
+      "epoch": 0.1416,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.653387096774193e-05,
+      "loss": 0.1513,
+      "step": 8850
+    },
+    {
+      "epoch": 0.141616,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.653225806451613e-05,
+      "loss": 0.1976,
+      "step": 8851
+    },
+    {
+      "epoch": 0.141632,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.653064516129033e-05,
+      "loss": 0.187,
+      "step": 8852
+    },
+    {
+      "epoch": 0.141648,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.652903225806453e-05,
+      "loss": 0.1668,
+      "step": 8853
+    },
+    {
+      "epoch": 0.141664,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.652741935483872e-05,
+      "loss": 0.1307,
+      "step": 8854
+    },
+    {
+      "epoch": 0.14168,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.652580645161292e-05,
+      "loss": 0.1916,
+      "step": 8855
+    },
+    {
+      "epoch": 0.141696,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.65241935483871e-05,
+      "loss": 0.1826,
+      "step": 8856
+    },
+    {
+      "epoch": 0.141712,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.65225806451613e-05,
+      "loss": 0.197,
+      "step": 8857
+    },
+    {
+      "epoch": 0.141728,
+      "grad_norm": 1.3359375,
+      "learning_rate": 8.652096774193549e-05,
+      "loss": 0.1891,
+      "step": 8858
+    },
+    {
+      "epoch": 0.141744,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.651935483870967e-05,
+      "loss": 0.1756,
+      "step": 8859
+    },
+    {
+      "epoch": 0.14176,
+      "grad_norm": 1.8828125,
+      "learning_rate": 8.651774193548387e-05,
+      "loss": 0.1948,
+      "step": 8860
+    },
+    {
+      "epoch": 0.141776,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.651612903225806e-05,
+      "loss": 0.1854,
+      "step": 8861
+    },
+    {
+      "epoch": 0.141792,
+      "grad_norm": 0.75,
+      "learning_rate": 8.651451612903226e-05,
+      "loss": 0.1343,
+      "step": 8862
+    },
+    {
+      "epoch": 0.141808,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.651290322580645e-05,
+      "loss": 0.192,
+      "step": 8863
+    },
+    {
+      "epoch": 0.141824,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.651129032258065e-05,
+      "loss": 0.1342,
+      "step": 8864
+    },
+    {
+      "epoch": 0.14184,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.650967741935484e-05,
+      "loss": 0.196,
+      "step": 8865
+    },
+    {
+      "epoch": 0.141856,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.650806451612904e-05,
+      "loss": 0.1526,
+      "step": 8866
+    },
+    {
+      "epoch": 0.141872,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.650645161290323e-05,
+      "loss": 0.1888,
+      "step": 8867
+    },
+    {
+      "epoch": 0.141888,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.650483870967743e-05,
+      "loss": 0.1741,
+      "step": 8868
+    },
+    {
+      "epoch": 0.141904,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.650322580645162e-05,
+      "loss": 0.186,
+      "step": 8869
+    },
+    {
+      "epoch": 0.14192,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.650161290322582e-05,
+      "loss": 0.2051,
+      "step": 8870
+    },
+    {
+      "epoch": 0.141936,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.65e-05,
+      "loss": 0.1694,
+      "step": 8871
+    },
+    {
+      "epoch": 0.141952,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.64983870967742e-05,
+      "loss": 0.2186,
+      "step": 8872
+    },
+    {
+      "epoch": 0.141968,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.649677419354839e-05,
+      "loss": 0.1414,
+      "step": 8873
+    },
+    {
+      "epoch": 0.141984,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.649516129032257e-05,
+      "loss": 0.1255,
+      "step": 8874
+    },
+    {
+      "epoch": 0.142,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.649354838709677e-05,
+      "loss": 0.1562,
+      "step": 8875
+    },
+    {
+      "epoch": 0.142016,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.649193548387097e-05,
+      "loss": 0.1657,
+      "step": 8876
+    },
+    {
+      "epoch": 0.142032,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.649032258064517e-05,
+      "loss": 0.1612,
+      "step": 8877
+    },
+    {
+      "epoch": 0.142048,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.648870967741936e-05,
+      "loss": 0.1502,
+      "step": 8878
+    },
+    {
+      "epoch": 0.142064,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.648709677419356e-05,
+      "loss": 0.1613,
+      "step": 8879
+    },
+    {
+      "epoch": 0.14208,
+      "grad_norm": 1.4296875,
+      "learning_rate": 8.648548387096774e-05,
+      "loss": 0.1633,
+      "step": 8880
+    },
+    {
+      "epoch": 0.142096,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.648387096774194e-05,
+      "loss": 0.2072,
+      "step": 8881
+    },
+    {
+      "epoch": 0.142112,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.648225806451613e-05,
+      "loss": 0.1796,
+      "step": 8882
+    },
+    {
+      "epoch": 0.142128,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.648064516129033e-05,
+      "loss": 0.1909,
+      "step": 8883
+    },
+    {
+      "epoch": 0.142144,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.647903225806452e-05,
+      "loss": 0.1859,
+      "step": 8884
+    },
+    {
+      "epoch": 0.14216,
+      "grad_norm": 0.875,
+      "learning_rate": 8.647741935483872e-05,
+      "loss": 0.2056,
+      "step": 8885
+    },
+    {
+      "epoch": 0.142176,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.64758064516129e-05,
+      "loss": 0.1399,
+      "step": 8886
+    },
+    {
+      "epoch": 0.142192,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.64741935483871e-05,
+      "loss": 0.1957,
+      "step": 8887
+    },
+    {
+      "epoch": 0.142208,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.64725806451613e-05,
+      "loss": 0.1799,
+      "step": 8888
+    },
+    {
+      "epoch": 0.142224,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.647096774193549e-05,
+      "loss": 0.1752,
+      "step": 8889
+    },
+    {
+      "epoch": 0.14224,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.646935483870969e-05,
+      "loss": 0.1618,
+      "step": 8890
+    },
+    {
+      "epoch": 0.142256,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.646774193548387e-05,
+      "loss": 0.1654,
+      "step": 8891
+    },
+    {
+      "epoch": 0.142272,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.646612903225807e-05,
+      "loss": 0.1563,
+      "step": 8892
+    },
+    {
+      "epoch": 0.142288,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.646451612903226e-05,
+      "loss": 0.2092,
+      "step": 8893
+    },
+    {
+      "epoch": 0.142304,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.646290322580646e-05,
+      "loss": 0.1697,
+      "step": 8894
+    },
+    {
+      "epoch": 0.14232,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.646129032258064e-05,
+      "loss": 0.1722,
+      "step": 8895
+    },
+    {
+      "epoch": 0.142336,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.645967741935484e-05,
+      "loss": 0.1943,
+      "step": 8896
+    },
+    {
+      "epoch": 0.142352,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.645806451612903e-05,
+      "loss": 0.2033,
+      "step": 8897
+    },
+    {
+      "epoch": 0.142368,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.645645161290323e-05,
+      "loss": 0.1561,
+      "step": 8898
+    },
+    {
+      "epoch": 0.142384,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.645483870967742e-05,
+      "loss": 0.1566,
+      "step": 8899
+    },
+    {
+      "epoch": 0.1424,
+      "grad_norm": 1.3359375,
+      "learning_rate": 8.645322580645161e-05,
+      "loss": 0.1621,
+      "step": 8900
+    },
+    {
+      "epoch": 0.142416,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.645161290322581e-05,
+      "loss": 0.205,
+      "step": 8901
+    },
+    {
+      "epoch": 0.142432,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.645000000000001e-05,
+      "loss": 0.1788,
+      "step": 8902
+    },
+    {
+      "epoch": 0.142448,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.64483870967742e-05,
+      "loss": 0.193,
+      "step": 8903
+    },
+    {
+      "epoch": 0.142464,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.64467741935484e-05,
+      "loss": 0.16,
+      "step": 8904
+    },
+    {
+      "epoch": 0.14248,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.644516129032259e-05,
+      "loss": 0.197,
+      "step": 8905
+    },
+    {
+      "epoch": 0.142496,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.644354838709677e-05,
+      "loss": 0.1537,
+      "step": 8906
+    },
+    {
+      "epoch": 0.142512,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.644193548387097e-05,
+      "loss": 0.1938,
+      "step": 8907
+    },
+    {
+      "epoch": 0.142528,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.644032258064516e-05,
+      "loss": 0.1433,
+      "step": 8908
+    },
+    {
+      "epoch": 0.142544,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.643870967741936e-05,
+      "loss": 0.1741,
+      "step": 8909
+    },
+    {
+      "epoch": 0.14256,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.643709677419354e-05,
+      "loss": 0.1431,
+      "step": 8910
+    },
+    {
+      "epoch": 0.142576,
+      "grad_norm": 1.0,
+      "learning_rate": 8.643548387096774e-05,
+      "loss": 0.1988,
+      "step": 8911
+    },
+    {
+      "epoch": 0.142592,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.643387096774194e-05,
+      "loss": 0.2238,
+      "step": 8912
+    },
+    {
+      "epoch": 0.142608,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.643225806451614e-05,
+      "loss": 0.1427,
+      "step": 8913
+    },
+    {
+      "epoch": 0.142624,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.643064516129033e-05,
+      "loss": 0.1889,
+      "step": 8914
+    },
+    {
+      "epoch": 0.14264,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.642903225806453e-05,
+      "loss": 0.1624,
+      "step": 8915
+    },
+    {
+      "epoch": 0.142656,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.642741935483871e-05,
+      "loss": 0.1858,
+      "step": 8916
+    },
+    {
+      "epoch": 0.142672,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.642580645161291e-05,
+      "loss": 0.178,
+      "step": 8917
+    },
+    {
+      "epoch": 0.142688,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.64241935483871e-05,
+      "loss": 0.1516,
+      "step": 8918
+    },
+    {
+      "epoch": 0.142704,
+      "grad_norm": 0.625,
+      "learning_rate": 8.64225806451613e-05,
+      "loss": 0.1874,
+      "step": 8919
+    },
+    {
+      "epoch": 0.14272,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.642096774193549e-05,
+      "loss": 0.1666,
+      "step": 8920
+    },
+    {
+      "epoch": 0.142736,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.641935483870967e-05,
+      "loss": 0.1814,
+      "step": 8921
+    },
+    {
+      "epoch": 0.142752,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.641774193548387e-05,
+      "loss": 0.2178,
+      "step": 8922
+    },
+    {
+      "epoch": 0.142768,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.641612903225806e-05,
+      "loss": 0.1708,
+      "step": 8923
+    },
+    {
+      "epoch": 0.142784,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.641451612903226e-05,
+      "loss": 0.1756,
+      "step": 8924
+    },
+    {
+      "epoch": 0.1428,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.641290322580646e-05,
+      "loss": 0.151,
+      "step": 8925
+    },
+    {
+      "epoch": 0.142816,
+      "grad_norm": 1.3125,
+      "learning_rate": 8.641129032258066e-05,
+      "loss": 0.1654,
+      "step": 8926
+    },
+    {
+      "epoch": 0.142832,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.640967741935484e-05,
+      "loss": 0.1768,
+      "step": 8927
+    },
+    {
+      "epoch": 0.142848,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.640806451612904e-05,
+      "loss": 0.1296,
+      "step": 8928
+    },
+    {
+      "epoch": 0.142864,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.640645161290323e-05,
+      "loss": 0.1803,
+      "step": 8929
+    },
+    {
+      "epoch": 0.14288,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.640483870967743e-05,
+      "loss": 0.1826,
+      "step": 8930
+    },
+    {
+      "epoch": 0.142896,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.640322580645161e-05,
+      "loss": 0.137,
+      "step": 8931
+    },
+    {
+      "epoch": 0.142912,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.640161290322581e-05,
+      "loss": 0.1884,
+      "step": 8932
+    },
+    {
+      "epoch": 0.142928,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.64e-05,
+      "loss": 0.1798,
+      "step": 8933
+    },
+    {
+      "epoch": 0.142944,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.63983870967742e-05,
+      "loss": 0.174,
+      "step": 8934
+    },
+    {
+      "epoch": 0.14296,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.639677419354839e-05,
+      "loss": 0.152,
+      "step": 8935
+    },
+    {
+      "epoch": 0.142976,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.639516129032258e-05,
+      "loss": 0.1818,
+      "step": 8936
+    },
+    {
+      "epoch": 0.142992,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.639354838709678e-05,
+      "loss": 0.1653,
+      "step": 8937
+    },
+    {
+      "epoch": 0.143008,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.639193548387097e-05,
+      "loss": 0.1606,
+      "step": 8938
+    },
+    {
+      "epoch": 0.143024,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.639032258064517e-05,
+      "loss": 0.1349,
+      "step": 8939
+    },
+    {
+      "epoch": 0.14304,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.638870967741936e-05,
+      "loss": 0.1571,
+      "step": 8940
+    },
+    {
+      "epoch": 0.143056,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.638709677419356e-05,
+      "loss": 0.1793,
+      "step": 8941
+    },
+    {
+      "epoch": 0.143072,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.638548387096774e-05,
+      "loss": 0.16,
+      "step": 8942
+    },
+    {
+      "epoch": 0.143088,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.638387096774194e-05,
+      "loss": 0.1773,
+      "step": 8943
+    },
+    {
+      "epoch": 0.143104,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.638225806451613e-05,
+      "loss": 0.2113,
+      "step": 8944
+    },
+    {
+      "epoch": 0.14312,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.638064516129033e-05,
+      "loss": 0.1431,
+      "step": 8945
+    },
+    {
+      "epoch": 0.143136,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.637903225806451e-05,
+      "loss": 0.2174,
+      "step": 8946
+    },
+    {
+      "epoch": 0.143152,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.637741935483871e-05,
+      "loss": 0.1835,
+      "step": 8947
+    },
+    {
+      "epoch": 0.143168,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.637580645161291e-05,
+      "loss": 0.1792,
+      "step": 8948
+    },
+    {
+      "epoch": 0.143184,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.637419354838711e-05,
+      "loss": 0.1729,
+      "step": 8949
+    },
+    {
+      "epoch": 0.1432,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.63725806451613e-05,
+      "loss": 0.1306,
+      "step": 8950
+    },
+    {
+      "epoch": 0.143216,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.63709677419355e-05,
+      "loss": 0.148,
+      "step": 8951
+    },
+    {
+      "epoch": 0.143232,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.636935483870968e-05,
+      "loss": 0.153,
+      "step": 8952
+    },
+    {
+      "epoch": 0.143248,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.636774193548387e-05,
+      "loss": 0.1747,
+      "step": 8953
+    },
+    {
+      "epoch": 0.143264,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.636612903225807e-05,
+      "loss": 0.1747,
+      "step": 8954
+    },
+    {
+      "epoch": 0.14328,
+      "grad_norm": 1.28125,
+      "learning_rate": 8.636451612903226e-05,
+      "loss": 0.1894,
+      "step": 8955
+    },
+    {
+      "epoch": 0.143296,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.636290322580646e-05,
+      "loss": 0.178,
+      "step": 8956
+    },
+    {
+      "epoch": 0.143312,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.636129032258064e-05,
+      "loss": 0.1436,
+      "step": 8957
+    },
+    {
+      "epoch": 0.143328,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.635967741935484e-05,
+      "loss": 0.1881,
+      "step": 8958
+    },
+    {
+      "epoch": 0.143344,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.635806451612903e-05,
+      "loss": 0.2209,
+      "step": 8959
+    },
+    {
+      "epoch": 0.14336,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.635645161290323e-05,
+      "loss": 0.1893,
+      "step": 8960
+    },
+    {
+      "epoch": 0.143376,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.635483870967743e-05,
+      "loss": 0.2,
+      "step": 8961
+    },
+    {
+      "epoch": 0.143392,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.635322580645163e-05,
+      "loss": 0.1627,
+      "step": 8962
+    },
+    {
+      "epoch": 0.143408,
+      "grad_norm": 0.875,
+      "learning_rate": 8.635161290322581e-05,
+      "loss": 0.1586,
+      "step": 8963
+    },
+    {
+      "epoch": 0.143424,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.635000000000001e-05,
+      "loss": 0.1946,
+      "step": 8964
+    },
+    {
+      "epoch": 0.14344,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.63483870967742e-05,
+      "loss": 0.2041,
+      "step": 8965
+    },
+    {
+      "epoch": 0.143456,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.63467741935484e-05,
+      "loss": 0.1984,
+      "step": 8966
+    },
+    {
+      "epoch": 0.143472,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.634516129032258e-05,
+      "loss": 0.1885,
+      "step": 8967
+    },
+    {
+      "epoch": 0.143488,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.634354838709677e-05,
+      "loss": 0.1636,
+      "step": 8968
+    },
+    {
+      "epoch": 0.143504,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.634193548387097e-05,
+      "loss": 0.1936,
+      "step": 8969
+    },
+    {
+      "epoch": 0.14352,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.634032258064516e-05,
+      "loss": 0.2175,
+      "step": 8970
+    },
+    {
+      "epoch": 0.143536,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.633870967741936e-05,
+      "loss": 0.1774,
+      "step": 8971
+    },
+    {
+      "epoch": 0.143552,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.633709677419355e-05,
+      "loss": 0.1752,
+      "step": 8972
+    },
+    {
+      "epoch": 0.143568,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.633548387096775e-05,
+      "loss": 0.1561,
+      "step": 8973
+    },
+    {
+      "epoch": 0.143584,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.633387096774194e-05,
+      "loss": 0.1584,
+      "step": 8974
+    },
+    {
+      "epoch": 0.1436,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.633225806451614e-05,
+      "loss": 0.1557,
+      "step": 8975
+    },
+    {
+      "epoch": 0.143616,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.633064516129033e-05,
+      "loss": 0.1718,
+      "step": 8976
+    },
+    {
+      "epoch": 0.143632,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.632903225806453e-05,
+      "loss": 0.1434,
+      "step": 8977
+    },
+    {
+      "epoch": 0.143648,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.632741935483871e-05,
+      "loss": 0.183,
+      "step": 8978
+    },
+    {
+      "epoch": 0.143664,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.632580645161291e-05,
+      "loss": 0.1621,
+      "step": 8979
+    },
+    {
+      "epoch": 0.14368,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.63241935483871e-05,
+      "loss": 0.1537,
+      "step": 8980
+    },
+    {
+      "epoch": 0.143696,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.63225806451613e-05,
+      "loss": 0.1967,
+      "step": 8981
+    },
+    {
+      "epoch": 0.143712,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.632096774193548e-05,
+      "loss": 0.1575,
+      "step": 8982
+    },
+    {
+      "epoch": 0.143728,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.631935483870968e-05,
+      "loss": 0.2099,
+      "step": 8983
+    },
+    {
+      "epoch": 0.143744,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.631774193548388e-05,
+      "loss": 0.1713,
+      "step": 8984
+    },
+    {
+      "epoch": 0.14376,
+      "grad_norm": 0.875,
+      "learning_rate": 8.631612903225807e-05,
+      "loss": 0.1772,
+      "step": 8985
+    },
+    {
+      "epoch": 0.143776,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.631451612903227e-05,
+      "loss": 0.2033,
+      "step": 8986
+    },
+    {
+      "epoch": 0.143792,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.631290322580645e-05,
+      "loss": 0.1491,
+      "step": 8987
+    },
+    {
+      "epoch": 0.143808,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.631129032258065e-05,
+      "loss": 0.1727,
+      "step": 8988
+    },
+    {
+      "epoch": 0.143824,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.630967741935484e-05,
+      "loss": 0.1672,
+      "step": 8989
+    },
+    {
+      "epoch": 0.14384,
+      "grad_norm": 1.546875,
+      "learning_rate": 8.630806451612904e-05,
+      "loss": 0.1801,
+      "step": 8990
+    },
+    {
+      "epoch": 0.143856,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.630645161290323e-05,
+      "loss": 0.1544,
+      "step": 8991
+    },
+    {
+      "epoch": 0.143872,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.630483870967743e-05,
+      "loss": 0.1616,
+      "step": 8992
+    },
+    {
+      "epoch": 0.143888,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.630322580645161e-05,
+      "loss": 0.1664,
+      "step": 8993
+    },
+    {
+      "epoch": 0.143904,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.630161290322581e-05,
+      "loss": 0.1481,
+      "step": 8994
+    },
+    {
+      "epoch": 0.14392,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.63e-05,
+      "loss": 0.1587,
+      "step": 8995
+    },
+    {
+      "epoch": 0.143936,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.62983870967742e-05,
+      "loss": 0.1931,
+      "step": 8996
+    },
+    {
+      "epoch": 0.143952,
+      "grad_norm": 1.4296875,
+      "learning_rate": 8.62967741935484e-05,
+      "loss": 0.2035,
+      "step": 8997
+    },
+    {
+      "epoch": 0.143968,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.629516129032258e-05,
+      "loss": 0.1849,
+      "step": 8998
+    },
+    {
+      "epoch": 0.143984,
+      "grad_norm": 1.375,
+      "learning_rate": 8.629354838709678e-05,
+      "loss": 0.1773,
+      "step": 8999
+    },
+    {
+      "epoch": 0.144,
+      "grad_norm": 1.0,
+      "learning_rate": 8.629193548387097e-05,
+      "loss": 0.1586,
+      "step": 9000
+    },
+    {
+      "epoch": 0.144016,
+      "grad_norm": 1.28125,
+      "learning_rate": 8.629032258064517e-05,
+      "loss": 0.1694,
+      "step": 9001
+    },
+    {
+      "epoch": 0.144032,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.628870967741935e-05,
+      "loss": 0.1646,
+      "step": 9002
+    },
+    {
+      "epoch": 0.144048,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.628709677419355e-05,
+      "loss": 0.1349,
+      "step": 9003
+    },
+    {
+      "epoch": 0.144064,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.628548387096774e-05,
+      "loss": 0.2061,
+      "step": 9004
+    },
+    {
+      "epoch": 0.14408,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.628387096774194e-05,
+      "loss": 0.1602,
+      "step": 9005
+    },
+    {
+      "epoch": 0.144096,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.628225806451613e-05,
+      "loss": 0.1896,
+      "step": 9006
+    },
+    {
+      "epoch": 0.144112,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.628064516129032e-05,
+      "loss": 0.2073,
+      "step": 9007
+    },
+    {
+      "epoch": 0.144128,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.627903225806452e-05,
+      "loss": 0.1973,
+      "step": 9008
+    },
+    {
+      "epoch": 0.144144,
+      "grad_norm": 1.4453125,
+      "learning_rate": 8.627741935483872e-05,
+      "loss": 0.18,
+      "step": 9009
+    },
+    {
+      "epoch": 0.14416,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.627580645161291e-05,
+      "loss": 0.1615,
+      "step": 9010
+    },
+    {
+      "epoch": 0.144176,
+      "grad_norm": 1.28125,
+      "learning_rate": 8.627419354838711e-05,
+      "loss": 0.2138,
+      "step": 9011
+    },
+    {
+      "epoch": 0.144192,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.62725806451613e-05,
+      "loss": 0.1623,
+      "step": 9012
+    },
+    {
+      "epoch": 0.144208,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.62709677419355e-05,
+      "loss": 0.1743,
+      "step": 9013
+    },
+    {
+      "epoch": 0.144224,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.626935483870968e-05,
+      "loss": 0.1758,
+      "step": 9014
+    },
+    {
+      "epoch": 0.14424,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.626774193548387e-05,
+      "loss": 0.1532,
+      "step": 9015
+    },
+    {
+      "epoch": 0.144256,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.626612903225807e-05,
+      "loss": 0.1658,
+      "step": 9016
+    },
+    {
+      "epoch": 0.144272,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.626451612903225e-05,
+      "loss": 0.1666,
+      "step": 9017
+    },
+    {
+      "epoch": 0.144288,
+      "grad_norm": 1.3515625,
+      "learning_rate": 8.626290322580645e-05,
+      "loss": 0.1226,
+      "step": 9018
+    },
+    {
+      "epoch": 0.144304,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.626129032258064e-05,
+      "loss": 0.2015,
+      "step": 9019
+    },
+    {
+      "epoch": 0.14432,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.625967741935484e-05,
+      "loss": 0.1805,
+      "step": 9020
+    },
+    {
+      "epoch": 0.144336,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.625806451612904e-05,
+      "loss": 0.1721,
+      "step": 9021
+    },
+    {
+      "epoch": 0.144352,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.625645161290324e-05,
+      "loss": 0.178,
+      "step": 9022
+    },
+    {
+      "epoch": 0.144368,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.625483870967742e-05,
+      "loss": 0.1761,
+      "step": 9023
+    },
+    {
+      "epoch": 0.144384,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.625322580645162e-05,
+      "loss": 0.188,
+      "step": 9024
+    },
+    {
+      "epoch": 0.1444,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.625161290322581e-05,
+      "loss": 0.2036,
+      "step": 9025
+    },
+    {
+      "epoch": 0.144416,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.625000000000001e-05,
+      "loss": 0.1882,
+      "step": 9026
+    },
+    {
+      "epoch": 0.144432,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.62483870967742e-05,
+      "loss": 0.1685,
+      "step": 9027
+    },
+    {
+      "epoch": 0.144448,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.62467741935484e-05,
+      "loss": 0.1803,
+      "step": 9028
+    },
+    {
+      "epoch": 0.144464,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.624516129032258e-05,
+      "loss": 0.1597,
+      "step": 9029
+    },
+    {
+      "epoch": 0.14448,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.624354838709677e-05,
+      "loss": 0.1694,
+      "step": 9030
+    },
+    {
+      "epoch": 0.144496,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.624193548387097e-05,
+      "loss": 0.222,
+      "step": 9031
+    },
+    {
+      "epoch": 0.144512,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.624032258064517e-05,
+      "loss": 0.1712,
+      "step": 9032
+    },
+    {
+      "epoch": 0.144528,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.623870967741937e-05,
+      "loss": 0.1484,
+      "step": 9033
+    },
+    {
+      "epoch": 0.144544,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.623709677419355e-05,
+      "loss": 0.1818,
+      "step": 9034
+    },
+    {
+      "epoch": 0.14456,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.623548387096775e-05,
+      "loss": 0.1432,
+      "step": 9035
+    },
+    {
+      "epoch": 0.144576,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.623387096774194e-05,
+      "loss": 0.2101,
+      "step": 9036
+    },
+    {
+      "epoch": 0.144592,
+      "grad_norm": 1.234375,
+      "learning_rate": 8.623225806451614e-05,
+      "loss": 0.1907,
+      "step": 9037
+    },
+    {
+      "epoch": 0.144608,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.623064516129032e-05,
+      "loss": 0.1585,
+      "step": 9038
+    },
+    {
+      "epoch": 0.144624,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.622903225806452e-05,
+      "loss": 0.2153,
+      "step": 9039
+    },
+    {
+      "epoch": 0.14464,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.622741935483871e-05,
+      "loss": 0.1721,
+      "step": 9040
+    },
+    {
+      "epoch": 0.144656,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.622580645161291e-05,
+      "loss": 0.1413,
+      "step": 9041
+    },
+    {
+      "epoch": 0.144672,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.62241935483871e-05,
+      "loss": 0.1656,
+      "step": 9042
+    },
+    {
+      "epoch": 0.144688,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.62225806451613e-05,
+      "loss": 0.1771,
+      "step": 9043
+    },
+    {
+      "epoch": 0.144704,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.62209677419355e-05,
+      "loss": 0.2089,
+      "step": 9044
+    },
+    {
+      "epoch": 0.14472,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.621935483870968e-05,
+      "loss": 0.2009,
+      "step": 9045
+    },
+    {
+      "epoch": 0.144736,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.621774193548388e-05,
+      "loss": 0.1622,
+      "step": 9046
+    },
+    {
+      "epoch": 0.144752,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.621612903225807e-05,
+      "loss": 0.1583,
+      "step": 9047
+    },
+    {
+      "epoch": 0.144768,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.621451612903227e-05,
+      "loss": 0.1679,
+      "step": 9048
+    },
+    {
+      "epoch": 0.144784,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.621290322580645e-05,
+      "loss": 0.1884,
+      "step": 9049
+    },
+    {
+      "epoch": 0.1448,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.621129032258065e-05,
+      "loss": 0.1519,
+      "step": 9050
+    },
+    {
+      "epoch": 0.144816,
+      "grad_norm": 1.2890625,
+      "learning_rate": 8.620967741935484e-05,
+      "loss": 0.2044,
+      "step": 9051
+    },
+    {
+      "epoch": 0.144832,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.620806451612904e-05,
+      "loss": 0.1578,
+      "step": 9052
+    },
+    {
+      "epoch": 0.144848,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.620645161290322e-05,
+      "loss": 0.1637,
+      "step": 9053
+    },
+    {
+      "epoch": 0.144864,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.620483870967742e-05,
+      "loss": 0.1847,
+      "step": 9054
+    },
+    {
+      "epoch": 0.14488,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.620322580645161e-05,
+      "loss": 0.1892,
+      "step": 9055
+    },
+    {
+      "epoch": 0.144896,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.620161290322581e-05,
+      "loss": 0.1402,
+      "step": 9056
+    },
+    {
+      "epoch": 0.144912,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.620000000000001e-05,
+      "loss": 0.1851,
+      "step": 9057
+    },
+    {
+      "epoch": 0.144928,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.619838709677421e-05,
+      "loss": 0.1735,
+      "step": 9058
+    },
+    {
+      "epoch": 0.144944,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.61967741935484e-05,
+      "loss": 0.227,
+      "step": 9059
+    },
+    {
+      "epoch": 0.14496,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.61951612903226e-05,
+      "loss": 0.1632,
+      "step": 9060
+    },
+    {
+      "epoch": 0.144976,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.619354838709678e-05,
+      "loss": 0.2003,
+      "step": 9061
+    },
+    {
+      "epoch": 0.144992,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.619193548387097e-05,
+      "loss": 0.1489,
+      "step": 9062
+    },
+    {
+      "epoch": 0.145008,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.619032258064517e-05,
+      "loss": 0.2041,
+      "step": 9063
+    },
+    {
+      "epoch": 0.145024,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.618870967741935e-05,
+      "loss": 0.1682,
+      "step": 9064
+    },
+    {
+      "epoch": 0.14504,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.618709677419355e-05,
+      "loss": 0.1485,
+      "step": 9065
+    },
+    {
+      "epoch": 0.145056,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.618548387096774e-05,
+      "loss": 0.1767,
+      "step": 9066
+    },
+    {
+      "epoch": 0.145072,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.618387096774194e-05,
+      "loss": 0.1837,
+      "step": 9067
+    },
+    {
+      "epoch": 0.145088,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.618225806451614e-05,
+      "loss": 0.1785,
+      "step": 9068
+    },
+    {
+      "epoch": 0.145104,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.618064516129034e-05,
+      "loss": 0.1501,
+      "step": 9069
+    },
+    {
+      "epoch": 0.14512,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.617903225806452e-05,
+      "loss": 0.1713,
+      "step": 9070
+    },
+    {
+      "epoch": 0.145136,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.617741935483872e-05,
+      "loss": 0.1923,
+      "step": 9071
+    },
+    {
+      "epoch": 0.145152,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.617580645161291e-05,
+      "loss": 0.1849,
+      "step": 9072
+    },
+    {
+      "epoch": 0.145168,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.617419354838711e-05,
+      "loss": 0.1631,
+      "step": 9073
+    },
+    {
+      "epoch": 0.145184,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.61725806451613e-05,
+      "loss": 0.1517,
+      "step": 9074
+    },
+    {
+      "epoch": 0.1452,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.617096774193549e-05,
+      "loss": 0.1913,
+      "step": 9075
+    },
+    {
+      "epoch": 0.145216,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.616935483870968e-05,
+      "loss": 0.1758,
+      "step": 9076
+    },
+    {
+      "epoch": 0.145232,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.616774193548387e-05,
+      "loss": 0.188,
+      "step": 9077
+    },
+    {
+      "epoch": 0.145248,
+      "grad_norm": 1.0,
+      "learning_rate": 8.616612903225806e-05,
+      "loss": 0.185,
+      "step": 9078
+    },
+    {
+      "epoch": 0.145264,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.616451612903226e-05,
+      "loss": 0.1797,
+      "step": 9079
+    },
+    {
+      "epoch": 0.14528,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.616290322580645e-05,
+      "loss": 0.1458,
+      "step": 9080
+    },
+    {
+      "epoch": 0.145296,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.616129032258065e-05,
+      "loss": 0.2259,
+      "step": 9081
+    },
+    {
+      "epoch": 0.145312,
+      "grad_norm": 1.296875,
+      "learning_rate": 8.615967741935485e-05,
+      "loss": 0.1769,
+      "step": 9082
+    },
+    {
+      "epoch": 0.145328,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.615806451612904e-05,
+      "loss": 0.1626,
+      "step": 9083
+    },
+    {
+      "epoch": 0.145344,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.615645161290324e-05,
+      "loss": 0.1945,
+      "step": 9084
+    },
+    {
+      "epoch": 0.14536,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.615483870967742e-05,
+      "loss": 0.1737,
+      "step": 9085
+    },
+    {
+      "epoch": 0.145376,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.615322580645162e-05,
+      "loss": 0.1617,
+      "step": 9086
+    },
+    {
+      "epoch": 0.145392,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.615161290322581e-05,
+      "loss": 0.1469,
+      "step": 9087
+    },
+    {
+      "epoch": 0.145408,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.615000000000001e-05,
+      "loss": 0.1466,
+      "step": 9088
+    },
+    {
+      "epoch": 0.145424,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.614838709677419e-05,
+      "loss": 0.1693,
+      "step": 9089
+    },
+    {
+      "epoch": 0.14544,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.614677419354839e-05,
+      "loss": 0.156,
+      "step": 9090
+    },
+    {
+      "epoch": 0.145456,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.614516129032258e-05,
+      "loss": 0.1483,
+      "step": 9091
+    },
+    {
+      "epoch": 0.145472,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.614354838709678e-05,
+      "loss": 0.1953,
+      "step": 9092
+    },
+    {
+      "epoch": 0.145488,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.614193548387098e-05,
+      "loss": 0.1553,
+      "step": 9093
+    },
+    {
+      "epoch": 0.145504,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.614032258064516e-05,
+      "loss": 0.17,
+      "step": 9094
+    },
+    {
+      "epoch": 0.14552,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.613870967741936e-05,
+      "loss": 0.1781,
+      "step": 9095
+    },
+    {
+      "epoch": 0.145536,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.613709677419355e-05,
+      "loss": 0.176,
+      "step": 9096
+    },
+    {
+      "epoch": 0.145552,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.613548387096775e-05,
+      "loss": 0.1877,
+      "step": 9097
+    },
+    {
+      "epoch": 0.145568,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.613387096774194e-05,
+      "loss": 0.1797,
+      "step": 9098
+    },
+    {
+      "epoch": 0.145584,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.613225806451614e-05,
+      "loss": 0.1686,
+      "step": 9099
+    },
+    {
+      "epoch": 0.1456,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.613064516129032e-05,
+      "loss": 0.1359,
+      "step": 9100
+    },
+    {
+      "epoch": 0.145616,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.612903225806452e-05,
+      "loss": 0.1479,
+      "step": 9101
+    },
+    {
+      "epoch": 0.145632,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.612741935483871e-05,
+      "loss": 0.1775,
+      "step": 9102
+    },
+    {
+      "epoch": 0.145648,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.61258064516129e-05,
+      "loss": 0.1886,
+      "step": 9103
+    },
+    {
+      "epoch": 0.145664,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.61241935483871e-05,
+      "loss": 0.1499,
+      "step": 9104
+    },
+    {
+      "epoch": 0.14568,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.61225806451613e-05,
+      "loss": 0.1142,
+      "step": 9105
+    },
+    {
+      "epoch": 0.145696,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.612096774193549e-05,
+      "loss": 0.1547,
+      "step": 9106
+    },
+    {
+      "epoch": 0.145712,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.611935483870968e-05,
+      "loss": 0.1607,
+      "step": 9107
+    },
+    {
+      "epoch": 0.145728,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.611774193548388e-05,
+      "loss": 0.1358,
+      "step": 9108
+    },
+    {
+      "epoch": 0.145744,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.611612903225806e-05,
+      "loss": 0.144,
+      "step": 9109
+    },
+    {
+      "epoch": 0.14576,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.611451612903226e-05,
+      "loss": 0.1743,
+      "step": 9110
+    },
+    {
+      "epoch": 0.145776,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.611290322580645e-05,
+      "loss": 0.2062,
+      "step": 9111
+    },
+    {
+      "epoch": 0.145792,
+      "grad_norm": 0.478515625,
+      "learning_rate": 8.611129032258065e-05,
+      "loss": 0.1362,
+      "step": 9112
+    },
+    {
+      "epoch": 0.145808,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.610967741935484e-05,
+      "loss": 0.2084,
+      "step": 9113
+    },
+    {
+      "epoch": 0.145824,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.610806451612903e-05,
+      "loss": 0.1387,
+      "step": 9114
+    },
+    {
+      "epoch": 0.14584,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.610645161290322e-05,
+      "loss": 0.1736,
+      "step": 9115
+    },
+    {
+      "epoch": 0.145856,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.610483870967742e-05,
+      "loss": 0.1942,
+      "step": 9116
+    },
+    {
+      "epoch": 0.145872,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.610322580645162e-05,
+      "loss": 0.1565,
+      "step": 9117
+    },
+    {
+      "epoch": 0.145888,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.610161290322582e-05,
+      "loss": 0.1864,
+      "step": 9118
+    },
+    {
+      "epoch": 0.145904,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.61e-05,
+      "loss": 0.1698,
+      "step": 9119
+    },
+    {
+      "epoch": 0.14592,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.60983870967742e-05,
+      "loss": 0.1601,
+      "step": 9120
+    },
+    {
+      "epoch": 0.145936,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.609677419354839e-05,
+      "loss": 0.1421,
+      "step": 9121
+    },
+    {
+      "epoch": 0.145952,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.609516129032259e-05,
+      "loss": 0.1665,
+      "step": 9122
+    },
+    {
+      "epoch": 0.145968,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.609354838709678e-05,
+      "loss": 0.1321,
+      "step": 9123
+    },
+    {
+      "epoch": 0.145984,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.609193548387096e-05,
+      "loss": 0.1826,
+      "step": 9124
+    },
+    {
+      "epoch": 0.146,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.609032258064516e-05,
+      "loss": 0.1718,
+      "step": 9125
+    },
+    {
+      "epoch": 0.146016,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.608870967741935e-05,
+      "loss": 0.1622,
+      "step": 9126
+    },
+    {
+      "epoch": 0.146032,
+      "grad_norm": 0.95703125,
+      "learning_rate": 8.608709677419355e-05,
+      "loss": 0.1679,
+      "step": 9127
+    },
+    {
+      "epoch": 0.146048,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.608548387096775e-05,
+      "loss": 0.1671,
+      "step": 9128
+    },
+    {
+      "epoch": 0.146064,
+      "grad_norm": 1.46875,
+      "learning_rate": 8.608387096774195e-05,
+      "loss": 0.1829,
+      "step": 9129
+    },
+    {
+      "epoch": 0.14608,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.608225806451613e-05,
+      "loss": 0.1739,
+      "step": 9130
+    },
+    {
+      "epoch": 0.146096,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.608064516129033e-05,
+      "loss": 0.158,
+      "step": 9131
+    },
+    {
+      "epoch": 0.146112,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.607903225806452e-05,
+      "loss": 0.1863,
+      "step": 9132
+    },
+    {
+      "epoch": 0.146128,
+      "grad_norm": 1.265625,
+      "learning_rate": 8.607741935483872e-05,
+      "loss": 0.1724,
+      "step": 9133
+    },
+    {
+      "epoch": 0.146144,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.60758064516129e-05,
+      "loss": 0.1879,
+      "step": 9134
+    },
+    {
+      "epoch": 0.14616,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.60741935483871e-05,
+      "loss": 0.189,
+      "step": 9135
+    },
+    {
+      "epoch": 0.146176,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.607258064516129e-05,
+      "loss": 0.1519,
+      "step": 9136
+    },
+    {
+      "epoch": 0.146192,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.607096774193549e-05,
+      "loss": 0.1601,
+      "step": 9137
+    },
+    {
+      "epoch": 0.146208,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.606935483870968e-05,
+      "loss": 0.1603,
+      "step": 9138
+    },
+    {
+      "epoch": 0.146224,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.606774193548388e-05,
+      "loss": 0.2053,
+      "step": 9139
+    },
+    {
+      "epoch": 0.14624,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.606612903225808e-05,
+      "loss": 0.1391,
+      "step": 9140
+    },
+    {
+      "epoch": 0.146256,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.606451612903226e-05,
+      "loss": 0.1749,
+      "step": 9141
+    },
+    {
+      "epoch": 0.146272,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.606290322580646e-05,
+      "loss": 0.1781,
+      "step": 9142
+    },
+    {
+      "epoch": 0.146288,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.606129032258065e-05,
+      "loss": 0.1997,
+      "step": 9143
+    },
+    {
+      "epoch": 0.146304,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.605967741935485e-05,
+      "loss": 0.1769,
+      "step": 9144
+    },
+    {
+      "epoch": 0.14632,
+      "grad_norm": 0.498046875,
+      "learning_rate": 8.605806451612903e-05,
+      "loss": 0.1495,
+      "step": 9145
+    },
+    {
+      "epoch": 0.146336,
+      "grad_norm": 1.4296875,
+      "learning_rate": 8.605645161290323e-05,
+      "loss": 0.2052,
+      "step": 9146
+    },
+    {
+      "epoch": 0.146352,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.605483870967742e-05,
+      "loss": 0.1731,
+      "step": 9147
+    },
+    {
+      "epoch": 0.146368,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.605322580645162e-05,
+      "loss": 0.1395,
+      "step": 9148
+    },
+    {
+      "epoch": 0.146384,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.60516129032258e-05,
+      "loss": 0.1724,
+      "step": 9149
+    },
+    {
+      "epoch": 0.1464,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.605e-05,
+      "loss": 0.1602,
+      "step": 9150
+    },
+    {
+      "epoch": 0.146416,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.604838709677419e-05,
+      "loss": 0.1528,
+      "step": 9151
+    },
+    {
+      "epoch": 0.146432,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.604677419354839e-05,
+      "loss": 0.1521,
+      "step": 9152
+    },
+    {
+      "epoch": 0.146448,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.604516129032259e-05,
+      "loss": 0.155,
+      "step": 9153
+    },
+    {
+      "epoch": 0.146464,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.604354838709678e-05,
+      "loss": 0.1971,
+      "step": 9154
+    },
+    {
+      "epoch": 0.14648,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.604193548387098e-05,
+      "loss": 0.1632,
+      "step": 9155
+    },
+    {
+      "epoch": 0.146496,
+      "grad_norm": 1.265625,
+      "learning_rate": 8.604032258064516e-05,
+      "loss": 0.1798,
+      "step": 9156
+    },
+    {
+      "epoch": 0.146512,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.603870967741936e-05,
+      "loss": 0.1858,
+      "step": 9157
+    },
+    {
+      "epoch": 0.146528,
+      "grad_norm": 1.359375,
+      "learning_rate": 8.603709677419355e-05,
+      "loss": 0.2101,
+      "step": 9158
+    },
+    {
+      "epoch": 0.146544,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.603548387096775e-05,
+      "loss": 0.1515,
+      "step": 9159
+    },
+    {
+      "epoch": 0.14656,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.603387096774193e-05,
+      "loss": 0.1952,
+      "step": 9160
+    },
+    {
+      "epoch": 0.146576,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.603225806451613e-05,
+      "loss": 0.228,
+      "step": 9161
+    },
+    {
+      "epoch": 0.146592,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.603064516129032e-05,
+      "loss": 0.1566,
+      "step": 9162
+    },
+    {
+      "epoch": 0.146608,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.602903225806452e-05,
+      "loss": 0.1369,
+      "step": 9163
+    },
+    {
+      "epoch": 0.146624,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.602741935483872e-05,
+      "loss": 0.216,
+      "step": 9164
+    },
+    {
+      "epoch": 0.14664,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.602580645161292e-05,
+      "loss": 0.1753,
+      "step": 9165
+    },
+    {
+      "epoch": 0.146656,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.60241935483871e-05,
+      "loss": 0.1843,
+      "step": 9166
+    },
+    {
+      "epoch": 0.146672,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.60225806451613e-05,
+      "loss": 0.1779,
+      "step": 9167
+    },
+    {
+      "epoch": 0.146688,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.602096774193549e-05,
+      "loss": 0.1569,
+      "step": 9168
+    },
+    {
+      "epoch": 0.146704,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.601935483870969e-05,
+      "loss": 0.2055,
+      "step": 9169
+    },
+    {
+      "epoch": 0.14672,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.601774193548388e-05,
+      "loss": 0.1594,
+      "step": 9170
+    },
+    {
+      "epoch": 0.146736,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.601612903225806e-05,
+      "loss": 0.1537,
+      "step": 9171
+    },
+    {
+      "epoch": 0.146752,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.601451612903226e-05,
+      "loss": 0.2112,
+      "step": 9172
+    },
+    {
+      "epoch": 0.146768,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.601290322580645e-05,
+      "loss": 0.1791,
+      "step": 9173
+    },
+    {
+      "epoch": 0.146784,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.601129032258065e-05,
+      "loss": 0.1534,
+      "step": 9174
+    },
+    {
+      "epoch": 0.1468,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.600967741935483e-05,
+      "loss": 0.1721,
+      "step": 9175
+    },
+    {
+      "epoch": 0.146816,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.600806451612903e-05,
+      "loss": 0.2203,
+      "step": 9176
+    },
+    {
+      "epoch": 0.146832,
+      "grad_norm": 0.4296875,
+      "learning_rate": 8.600645161290323e-05,
+      "loss": 0.1611,
+      "step": 9177
+    },
+    {
+      "epoch": 0.146848,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.600483870967743e-05,
+      "loss": 0.141,
+      "step": 9178
+    },
+    {
+      "epoch": 0.146864,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.600322580645162e-05,
+      "loss": 0.188,
+      "step": 9179
+    },
+    {
+      "epoch": 0.14688,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.600161290322582e-05,
+      "loss": 0.1568,
+      "step": 9180
+    },
+    {
+      "epoch": 0.146896,
+      "grad_norm": 1.2734375,
+      "learning_rate": 8.6e-05,
+      "loss": 0.2025,
+      "step": 9181
+    },
+    {
+      "epoch": 0.146912,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.59983870967742e-05,
+      "loss": 0.1848,
+      "step": 9182
+    },
+    {
+      "epoch": 0.146928,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.599677419354839e-05,
+      "loss": 0.1272,
+      "step": 9183
+    },
+    {
+      "epoch": 0.146944,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.599516129032259e-05,
+      "loss": 0.1745,
+      "step": 9184
+    },
+    {
+      "epoch": 0.14696,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.599354838709677e-05,
+      "loss": 0.2325,
+      "step": 9185
+    },
+    {
+      "epoch": 0.146976,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.599193548387096e-05,
+      "loss": 0.1435,
+      "step": 9186
+    },
+    {
+      "epoch": 0.146992,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.599032258064516e-05,
+      "loss": 0.1596,
+      "step": 9187
+    },
+    {
+      "epoch": 0.147008,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.598870967741936e-05,
+      "loss": 0.1722,
+      "step": 9188
+    },
+    {
+      "epoch": 0.147024,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.598709677419356e-05,
+      "loss": 0.1948,
+      "step": 9189
+    },
+    {
+      "epoch": 0.14704,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.598548387096775e-05,
+      "loss": 0.1498,
+      "step": 9190
+    },
+    {
+      "epoch": 0.147056,
+      "grad_norm": 1.4140625,
+      "learning_rate": 8.598387096774195e-05,
+      "loss": 0.1913,
+      "step": 9191
+    },
+    {
+      "epoch": 0.147072,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.598225806451613e-05,
+      "loss": 0.1951,
+      "step": 9192
+    },
+    {
+      "epoch": 0.147088,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.598064516129033e-05,
+      "loss": 0.179,
+      "step": 9193
+    },
+    {
+      "epoch": 0.147104,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.597903225806452e-05,
+      "loss": 0.1499,
+      "step": 9194
+    },
+    {
+      "epoch": 0.14712,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.597741935483872e-05,
+      "loss": 0.1946,
+      "step": 9195
+    },
+    {
+      "epoch": 0.147136,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.59758064516129e-05,
+      "loss": 0.1702,
+      "step": 9196
+    },
+    {
+      "epoch": 0.147152,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.59741935483871e-05,
+      "loss": 0.1762,
+      "step": 9197
+    },
+    {
+      "epoch": 0.147168,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.597258064516129e-05,
+      "loss": 0.1879,
+      "step": 9198
+    },
+    {
+      "epoch": 0.147184,
+      "grad_norm": 0.75,
+      "learning_rate": 8.597096774193549e-05,
+      "loss": 0.1859,
+      "step": 9199
+    },
+    {
+      "epoch": 0.1472,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.596935483870969e-05,
+      "loss": 0.1545,
+      "step": 9200
+    },
+    {
+      "epoch": 0.147216,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.596774193548387e-05,
+      "loss": 0.1574,
+      "step": 9201
+    },
+    {
+      "epoch": 0.147232,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.596612903225807e-05,
+      "loss": 0.1702,
+      "step": 9202
+    },
+    {
+      "epoch": 0.147248,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.596451612903226e-05,
+      "loss": 0.1675,
+      "step": 9203
+    },
+    {
+      "epoch": 0.147264,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.596290322580646e-05,
+      "loss": 0.2006,
+      "step": 9204
+    },
+    {
+      "epoch": 0.14728,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.596129032258065e-05,
+      "loss": 0.1428,
+      "step": 9205
+    },
+    {
+      "epoch": 0.147296,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.595967741935484e-05,
+      "loss": 0.2267,
+      "step": 9206
+    },
+    {
+      "epoch": 0.147312,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.595806451612903e-05,
+      "loss": 0.1549,
+      "step": 9207
+    },
+    {
+      "epoch": 0.147328,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.595645161290323e-05,
+      "loss": 0.1386,
+      "step": 9208
+    },
+    {
+      "epoch": 0.147344,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.595483870967742e-05,
+      "loss": 0.1782,
+      "step": 9209
+    },
+    {
+      "epoch": 0.14736,
+      "grad_norm": 0.75,
+      "learning_rate": 8.595322580645162e-05,
+      "loss": 0.2008,
+      "step": 9210
+    },
+    {
+      "epoch": 0.147376,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.59516129032258e-05,
+      "loss": 0.1717,
+      "step": 9211
+    },
+    {
+      "epoch": 0.147392,
+      "grad_norm": 0.75,
+      "learning_rate": 8.595e-05,
+      "loss": 0.1618,
+      "step": 9212
+    },
+    {
+      "epoch": 0.147408,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.59483870967742e-05,
+      "loss": 0.2014,
+      "step": 9213
+    },
+    {
+      "epoch": 0.147424,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.59467741935484e-05,
+      "loss": 0.1631,
+      "step": 9214
+    },
+    {
+      "epoch": 0.14744,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.594516129032259e-05,
+      "loss": 0.1467,
+      "step": 9215
+    },
+    {
+      "epoch": 0.147456,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.594354838709677e-05,
+      "loss": 0.1764,
+      "step": 9216
+    },
+    {
+      "epoch": 0.147472,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.594193548387097e-05,
+      "loss": 0.1736,
+      "step": 9217
+    },
+    {
+      "epoch": 0.147488,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.594032258064516e-05,
+      "loss": 0.1858,
+      "step": 9218
+    },
+    {
+      "epoch": 0.147504,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.593870967741936e-05,
+      "loss": 0.1631,
+      "step": 9219
+    },
+    {
+      "epoch": 0.14752,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.593709677419354e-05,
+      "loss": 0.2124,
+      "step": 9220
+    },
+    {
+      "epoch": 0.147536,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.593548387096774e-05,
+      "loss": 0.2131,
+      "step": 9221
+    },
+    {
+      "epoch": 0.147552,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.593387096774193e-05,
+      "loss": 0.1474,
+      "step": 9222
+    },
+    {
+      "epoch": 0.147568,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.593225806451613e-05,
+      "loss": 0.1725,
+      "step": 9223
+    },
+    {
+      "epoch": 0.147584,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.593064516129033e-05,
+      "loss": 0.2053,
+      "step": 9224
+    },
+    {
+      "epoch": 0.1476,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.592903225806453e-05,
+      "loss": 0.1725,
+      "step": 9225
+    },
+    {
+      "epoch": 0.147616,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.592741935483872e-05,
+      "loss": 0.1889,
+      "step": 9226
+    },
+    {
+      "epoch": 0.147632,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.592580645161292e-05,
+      "loss": 0.1908,
+      "step": 9227
+    },
+    {
+      "epoch": 0.147648,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.59241935483871e-05,
+      "loss": 0.1545,
+      "step": 9228
+    },
+    {
+      "epoch": 0.147664,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.59225806451613e-05,
+      "loss": 0.1761,
+      "step": 9229
+    },
+    {
+      "epoch": 0.14768,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.592096774193549e-05,
+      "loss": 0.1808,
+      "step": 9230
+    },
+    {
+      "epoch": 0.147696,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.591935483870969e-05,
+      "loss": 0.1533,
+      "step": 9231
+    },
+    {
+      "epoch": 0.147712,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.591774193548387e-05,
+      "loss": 0.1519,
+      "step": 9232
+    },
+    {
+      "epoch": 0.147728,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.591612903225806e-05,
+      "loss": 0.2157,
+      "step": 9233
+    },
+    {
+      "epoch": 0.147744,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.591451612903226e-05,
+      "loss": 0.1692,
+      "step": 9234
+    },
+    {
+      "epoch": 0.14776,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.591290322580646e-05,
+      "loss": 0.1762,
+      "step": 9235
+    },
+    {
+      "epoch": 0.147776,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.591129032258066e-05,
+      "loss": 0.1669,
+      "step": 9236
+    },
+    {
+      "epoch": 0.147792,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.590967741935484e-05,
+      "loss": 0.1785,
+      "step": 9237
+    },
+    {
+      "epoch": 0.147808,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.590806451612904e-05,
+      "loss": 0.148,
+      "step": 9238
+    },
+    {
+      "epoch": 0.147824,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.590645161290323e-05,
+      "loss": 0.175,
+      "step": 9239
+    },
+    {
+      "epoch": 0.14784,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.590483870967743e-05,
+      "loss": 0.179,
+      "step": 9240
+    },
+    {
+      "epoch": 0.147856,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.590322580645162e-05,
+      "loss": 0.1422,
+      "step": 9241
+    },
+    {
+      "epoch": 0.147872,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.590161290322581e-05,
+      "loss": 0.1632,
+      "step": 9242
+    },
+    {
+      "epoch": 0.147888,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.59e-05,
+      "loss": 0.1636,
+      "step": 9243
+    },
+    {
+      "epoch": 0.147904,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.58983870967742e-05,
+      "loss": 0.1696,
+      "step": 9244
+    },
+    {
+      "epoch": 0.14792,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.589677419354839e-05,
+      "loss": 0.1954,
+      "step": 9245
+    },
+    {
+      "epoch": 0.147936,
+      "grad_norm": 0.75,
+      "learning_rate": 8.589516129032259e-05,
+      "loss": 0.1825,
+      "step": 9246
+    },
+    {
+      "epoch": 0.147952,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.589354838709677e-05,
+      "loss": 0.2036,
+      "step": 9247
+    },
+    {
+      "epoch": 0.147968,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.589193548387097e-05,
+      "loss": 0.1479,
+      "step": 9248
+    },
+    {
+      "epoch": 0.147984,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.589032258064517e-05,
+      "loss": 0.2114,
+      "step": 9249
+    },
+    {
+      "epoch": 0.148,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.588870967741936e-05,
+      "loss": 0.1842,
+      "step": 9250
+    },
+    {
+      "epoch": 0.148016,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.588709677419356e-05,
+      "loss": 0.149,
+      "step": 9251
+    },
+    {
+      "epoch": 0.148032,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.588548387096774e-05,
+      "loss": 0.1695,
+      "step": 9252
+    },
+    {
+      "epoch": 0.148048,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.588387096774194e-05,
+      "loss": 0.188,
+      "step": 9253
+    },
+    {
+      "epoch": 0.148064,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.588225806451613e-05,
+      "loss": 0.2128,
+      "step": 9254
+    },
+    {
+      "epoch": 0.14808,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.588064516129033e-05,
+      "loss": 0.1754,
+      "step": 9255
+    },
+    {
+      "epoch": 0.148096,
+      "grad_norm": 0.95703125,
+      "learning_rate": 8.587903225806451e-05,
+      "loss": 0.1735,
+      "step": 9256
+    },
+    {
+      "epoch": 0.148112,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.587741935483871e-05,
+      "loss": 0.19,
+      "step": 9257
+    },
+    {
+      "epoch": 0.148128,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.58758064516129e-05,
+      "loss": 0.1646,
+      "step": 9258
+    },
+    {
+      "epoch": 0.148144,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.58741935483871e-05,
+      "loss": 0.2085,
+      "step": 9259
+    },
+    {
+      "epoch": 0.14816,
+      "grad_norm": 1.4375,
+      "learning_rate": 8.58725806451613e-05,
+      "loss": 0.2251,
+      "step": 9260
+    },
+    {
+      "epoch": 0.148176,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.58709677419355e-05,
+      "loss": 0.1859,
+      "step": 9261
+    },
+    {
+      "epoch": 0.148192,
+      "grad_norm": 0.625,
+      "learning_rate": 8.586935483870969e-05,
+      "loss": 0.1796,
+      "step": 9262
+    },
+    {
+      "epoch": 0.148208,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.586774193548387e-05,
+      "loss": 0.1844,
+      "step": 9263
+    },
+    {
+      "epoch": 0.148224,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.586612903225807e-05,
+      "loss": 0.1322,
+      "step": 9264
+    },
+    {
+      "epoch": 0.14824,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.586451612903226e-05,
+      "loss": 0.1845,
+      "step": 9265
+    },
+    {
+      "epoch": 0.148256,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.586290322580646e-05,
+      "loss": 0.1547,
+      "step": 9266
+    },
+    {
+      "epoch": 0.148272,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.586129032258064e-05,
+      "loss": 0.2331,
+      "step": 9267
+    },
+    {
+      "epoch": 0.148288,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.585967741935484e-05,
+      "loss": 0.1537,
+      "step": 9268
+    },
+    {
+      "epoch": 0.148304,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.585806451612903e-05,
+      "loss": 0.1745,
+      "step": 9269
+    },
+    {
+      "epoch": 0.14832,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.585645161290323e-05,
+      "loss": 0.2005,
+      "step": 9270
+    },
+    {
+      "epoch": 0.148336,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.585483870967741e-05,
+      "loss": 0.1709,
+      "step": 9271
+    },
+    {
+      "epoch": 0.148352,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.585322580645161e-05,
+      "loss": 0.1945,
+      "step": 9272
+    },
+    {
+      "epoch": 0.148368,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.585161290322581e-05,
+      "loss": 0.1848,
+      "step": 9273
+    },
+    {
+      "epoch": 0.148384,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.585000000000001e-05,
+      "loss": 0.2025,
+      "step": 9274
+    },
+    {
+      "epoch": 0.1484,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.58483870967742e-05,
+      "loss": 0.2376,
+      "step": 9275
+    },
+    {
+      "epoch": 0.148416,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.58467741935484e-05,
+      "loss": 0.1517,
+      "step": 9276
+    },
+    {
+      "epoch": 0.148432,
+      "grad_norm": 1.2734375,
+      "learning_rate": 8.584516129032258e-05,
+      "loss": 0.1713,
+      "step": 9277
+    },
+    {
+      "epoch": 0.148448,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.584354838709678e-05,
+      "loss": 0.1982,
+      "step": 9278
+    },
+    {
+      "epoch": 0.148464,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.584193548387097e-05,
+      "loss": 0.2247,
+      "step": 9279
+    },
+    {
+      "epoch": 0.14848,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.584032258064516e-05,
+      "loss": 0.1365,
+      "step": 9280
+    },
+    {
+      "epoch": 0.148496,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.583870967741936e-05,
+      "loss": 0.2145,
+      "step": 9281
+    },
+    {
+      "epoch": 0.148512,
+      "grad_norm": 1.453125,
+      "learning_rate": 8.583709677419354e-05,
+      "loss": 0.1476,
+      "step": 9282
+    },
+    {
+      "epoch": 0.148528,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.583548387096774e-05,
+      "loss": 0.1383,
+      "step": 9283
+    },
+    {
+      "epoch": 0.148544,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.583387096774194e-05,
+      "loss": 0.2236,
+      "step": 9284
+    },
+    {
+      "epoch": 0.14856,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.583225806451614e-05,
+      "loss": 0.1618,
+      "step": 9285
+    },
+    {
+      "epoch": 0.148576,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.583064516129033e-05,
+      "loss": 0.1644,
+      "step": 9286
+    },
+    {
+      "epoch": 0.148592,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.582903225806453e-05,
+      "loss": 0.1672,
+      "step": 9287
+    },
+    {
+      "epoch": 0.148608,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.582741935483871e-05,
+      "loss": 0.1854,
+      "step": 9288
+    },
+    {
+      "epoch": 0.148624,
+      "grad_norm": 1.3203125,
+      "learning_rate": 8.582580645161291e-05,
+      "loss": 0.174,
+      "step": 9289
+    },
+    {
+      "epoch": 0.14864,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.58241935483871e-05,
+      "loss": 0.1692,
+      "step": 9290
+    },
+    {
+      "epoch": 0.148656,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.58225806451613e-05,
+      "loss": 0.1433,
+      "step": 9291
+    },
+    {
+      "epoch": 0.148672,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.582096774193548e-05,
+      "loss": 0.2088,
+      "step": 9292
+    },
+    {
+      "epoch": 0.148688,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.581935483870968e-05,
+      "loss": 0.1529,
+      "step": 9293
+    },
+    {
+      "epoch": 0.148704,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.581774193548387e-05,
+      "loss": 0.1789,
+      "step": 9294
+    },
+    {
+      "epoch": 0.14872,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.581612903225807e-05,
+      "loss": 0.1746,
+      "step": 9295
+    },
+    {
+      "epoch": 0.148736,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.581451612903227e-05,
+      "loss": 0.1584,
+      "step": 9296
+    },
+    {
+      "epoch": 0.148752,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.581290322580646e-05,
+      "loss": 0.2083,
+      "step": 9297
+    },
+    {
+      "epoch": 0.148768,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.581129032258066e-05,
+      "loss": 0.1449,
+      "step": 9298
+    },
+    {
+      "epoch": 0.148784,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.580967741935484e-05,
+      "loss": 0.2044,
+      "step": 9299
+    },
+    {
+      "epoch": 0.1488,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.580806451612904e-05,
+      "loss": 0.1985,
+      "step": 9300
+    },
+    {
+      "epoch": 0.148816,
+      "grad_norm": 1.484375,
+      "learning_rate": 8.580645161290323e-05,
+      "loss": 0.1821,
+      "step": 9301
+    },
+    {
+      "epoch": 0.148832,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.580483870967743e-05,
+      "loss": 0.1357,
+      "step": 9302
+    },
+    {
+      "epoch": 0.148848,
+      "grad_norm": 0.75,
+      "learning_rate": 8.580322580645161e-05,
+      "loss": 0.1775,
+      "step": 9303
+    },
+    {
+      "epoch": 0.148864,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.580161290322581e-05,
+      "loss": 0.1783,
+      "step": 9304
+    },
+    {
+      "epoch": 0.14888,
+      "grad_norm": 1.3671875,
+      "learning_rate": 8.58e-05,
+      "loss": 0.1603,
+      "step": 9305
+    },
+    {
+      "epoch": 0.148896,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.57983870967742e-05,
+      "loss": 0.1593,
+      "step": 9306
+    },
+    {
+      "epoch": 0.148912,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.579677419354838e-05,
+      "loss": 0.217,
+      "step": 9307
+    },
+    {
+      "epoch": 0.148928,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.579516129032258e-05,
+      "loss": 0.1615,
+      "step": 9308
+    },
+    {
+      "epoch": 0.148944,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.579354838709678e-05,
+      "loss": 0.1526,
+      "step": 9309
+    },
+    {
+      "epoch": 0.14896,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.579193548387097e-05,
+      "loss": 0.1849,
+      "step": 9310
+    },
+    {
+      "epoch": 0.148976,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.579032258064517e-05,
+      "loss": 0.1426,
+      "step": 9311
+    },
+    {
+      "epoch": 0.148992,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.578870967741936e-05,
+      "loss": 0.165,
+      "step": 9312
+    },
+    {
+      "epoch": 0.149008,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.578709677419355e-05,
+      "loss": 0.1934,
+      "step": 9313
+    },
+    {
+      "epoch": 0.149024,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.578548387096774e-05,
+      "loss": 0.172,
+      "step": 9314
+    },
+    {
+      "epoch": 0.14904,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.578387096774194e-05,
+      "loss": 0.1915,
+      "step": 9315
+    },
+    {
+      "epoch": 0.149056,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.578225806451613e-05,
+      "loss": 0.1823,
+      "step": 9316
+    },
+    {
+      "epoch": 0.149072,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.578064516129033e-05,
+      "loss": 0.2217,
+      "step": 9317
+    },
+    {
+      "epoch": 0.149088,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.577903225806451e-05,
+      "loss": 0.191,
+      "step": 9318
+    },
+    {
+      "epoch": 0.149104,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.577741935483871e-05,
+      "loss": 0.1775,
+      "step": 9319
+    },
+    {
+      "epoch": 0.14912,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.577580645161291e-05,
+      "loss": 0.1297,
+      "step": 9320
+    },
+    {
+      "epoch": 0.149136,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.577419354838711e-05,
+      "loss": 0.1536,
+      "step": 9321
+    },
+    {
+      "epoch": 0.149152,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.57725806451613e-05,
+      "loss": 0.1684,
+      "step": 9322
+    },
+    {
+      "epoch": 0.149168,
+      "grad_norm": 1.484375,
+      "learning_rate": 8.57709677419355e-05,
+      "loss": 0.1786,
+      "step": 9323
+    },
+    {
+      "epoch": 0.149184,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.576935483870968e-05,
+      "loss": 0.189,
+      "step": 9324
+    },
+    {
+      "epoch": 0.1492,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.576774193548387e-05,
+      "loss": 0.1463,
+      "step": 9325
+    },
+    {
+      "epoch": 0.149216,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.576612903225807e-05,
+      "loss": 0.186,
+      "step": 9326
+    },
+    {
+      "epoch": 0.149232,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.576451612903225e-05,
+      "loss": 0.1734,
+      "step": 9327
+    },
+    {
+      "epoch": 0.149248,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.576290322580645e-05,
+      "loss": 0.172,
+      "step": 9328
+    },
+    {
+      "epoch": 0.149264,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.576129032258064e-05,
+      "loss": 0.207,
+      "step": 9329
+    },
+    {
+      "epoch": 0.14928,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.575967741935484e-05,
+      "loss": 0.17,
+      "step": 9330
+    },
+    {
+      "epoch": 0.149296,
+      "grad_norm": 0.625,
+      "learning_rate": 8.575806451612904e-05,
+      "loss": 0.1743,
+      "step": 9331
+    },
+    {
+      "epoch": 0.149312,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.575645161290323e-05,
+      "loss": 0.1624,
+      "step": 9332
+    },
+    {
+      "epoch": 0.149328,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.575483870967743e-05,
+      "loss": 0.156,
+      "step": 9333
+    },
+    {
+      "epoch": 0.149344,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.575322580645163e-05,
+      "loss": 0.1639,
+      "step": 9334
+    },
+    {
+      "epoch": 0.14936,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.575161290322581e-05,
+      "loss": 0.1681,
+      "step": 9335
+    },
+    {
+      "epoch": 0.149376,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.575000000000001e-05,
+      "loss": 0.194,
+      "step": 9336
+    },
+    {
+      "epoch": 0.149392,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.57483870967742e-05,
+      "loss": 0.1976,
+      "step": 9337
+    },
+    {
+      "epoch": 0.149408,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.57467741935484e-05,
+      "loss": 0.186,
+      "step": 9338
+    },
+    {
+      "epoch": 0.149424,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.574516129032258e-05,
+      "loss": 0.1701,
+      "step": 9339
+    },
+    {
+      "epoch": 0.14944,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.574354838709678e-05,
+      "loss": 0.1763,
+      "step": 9340
+    },
+    {
+      "epoch": 0.149456,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.574193548387097e-05,
+      "loss": 0.1681,
+      "step": 9341
+    },
+    {
+      "epoch": 0.149472,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.574032258064515e-05,
+      "loss": 0.1704,
+      "step": 9342
+    },
+    {
+      "epoch": 0.149488,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.573870967741935e-05,
+      "loss": 0.2408,
+      "step": 9343
+    },
+    {
+      "epoch": 0.149504,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.573709677419355e-05,
+      "loss": 0.1683,
+      "step": 9344
+    },
+    {
+      "epoch": 0.14952,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.573548387096775e-05,
+      "loss": 0.1707,
+      "step": 9345
+    },
+    {
+      "epoch": 0.149536,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.573387096774194e-05,
+      "loss": 0.1517,
+      "step": 9346
+    },
+    {
+      "epoch": 0.149552,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.573225806451614e-05,
+      "loss": 0.2082,
+      "step": 9347
+    },
+    {
+      "epoch": 0.149568,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.573064516129033e-05,
+      "loss": 0.1498,
+      "step": 9348
+    },
+    {
+      "epoch": 0.149584,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.572903225806452e-05,
+      "loss": 0.196,
+      "step": 9349
+    },
+    {
+      "epoch": 0.1496,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.572741935483871e-05,
+      "loss": 0.1956,
+      "step": 9350
+    },
+    {
+      "epoch": 0.149616,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.572580645161291e-05,
+      "loss": 0.1793,
+      "step": 9351
+    },
+    {
+      "epoch": 0.149632,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.57241935483871e-05,
+      "loss": 0.1353,
+      "step": 9352
+    },
+    {
+      "epoch": 0.149648,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.57225806451613e-05,
+      "loss": 0.1653,
+      "step": 9353
+    },
+    {
+      "epoch": 0.149664,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.572096774193548e-05,
+      "loss": 0.1943,
+      "step": 9354
+    },
+    {
+      "epoch": 0.14968,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.571935483870968e-05,
+      "loss": 0.1833,
+      "step": 9355
+    },
+    {
+      "epoch": 0.149696,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.571774193548388e-05,
+      "loss": 0.1523,
+      "step": 9356
+    },
+    {
+      "epoch": 0.149712,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.571612903225807e-05,
+      "loss": 0.1827,
+      "step": 9357
+    },
+    {
+      "epoch": 0.149728,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.571451612903227e-05,
+      "loss": 0.1739,
+      "step": 9358
+    },
+    {
+      "epoch": 0.149744,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.571290322580645e-05,
+      "loss": 0.1301,
+      "step": 9359
+    },
+    {
+      "epoch": 0.14976,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.571129032258065e-05,
+      "loss": 0.158,
+      "step": 9360
+    },
+    {
+      "epoch": 0.149776,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.570967741935484e-05,
+      "loss": 0.2253,
+      "step": 9361
+    },
+    {
+      "epoch": 0.149792,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.570806451612904e-05,
+      "loss": 0.1911,
+      "step": 9362
+    },
+    {
+      "epoch": 0.149808,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.570645161290322e-05,
+      "loss": 0.1616,
+      "step": 9363
+    },
+    {
+      "epoch": 0.149824,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.570483870967742e-05,
+      "loss": 0.1744,
+      "step": 9364
+    },
+    {
+      "epoch": 0.14984,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.570322580645161e-05,
+      "loss": 0.2221,
+      "step": 9365
+    },
+    {
+      "epoch": 0.149856,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.570161290322581e-05,
+      "loss": 0.1657,
+      "step": 9366
+    },
+    {
+      "epoch": 0.149872,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.57e-05,
+      "loss": 0.1619,
+      "step": 9367
+    },
+    {
+      "epoch": 0.149888,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.56983870967742e-05,
+      "loss": 0.1575,
+      "step": 9368
+    },
+    {
+      "epoch": 0.149904,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.56967741935484e-05,
+      "loss": 0.1861,
+      "step": 9369
+    },
+    {
+      "epoch": 0.14992,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.56951612903226e-05,
+      "loss": 0.1803,
+      "step": 9370
+    },
+    {
+      "epoch": 0.149936,
+      "grad_norm": 0.75,
+      "learning_rate": 8.569354838709678e-05,
+      "loss": 0.1813,
+      "step": 9371
+    },
+    {
+      "epoch": 0.149952,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.569193548387097e-05,
+      "loss": 0.174,
+      "step": 9372
+    },
+    {
+      "epoch": 0.149968,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.569032258064517e-05,
+      "loss": 0.1911,
+      "step": 9373
+    },
+    {
+      "epoch": 0.149984,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.568870967741935e-05,
+      "loss": 0.163,
+      "step": 9374
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.568709677419355e-05,
+      "loss": 0.1633,
+      "step": 9375
+    },
+    {
+      "epoch": 0.150016,
+      "grad_norm": 1.3515625,
+      "learning_rate": 8.568548387096774e-05,
+      "loss": 0.1727,
+      "step": 9376
+    },
+    {
+      "epoch": 0.150032,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.568387096774194e-05,
+      "loss": 0.1606,
+      "step": 9377
+    },
+    {
+      "epoch": 0.150048,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.568225806451612e-05,
+      "loss": 0.1532,
+      "step": 9378
+    },
+    {
+      "epoch": 0.150064,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.568064516129032e-05,
+      "loss": 0.2024,
+      "step": 9379
+    },
+    {
+      "epoch": 0.15008,
+      "grad_norm": 0.875,
+      "learning_rate": 8.567903225806452e-05,
+      "loss": 0.1761,
+      "step": 9380
+    },
+    {
+      "epoch": 0.150096,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.567741935483872e-05,
+      "loss": 0.1766,
+      "step": 9381
+    },
+    {
+      "epoch": 0.150112,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.567580645161291e-05,
+      "loss": 0.173,
+      "step": 9382
+    },
+    {
+      "epoch": 0.150128,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.567419354838711e-05,
+      "loss": 0.174,
+      "step": 9383
+    },
+    {
+      "epoch": 0.150144,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.56725806451613e-05,
+      "loss": 0.1764,
+      "step": 9384
+    },
+    {
+      "epoch": 0.15016,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.56709677419355e-05,
+      "loss": 0.1629,
+      "step": 9385
+    },
+    {
+      "epoch": 0.150176,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.566935483870968e-05,
+      "loss": 0.2408,
+      "step": 9386
+    },
+    {
+      "epoch": 0.150192,
+      "grad_norm": 1.2734375,
+      "learning_rate": 8.566774193548388e-05,
+      "loss": 0.1644,
+      "step": 9387
+    },
+    {
+      "epoch": 0.150208,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.566612903225807e-05,
+      "loss": 0.1444,
+      "step": 9388
+    },
+    {
+      "epoch": 0.150224,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.566451612903225e-05,
+      "loss": 0.2106,
+      "step": 9389
+    },
+    {
+      "epoch": 0.15024,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.566290322580645e-05,
+      "loss": 0.1757,
+      "step": 9390
+    },
+    {
+      "epoch": 0.150256,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.566129032258065e-05,
+      "loss": 0.1385,
+      "step": 9391
+    },
+    {
+      "epoch": 0.150272,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.565967741935485e-05,
+      "loss": 0.1712,
+      "step": 9392
+    },
+    {
+      "epoch": 0.150288,
+      "grad_norm": 1.234375,
+      "learning_rate": 8.565806451612904e-05,
+      "loss": 0.2336,
+      "step": 9393
+    },
+    {
+      "epoch": 0.150304,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.565645161290324e-05,
+      "loss": 0.1727,
+      "step": 9394
+    },
+    {
+      "epoch": 0.15032,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.565483870967742e-05,
+      "loss": 0.1614,
+      "step": 9395
+    },
+    {
+      "epoch": 0.150336,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.565322580645162e-05,
+      "loss": 0.1353,
+      "step": 9396
+    },
+    {
+      "epoch": 0.150352,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.565161290322581e-05,
+      "loss": 0.1606,
+      "step": 9397
+    },
+    {
+      "epoch": 0.150368,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.565000000000001e-05,
+      "loss": 0.1371,
+      "step": 9398
+    },
+    {
+      "epoch": 0.150384,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.56483870967742e-05,
+      "loss": 0.1696,
+      "step": 9399
+    },
+    {
+      "epoch": 0.1504,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.56467741935484e-05,
+      "loss": 0.1822,
+      "step": 9400
+    },
+    {
+      "epoch": 0.150416,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.564516129032258e-05,
+      "loss": 0.1714,
+      "step": 9401
+    },
+    {
+      "epoch": 0.150432,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.564354838709678e-05,
+      "loss": 0.1716,
+      "step": 9402
+    },
+    {
+      "epoch": 0.150448,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.564193548387097e-05,
+      "loss": 0.1891,
+      "step": 9403
+    },
+    {
+      "epoch": 0.150464,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.564032258064517e-05,
+      "loss": 0.2035,
+      "step": 9404
+    },
+    {
+      "epoch": 0.15048,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.563870967741937e-05,
+      "loss": 0.1404,
+      "step": 9405
+    },
+    {
+      "epoch": 0.150496,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.563709677419355e-05,
+      "loss": 0.1815,
+      "step": 9406
+    },
+    {
+      "epoch": 0.150512,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.563548387096775e-05,
+      "loss": 0.184,
+      "step": 9407
+    },
+    {
+      "epoch": 0.150528,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.563387096774194e-05,
+      "loss": 0.216,
+      "step": 9408
+    },
+    {
+      "epoch": 0.150544,
+      "grad_norm": 1.390625,
+      "learning_rate": 8.563225806451614e-05,
+      "loss": 0.1945,
+      "step": 9409
+    },
+    {
+      "epoch": 0.15056,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.563064516129032e-05,
+      "loss": 0.1806,
+      "step": 9410
+    },
+    {
+      "epoch": 0.150576,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.562903225806452e-05,
+      "loss": 0.1391,
+      "step": 9411
+    },
+    {
+      "epoch": 0.150592,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.562741935483871e-05,
+      "loss": 0.1709,
+      "step": 9412
+    },
+    {
+      "epoch": 0.150608,
+      "grad_norm": 1.0,
+      "learning_rate": 8.562580645161291e-05,
+      "loss": 0.2112,
+      "step": 9413
+    },
+    {
+      "epoch": 0.150624,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.56241935483871e-05,
+      "loss": 0.2165,
+      "step": 9414
+    },
+    {
+      "epoch": 0.15064,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.56225806451613e-05,
+      "loss": 0.2233,
+      "step": 9415
+    },
+    {
+      "epoch": 0.150656,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.56209677419355e-05,
+      "loss": 0.1594,
+      "step": 9416
+    },
+    {
+      "epoch": 0.150672,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.561935483870969e-05,
+      "loss": 0.1732,
+      "step": 9417
+    },
+    {
+      "epoch": 0.150688,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.561774193548388e-05,
+      "loss": 0.1991,
+      "step": 9418
+    },
+    {
+      "epoch": 0.150704,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.561612903225807e-05,
+      "loss": 0.1762,
+      "step": 9419
+    },
+    {
+      "epoch": 0.15072,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.561451612903226e-05,
+      "loss": 0.1748,
+      "step": 9420
+    },
+    {
+      "epoch": 0.150736,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.561290322580645e-05,
+      "loss": 0.1278,
+      "step": 9421
+    },
+    {
+      "epoch": 0.150752,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.561129032258065e-05,
+      "loss": 0.1779,
+      "step": 9422
+    },
+    {
+      "epoch": 0.150768,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.560967741935484e-05,
+      "loss": 0.1817,
+      "step": 9423
+    },
+    {
+      "epoch": 0.150784,
+      "grad_norm": 1.4140625,
+      "learning_rate": 8.560806451612904e-05,
+      "loss": 0.2275,
+      "step": 9424
+    },
+    {
+      "epoch": 0.1508,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.560645161290322e-05,
+      "loss": 0.1815,
+      "step": 9425
+    },
+    {
+      "epoch": 0.150816,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.560483870967742e-05,
+      "loss": 0.1904,
+      "step": 9426
+    },
+    {
+      "epoch": 0.150832,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.560322580645161e-05,
+      "loss": 0.173,
+      "step": 9427
+    },
+    {
+      "epoch": 0.150848,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.560161290322581e-05,
+      "loss": 0.2064,
+      "step": 9428
+    },
+    {
+      "epoch": 0.150864,
+      "grad_norm": 1.4140625,
+      "learning_rate": 8.560000000000001e-05,
+      "loss": 0.1987,
+      "step": 9429
+    },
+    {
+      "epoch": 0.15088,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.55983870967742e-05,
+      "loss": 0.1796,
+      "step": 9430
+    },
+    {
+      "epoch": 0.150896,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.559677419354839e-05,
+      "loss": 0.1833,
+      "step": 9431
+    },
+    {
+      "epoch": 0.150912,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.559516129032259e-05,
+      "loss": 0.1677,
+      "step": 9432
+    },
+    {
+      "epoch": 0.150928,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.559354838709678e-05,
+      "loss": 0.2181,
+      "step": 9433
+    },
+    {
+      "epoch": 0.150944,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.559193548387096e-05,
+      "loss": 0.1937,
+      "step": 9434
+    },
+    {
+      "epoch": 0.15096,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.559032258064516e-05,
+      "loss": 0.1727,
+      "step": 9435
+    },
+    {
+      "epoch": 0.150976,
+      "grad_norm": 1.125,
+      "learning_rate": 8.558870967741935e-05,
+      "loss": 0.2172,
+      "step": 9436
+    },
+    {
+      "epoch": 0.150992,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.558709677419355e-05,
+      "loss": 0.1818,
+      "step": 9437
+    },
+    {
+      "epoch": 0.151008,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.558548387096774e-05,
+      "loss": 0.1732,
+      "step": 9438
+    },
+    {
+      "epoch": 0.151024,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.558387096774194e-05,
+      "loss": 0.1891,
+      "step": 9439
+    },
+    {
+      "epoch": 0.15104,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.558225806451614e-05,
+      "loss": 0.1982,
+      "step": 9440
+    },
+    {
+      "epoch": 0.151056,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.558064516129033e-05,
+      "loss": 0.1835,
+      "step": 9441
+    },
+    {
+      "epoch": 0.151072,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.557903225806452e-05,
+      "loss": 0.1646,
+      "step": 9442
+    },
+    {
+      "epoch": 0.151088,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.557741935483872e-05,
+      "loss": 0.1628,
+      "step": 9443
+    },
+    {
+      "epoch": 0.151104,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.55758064516129e-05,
+      "loss": 0.1696,
+      "step": 9444
+    },
+    {
+      "epoch": 0.15112,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.55741935483871e-05,
+      "loss": 0.1302,
+      "step": 9445
+    },
+    {
+      "epoch": 0.151136,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.557258064516129e-05,
+      "loss": 0.1411,
+      "step": 9446
+    },
+    {
+      "epoch": 0.151152,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.557096774193549e-05,
+      "loss": 0.1632,
+      "step": 9447
+    },
+    {
+      "epoch": 0.151168,
+      "grad_norm": 0.75,
+      "learning_rate": 8.556935483870968e-05,
+      "loss": 0.1802,
+      "step": 9448
+    },
+    {
+      "epoch": 0.151184,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.556774193548388e-05,
+      "loss": 0.1865,
+      "step": 9449
+    },
+    {
+      "epoch": 0.1512,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.556612903225806e-05,
+      "loss": 0.1346,
+      "step": 9450
+    },
+    {
+      "epoch": 0.151216,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.556451612903226e-05,
+      "loss": 0.1715,
+      "step": 9451
+    },
+    {
+      "epoch": 0.151232,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.556290322580646e-05,
+      "loss": 0.1757,
+      "step": 9452
+    },
+    {
+      "epoch": 0.151248,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.556129032258065e-05,
+      "loss": 0.186,
+      "step": 9453
+    },
+    {
+      "epoch": 0.151264,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.555967741935485e-05,
+      "loss": 0.1396,
+      "step": 9454
+    },
+    {
+      "epoch": 0.15128,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.555806451612903e-05,
+      "loss": 0.2025,
+      "step": 9455
+    },
+    {
+      "epoch": 0.151296,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.555645161290323e-05,
+      "loss": 0.1561,
+      "step": 9456
+    },
+    {
+      "epoch": 0.151312,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.555483870967742e-05,
+      "loss": 0.1697,
+      "step": 9457
+    },
+    {
+      "epoch": 0.151328,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.555322580645162e-05,
+      "loss": 0.1809,
+      "step": 9458
+    },
+    {
+      "epoch": 0.151344,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.55516129032258e-05,
+      "loss": 0.1514,
+      "step": 9459
+    },
+    {
+      "epoch": 0.15136,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.555e-05,
+      "loss": 0.1665,
+      "step": 9460
+    },
+    {
+      "epoch": 0.151376,
+      "grad_norm": 1.3125,
+      "learning_rate": 8.554838709677419e-05,
+      "loss": 0.2005,
+      "step": 9461
+    },
+    {
+      "epoch": 0.151392,
+      "grad_norm": 1.125,
+      "learning_rate": 8.554677419354839e-05,
+      "loss": 0.1957,
+      "step": 9462
+    },
+    {
+      "epoch": 0.151408,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.554516129032258e-05,
+      "loss": 0.1795,
+      "step": 9463
+    },
+    {
+      "epoch": 0.151424,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.554354838709678e-05,
+      "loss": 0.1843,
+      "step": 9464
+    },
+    {
+      "epoch": 0.15144,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.554193548387098e-05,
+      "loss": 0.2006,
+      "step": 9465
+    },
+    {
+      "epoch": 0.151456,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.554032258064516e-05,
+      "loss": 0.225,
+      "step": 9466
+    },
+    {
+      "epoch": 0.151472,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.553870967741936e-05,
+      "loss": 0.2182,
+      "step": 9467
+    },
+    {
+      "epoch": 0.151488,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.553709677419355e-05,
+      "loss": 0.1492,
+      "step": 9468
+    },
+    {
+      "epoch": 0.151504,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.553548387096775e-05,
+      "loss": 0.1472,
+      "step": 9469
+    },
+    {
+      "epoch": 0.15152,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.553387096774193e-05,
+      "loss": 0.1565,
+      "step": 9470
+    },
+    {
+      "epoch": 0.151536,
+      "grad_norm": 0.490234375,
+      "learning_rate": 8.553225806451613e-05,
+      "loss": 0.134,
+      "step": 9471
+    },
+    {
+      "epoch": 0.151552,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.553064516129032e-05,
+      "loss": 0.1893,
+      "step": 9472
+    },
+    {
+      "epoch": 0.151568,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.552903225806452e-05,
+      "loss": 0.1809,
+      "step": 9473
+    },
+    {
+      "epoch": 0.151584,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.55274193548387e-05,
+      "loss": 0.2087,
+      "step": 9474
+    },
+    {
+      "epoch": 0.1516,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.55258064516129e-05,
+      "loss": 0.171,
+      "step": 9475
+    },
+    {
+      "epoch": 0.151616,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.55241935483871e-05,
+      "loss": 0.1637,
+      "step": 9476
+    },
+    {
+      "epoch": 0.151632,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.55225806451613e-05,
+      "loss": 0.1465,
+      "step": 9477
+    },
+    {
+      "epoch": 0.151648,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.552096774193549e-05,
+      "loss": 0.2088,
+      "step": 9478
+    },
+    {
+      "epoch": 0.151664,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.551935483870969e-05,
+      "loss": 0.1697,
+      "step": 9479
+    },
+    {
+      "epoch": 0.15168,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.551774193548388e-05,
+      "loss": 0.1842,
+      "step": 9480
+    },
+    {
+      "epoch": 0.151696,
+      "grad_norm": 0.4453125,
+      "learning_rate": 8.551612903225806e-05,
+      "loss": 0.108,
+      "step": 9481
+    },
+    {
+      "epoch": 0.151712,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.551451612903226e-05,
+      "loss": 0.1502,
+      "step": 9482
+    },
+    {
+      "epoch": 0.151728,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.551290322580645e-05,
+      "loss": 0.1577,
+      "step": 9483
+    },
+    {
+      "epoch": 0.151744,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.551129032258065e-05,
+      "loss": 0.1507,
+      "step": 9484
+    },
+    {
+      "epoch": 0.15176,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.550967741935483e-05,
+      "loss": 0.1572,
+      "step": 9485
+    },
+    {
+      "epoch": 0.151776,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.550806451612903e-05,
+      "loss": 0.1637,
+      "step": 9486
+    },
+    {
+      "epoch": 0.151792,
+      "grad_norm": 0.49609375,
+      "learning_rate": 8.550645161290323e-05,
+      "loss": 0.1634,
+      "step": 9487
+    },
+    {
+      "epoch": 0.151808,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.550483870967743e-05,
+      "loss": 0.1729,
+      "step": 9488
+    },
+    {
+      "epoch": 0.151824,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.550322580645162e-05,
+      "loss": 0.1672,
+      "step": 9489
+    },
+    {
+      "epoch": 0.15184,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.550161290322582e-05,
+      "loss": 0.1696,
+      "step": 9490
+    },
+    {
+      "epoch": 0.151856,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.55e-05,
+      "loss": 0.1748,
+      "step": 9491
+    },
+    {
+      "epoch": 0.151872,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.54983870967742e-05,
+      "loss": 0.1717,
+      "step": 9492
+    },
+    {
+      "epoch": 0.151888,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.549677419354839e-05,
+      "loss": 0.1851,
+      "step": 9493
+    },
+    {
+      "epoch": 0.151904,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.549516129032259e-05,
+      "loss": 0.1774,
+      "step": 9494
+    },
+    {
+      "epoch": 0.15192,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.549354838709678e-05,
+      "loss": 0.1457,
+      "step": 9495
+    },
+    {
+      "epoch": 0.151936,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.549193548387096e-05,
+      "loss": 0.162,
+      "step": 9496
+    },
+    {
+      "epoch": 0.151952,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.549032258064516e-05,
+      "loss": 0.2258,
+      "step": 9497
+    },
+    {
+      "epoch": 0.151968,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.548870967741935e-05,
+      "loss": 0.1453,
+      "step": 9498
+    },
+    {
+      "epoch": 0.151984,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.548709677419355e-05,
+      "loss": 0.1585,
+      "step": 9499
+    },
+    {
+      "epoch": 0.152,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.548548387096775e-05,
+      "loss": 0.1796,
+      "step": 9500
+    },
+    {
+      "epoch": 0.152016,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.548387096774195e-05,
+      "loss": 0.1722,
+      "step": 9501
+    },
+    {
+      "epoch": 0.152032,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.548225806451613e-05,
+      "loss": 0.2111,
+      "step": 9502
+    },
+    {
+      "epoch": 0.152048,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.548064516129033e-05,
+      "loss": 0.1554,
+      "step": 9503
+    },
+    {
+      "epoch": 0.152064,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.547903225806452e-05,
+      "loss": 0.1795,
+      "step": 9504
+    },
+    {
+      "epoch": 0.15208,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.547741935483872e-05,
+      "loss": 0.1764,
+      "step": 9505
+    },
+    {
+      "epoch": 0.152096,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.54758064516129e-05,
+      "loss": 0.1715,
+      "step": 9506
+    },
+    {
+      "epoch": 0.152112,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.54741935483871e-05,
+      "loss": 0.1692,
+      "step": 9507
+    },
+    {
+      "epoch": 0.152128,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.547258064516129e-05,
+      "loss": 0.1264,
+      "step": 9508
+    },
+    {
+      "epoch": 0.152144,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.547096774193549e-05,
+      "loss": 0.217,
+      "step": 9509
+    },
+    {
+      "epoch": 0.15216,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.546935483870968e-05,
+      "loss": 0.1596,
+      "step": 9510
+    },
+    {
+      "epoch": 0.152176,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.546774193548388e-05,
+      "loss": 0.1884,
+      "step": 9511
+    },
+    {
+      "epoch": 0.152192,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.546612903225807e-05,
+      "loss": 0.1585,
+      "step": 9512
+    },
+    {
+      "epoch": 0.152208,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.546451612903226e-05,
+      "loss": 0.1908,
+      "step": 9513
+    },
+    {
+      "epoch": 0.152224,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.546290322580646e-05,
+      "loss": 0.2143,
+      "step": 9514
+    },
+    {
+      "epoch": 0.15224,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.546129032258065e-05,
+      "loss": 0.1535,
+      "step": 9515
+    },
+    {
+      "epoch": 0.152256,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.545967741935485e-05,
+      "loss": 0.1883,
+      "step": 9516
+    },
+    {
+      "epoch": 0.152272,
+      "grad_norm": 0.75,
+      "learning_rate": 8.545806451612903e-05,
+      "loss": 0.1787,
+      "step": 9517
+    },
+    {
+      "epoch": 0.152288,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.545645161290323e-05,
+      "loss": 0.1479,
+      "step": 9518
+    },
+    {
+      "epoch": 0.152304,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.545483870967742e-05,
+      "loss": 0.1699,
+      "step": 9519
+    },
+    {
+      "epoch": 0.15232,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.545322580645162e-05,
+      "loss": 0.1988,
+      "step": 9520
+    },
+    {
+      "epoch": 0.152336,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.54516129032258e-05,
+      "loss": 0.1872,
+      "step": 9521
+    },
+    {
+      "epoch": 0.152352,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.545e-05,
+      "loss": 0.1729,
+      "step": 9522
+    },
+    {
+      "epoch": 0.152368,
+      "grad_norm": 1.3515625,
+      "learning_rate": 8.544838709677419e-05,
+      "loss": 0.1864,
+      "step": 9523
+    },
+    {
+      "epoch": 0.152384,
+      "grad_norm": 0.75,
+      "learning_rate": 8.544677419354839e-05,
+      "loss": 0.1382,
+      "step": 9524
+    },
+    {
+      "epoch": 0.1524,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.544516129032259e-05,
+      "loss": 0.147,
+      "step": 9525
+    },
+    {
+      "epoch": 0.152416,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.544354838709679e-05,
+      "loss": 0.2038,
+      "step": 9526
+    },
+    {
+      "epoch": 0.152432,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.544193548387097e-05,
+      "loss": 0.1375,
+      "step": 9527
+    },
+    {
+      "epoch": 0.152448,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.544032258064516e-05,
+      "loss": 0.1635,
+      "step": 9528
+    },
+    {
+      "epoch": 0.152464,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.543870967741936e-05,
+      "loss": 0.1805,
+      "step": 9529
+    },
+    {
+      "epoch": 0.15248,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.543709677419355e-05,
+      "loss": 0.208,
+      "step": 9530
+    },
+    {
+      "epoch": 0.152496,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.543548387096775e-05,
+      "loss": 0.1738,
+      "step": 9531
+    },
+    {
+      "epoch": 0.152512,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.543387096774193e-05,
+      "loss": 0.1956,
+      "step": 9532
+    },
+    {
+      "epoch": 0.152528,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.543225806451613e-05,
+      "loss": 0.1413,
+      "step": 9533
+    },
+    {
+      "epoch": 0.152544,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.543064516129032e-05,
+      "loss": 0.1717,
+      "step": 9534
+    },
+    {
+      "epoch": 0.15256,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.542903225806452e-05,
+      "loss": 0.2188,
+      "step": 9535
+    },
+    {
+      "epoch": 0.152576,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.542741935483872e-05,
+      "loss": 0.1855,
+      "step": 9536
+    },
+    {
+      "epoch": 0.152592,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.542580645161292e-05,
+      "loss": 0.2004,
+      "step": 9537
+    },
+    {
+      "epoch": 0.152608,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.54241935483871e-05,
+      "loss": 0.1265,
+      "step": 9538
+    },
+    {
+      "epoch": 0.152624,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.54225806451613e-05,
+      "loss": 0.1546,
+      "step": 9539
+    },
+    {
+      "epoch": 0.15264,
+      "grad_norm": 0.75,
+      "learning_rate": 8.542096774193549e-05,
+      "loss": 0.1522,
+      "step": 9540
+    },
+    {
+      "epoch": 0.152656,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.541935483870969e-05,
+      "loss": 0.1705,
+      "step": 9541
+    },
+    {
+      "epoch": 0.152672,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.541774193548387e-05,
+      "loss": 0.1684,
+      "step": 9542
+    },
+    {
+      "epoch": 0.152688,
+      "grad_norm": 0.625,
+      "learning_rate": 8.541612903225806e-05,
+      "loss": 0.1836,
+      "step": 9543
+    },
+    {
+      "epoch": 0.152704,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.541451612903226e-05,
+      "loss": 0.1786,
+      "step": 9544
+    },
+    {
+      "epoch": 0.15272,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.541290322580645e-05,
+      "loss": 0.1879,
+      "step": 9545
+    },
+    {
+      "epoch": 0.152736,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.541129032258065e-05,
+      "loss": 0.1595,
+      "step": 9546
+    },
+    {
+      "epoch": 0.152752,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.540967741935485e-05,
+      "loss": 0.1707,
+      "step": 9547
+    },
+    {
+      "epoch": 0.152768,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.540806451612904e-05,
+      "loss": 0.1702,
+      "step": 9548
+    },
+    {
+      "epoch": 0.152784,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.540645161290323e-05,
+      "loss": 0.1666,
+      "step": 9549
+    },
+    {
+      "epoch": 0.1528,
+      "grad_norm": 1.125,
+      "learning_rate": 8.540483870967743e-05,
+      "loss": 0.1656,
+      "step": 9550
+    },
+    {
+      "epoch": 0.152816,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.540322580645162e-05,
+      "loss": 0.1395,
+      "step": 9551
+    },
+    {
+      "epoch": 0.152832,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.540161290322582e-05,
+      "loss": 0.197,
+      "step": 9552
+    },
+    {
+      "epoch": 0.152848,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.54e-05,
+      "loss": 0.201,
+      "step": 9553
+    },
+    {
+      "epoch": 0.152864,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.53983870967742e-05,
+      "loss": 0.157,
+      "step": 9554
+    },
+    {
+      "epoch": 0.15288,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.539677419354839e-05,
+      "loss": 0.1738,
+      "step": 9555
+    },
+    {
+      "epoch": 0.152896,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.539516129032259e-05,
+      "loss": 0.1627,
+      "step": 9556
+    },
+    {
+      "epoch": 0.152912,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.539354838709677e-05,
+      "loss": 0.1687,
+      "step": 9557
+    },
+    {
+      "epoch": 0.152928,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.539193548387097e-05,
+      "loss": 0.1912,
+      "step": 9558
+    },
+    {
+      "epoch": 0.152944,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.539032258064516e-05,
+      "loss": 0.1859,
+      "step": 9559
+    },
+    {
+      "epoch": 0.15296,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.538870967741936e-05,
+      "loss": 0.1343,
+      "step": 9560
+    },
+    {
+      "epoch": 0.152976,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.538709677419356e-05,
+      "loss": 0.1818,
+      "step": 9561
+    },
+    {
+      "epoch": 0.152992,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.538548387096774e-05,
+      "loss": 0.1439,
+      "step": 9562
+    },
+    {
+      "epoch": 0.153008,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.538387096774194e-05,
+      "loss": 0.1671,
+      "step": 9563
+    },
+    {
+      "epoch": 0.153024,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.538225806451613e-05,
+      "loss": 0.1818,
+      "step": 9564
+    },
+    {
+      "epoch": 0.15304,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.538064516129033e-05,
+      "loss": 0.1828,
+      "step": 9565
+    },
+    {
+      "epoch": 0.153056,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.537903225806452e-05,
+      "loss": 0.167,
+      "step": 9566
+    },
+    {
+      "epoch": 0.153072,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.537741935483872e-05,
+      "loss": 0.1955,
+      "step": 9567
+    },
+    {
+      "epoch": 0.153088,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.53758064516129e-05,
+      "loss": 0.1926,
+      "step": 9568
+    },
+    {
+      "epoch": 0.153104,
+      "grad_norm": 1.0,
+      "learning_rate": 8.53741935483871e-05,
+      "loss": 0.1721,
+      "step": 9569
+    },
+    {
+      "epoch": 0.15312,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.537258064516129e-05,
+      "loss": 0.1622,
+      "step": 9570
+    },
+    {
+      "epoch": 0.153136,
+      "grad_norm": 1.4921875,
+      "learning_rate": 8.537096774193549e-05,
+      "loss": 0.1536,
+      "step": 9571
+    },
+    {
+      "epoch": 0.153152,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.536935483870969e-05,
+      "loss": 0.2177,
+      "step": 9572
+    },
+    {
+      "epoch": 0.153168,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.536774193548389e-05,
+      "loss": 0.1532,
+      "step": 9573
+    },
+    {
+      "epoch": 0.153184,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.536612903225807e-05,
+      "loss": 0.1818,
+      "step": 9574
+    },
+    {
+      "epoch": 0.1532,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.536451612903226e-05,
+      "loss": 0.1745,
+      "step": 9575
+    },
+    {
+      "epoch": 0.153216,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.536290322580646e-05,
+      "loss": 0.1936,
+      "step": 9576
+    },
+    {
+      "epoch": 0.153232,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.536129032258064e-05,
+      "loss": 0.1915,
+      "step": 9577
+    },
+    {
+      "epoch": 0.153248,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.535967741935484e-05,
+      "loss": 0.141,
+      "step": 9578
+    },
+    {
+      "epoch": 0.153264,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.535806451612903e-05,
+      "loss": 0.1674,
+      "step": 9579
+    },
+    {
+      "epoch": 0.15328,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.535645161290323e-05,
+      "loss": 0.1739,
+      "step": 9580
+    },
+    {
+      "epoch": 0.153296,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.535483870967742e-05,
+      "loss": 0.2014,
+      "step": 9581
+    },
+    {
+      "epoch": 0.153312,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.535322580645162e-05,
+      "loss": 0.2053,
+      "step": 9582
+    },
+    {
+      "epoch": 0.153328,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.535161290322581e-05,
+      "loss": 0.1737,
+      "step": 9583
+    },
+    {
+      "epoch": 0.153344,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.535e-05,
+      "loss": 0.175,
+      "step": 9584
+    },
+    {
+      "epoch": 0.15336,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.53483870967742e-05,
+      "loss": 0.1594,
+      "step": 9585
+    },
+    {
+      "epoch": 0.153376,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.53467741935484e-05,
+      "loss": 0.1609,
+      "step": 9586
+    },
+    {
+      "epoch": 0.153392,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.534516129032259e-05,
+      "loss": 0.1274,
+      "step": 9587
+    },
+    {
+      "epoch": 0.153408,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.534354838709679e-05,
+      "loss": 0.2117,
+      "step": 9588
+    },
+    {
+      "epoch": 0.153424,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.534193548387097e-05,
+      "loss": 0.1871,
+      "step": 9589
+    },
+    {
+      "epoch": 0.15344,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.534032258064516e-05,
+      "loss": 0.1701,
+      "step": 9590
+    },
+    {
+      "epoch": 0.153456,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.533870967741936e-05,
+      "loss": 0.172,
+      "step": 9591
+    },
+    {
+      "epoch": 0.153472,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.533709677419354e-05,
+      "loss": 0.1453,
+      "step": 9592
+    },
+    {
+      "epoch": 0.153488,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.533548387096774e-05,
+      "loss": 0.1272,
+      "step": 9593
+    },
+    {
+      "epoch": 0.153504,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.533387096774193e-05,
+      "loss": 0.1389,
+      "step": 9594
+    },
+    {
+      "epoch": 0.15352,
+      "grad_norm": 0.75,
+      "learning_rate": 8.533225806451613e-05,
+      "loss": 0.2078,
+      "step": 9595
+    },
+    {
+      "epoch": 0.153536,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.533064516129033e-05,
+      "loss": 0.1552,
+      "step": 9596
+    },
+    {
+      "epoch": 0.153552,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.532903225806453e-05,
+      "loss": 0.28,
+      "step": 9597
+    },
+    {
+      "epoch": 0.153568,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.532741935483871e-05,
+      "loss": 0.1797,
+      "step": 9598
+    },
+    {
+      "epoch": 0.153584,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.532580645161291e-05,
+      "loss": 0.1698,
+      "step": 9599
+    },
+    {
+      "epoch": 0.1536,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.53241935483871e-05,
+      "loss": 0.1498,
+      "step": 9600
+    },
+    {
+      "epoch": 0.153616,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.53225806451613e-05,
+      "loss": 0.1979,
+      "step": 9601
+    },
+    {
+      "epoch": 0.153632,
+      "grad_norm": 0.75,
+      "learning_rate": 8.532096774193549e-05,
+      "loss": 0.1652,
+      "step": 9602
+    },
+    {
+      "epoch": 0.153648,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.531935483870969e-05,
+      "loss": 0.1194,
+      "step": 9603
+    },
+    {
+      "epoch": 0.153664,
+      "grad_norm": 1.390625,
+      "learning_rate": 8.531774193548387e-05,
+      "loss": 0.2174,
+      "step": 9604
+    },
+    {
+      "epoch": 0.15368,
+      "grad_norm": 1.2578125,
+      "learning_rate": 8.531612903225806e-05,
+      "loss": 0.177,
+      "step": 9605
+    },
+    {
+      "epoch": 0.153696,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.531451612903226e-05,
+      "loss": 0.1633,
+      "step": 9606
+    },
+    {
+      "epoch": 0.153712,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.531290322580646e-05,
+      "loss": 0.1691,
+      "step": 9607
+    },
+    {
+      "epoch": 0.153728,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.531129032258066e-05,
+      "loss": 0.1749,
+      "step": 9608
+    },
+    {
+      "epoch": 0.153744,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.530967741935484e-05,
+      "loss": 0.1683,
+      "step": 9609
+    },
+    {
+      "epoch": 0.15376,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.530806451612904e-05,
+      "loss": 0.211,
+      "step": 9610
+    },
+    {
+      "epoch": 0.153776,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.530645161290323e-05,
+      "loss": 0.2147,
+      "step": 9611
+    },
+    {
+      "epoch": 0.153792,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.530483870967743e-05,
+      "loss": 0.2048,
+      "step": 9612
+    },
+    {
+      "epoch": 0.153808,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.530322580645161e-05,
+      "loss": 0.1729,
+      "step": 9613
+    },
+    {
+      "epoch": 0.153824,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.530161290322581e-05,
+      "loss": 0.1689,
+      "step": 9614
+    },
+    {
+      "epoch": 0.15384,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.53e-05,
+      "loss": 0.214,
+      "step": 9615
+    },
+    {
+      "epoch": 0.153856,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.52983870967742e-05,
+      "loss": 0.1716,
+      "step": 9616
+    },
+    {
+      "epoch": 0.153872,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.529677419354839e-05,
+      "loss": 0.1927,
+      "step": 9617
+    },
+    {
+      "epoch": 0.153888,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.529516129032259e-05,
+      "loss": 0.2084,
+      "step": 9618
+    },
+    {
+      "epoch": 0.153904,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.529354838709677e-05,
+      "loss": 0.1603,
+      "step": 9619
+    },
+    {
+      "epoch": 0.15392,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.529193548387097e-05,
+      "loss": 0.2044,
+      "step": 9620
+    },
+    {
+      "epoch": 0.153936,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.529032258064517e-05,
+      "loss": 0.1577,
+      "step": 9621
+    },
+    {
+      "epoch": 0.153952,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.528870967741936e-05,
+      "loss": 0.1512,
+      "step": 9622
+    },
+    {
+      "epoch": 0.153968,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.528709677419356e-05,
+      "loss": 0.1583,
+      "step": 9623
+    },
+    {
+      "epoch": 0.153984,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.528548387096774e-05,
+      "loss": 0.1671,
+      "step": 9624
+    },
+    {
+      "epoch": 0.154,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.528387096774194e-05,
+      "loss": 0.1754,
+      "step": 9625
+    },
+    {
+      "epoch": 0.154016,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.528225806451613e-05,
+      "loss": 0.2,
+      "step": 9626
+    },
+    {
+      "epoch": 0.154032,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.528064516129033e-05,
+      "loss": 0.1531,
+      "step": 9627
+    },
+    {
+      "epoch": 0.154048,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.527903225806451e-05,
+      "loss": 0.158,
+      "step": 9628
+    },
+    {
+      "epoch": 0.154064,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.527741935483871e-05,
+      "loss": 0.1856,
+      "step": 9629
+    },
+    {
+      "epoch": 0.15408,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.52758064516129e-05,
+      "loss": 0.1572,
+      "step": 9630
+    },
+    {
+      "epoch": 0.154096,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.52741935483871e-05,
+      "loss": 0.1636,
+      "step": 9631
+    },
+    {
+      "epoch": 0.154112,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.52725806451613e-05,
+      "loss": 0.2023,
+      "step": 9632
+    },
+    {
+      "epoch": 0.154128,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.52709677419355e-05,
+      "loss": 0.1904,
+      "step": 9633
+    },
+    {
+      "epoch": 0.154144,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.526935483870968e-05,
+      "loss": 0.1544,
+      "step": 9634
+    },
+    {
+      "epoch": 0.15416,
+      "grad_norm": 0.466796875,
+      "learning_rate": 8.526774193548388e-05,
+      "loss": 0.1366,
+      "step": 9635
+    },
+    {
+      "epoch": 0.154176,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.526612903225807e-05,
+      "loss": 0.1692,
+      "step": 9636
+    },
+    {
+      "epoch": 0.154192,
+      "grad_norm": 0.423828125,
+      "learning_rate": 8.526451612903226e-05,
+      "loss": 0.135,
+      "step": 9637
+    },
+    {
+      "epoch": 0.154208,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.526290322580646e-05,
+      "loss": 0.1717,
+      "step": 9638
+    },
+    {
+      "epoch": 0.154224,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.526129032258064e-05,
+      "loss": 0.1564,
+      "step": 9639
+    },
+    {
+      "epoch": 0.15424,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.525967741935484e-05,
+      "loss": 0.1665,
+      "step": 9640
+    },
+    {
+      "epoch": 0.154256,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.525806451612903e-05,
+      "loss": 0.1545,
+      "step": 9641
+    },
+    {
+      "epoch": 0.154272,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.525645161290323e-05,
+      "loss": 0.2267,
+      "step": 9642
+    },
+    {
+      "epoch": 0.154288,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.525483870967743e-05,
+      "loss": 0.176,
+      "step": 9643
+    },
+    {
+      "epoch": 0.154304,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.525322580645163e-05,
+      "loss": 0.1932,
+      "step": 9644
+    },
+    {
+      "epoch": 0.15432,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.525161290322581e-05,
+      "loss": 0.1851,
+      "step": 9645
+    },
+    {
+      "epoch": 0.154336,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.525000000000001e-05,
+      "loss": 0.186,
+      "step": 9646
+    },
+    {
+      "epoch": 0.154352,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.52483870967742e-05,
+      "loss": 0.1979,
+      "step": 9647
+    },
+    {
+      "epoch": 0.154368,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.52467741935484e-05,
+      "loss": 0.1379,
+      "step": 9648
+    },
+    {
+      "epoch": 0.154384,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.524516129032258e-05,
+      "loss": 0.1844,
+      "step": 9649
+    },
+    {
+      "epoch": 0.1544,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.524354838709678e-05,
+      "loss": 0.1747,
+      "step": 9650
+    },
+    {
+      "epoch": 0.154416,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.524193548387097e-05,
+      "loss": 0.1814,
+      "step": 9651
+    },
+    {
+      "epoch": 0.154432,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.524032258064516e-05,
+      "loss": 0.152,
+      "step": 9652
+    },
+    {
+      "epoch": 0.154448,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.523870967741936e-05,
+      "loss": 0.1391,
+      "step": 9653
+    },
+    {
+      "epoch": 0.154464,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.523709677419354e-05,
+      "loss": 0.1704,
+      "step": 9654
+    },
+    {
+      "epoch": 0.15448,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.523548387096774e-05,
+      "loss": 0.155,
+      "step": 9655
+    },
+    {
+      "epoch": 0.154496,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.523387096774194e-05,
+      "loss": 0.1626,
+      "step": 9656
+    },
+    {
+      "epoch": 0.154512,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.523225806451614e-05,
+      "loss": 0.2259,
+      "step": 9657
+    },
+    {
+      "epoch": 0.154528,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.523064516129033e-05,
+      "loss": 0.1606,
+      "step": 9658
+    },
+    {
+      "epoch": 0.154544,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.522903225806453e-05,
+      "loss": 0.1984,
+      "step": 9659
+    },
+    {
+      "epoch": 0.15456,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.522741935483871e-05,
+      "loss": 0.1833,
+      "step": 9660
+    },
+    {
+      "epoch": 0.154576,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.522580645161291e-05,
+      "loss": 0.173,
+      "step": 9661
+    },
+    {
+      "epoch": 0.154592,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.52241935483871e-05,
+      "loss": 0.2,
+      "step": 9662
+    },
+    {
+      "epoch": 0.154608,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.52225806451613e-05,
+      "loss": 0.1624,
+      "step": 9663
+    },
+    {
+      "epoch": 0.154624,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.522096774193548e-05,
+      "loss": 0.1483,
+      "step": 9664
+    },
+    {
+      "epoch": 0.15464,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.521935483870968e-05,
+      "loss": 0.1663,
+      "step": 9665
+    },
+    {
+      "epoch": 0.154656,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.521774193548387e-05,
+      "loss": 0.2033,
+      "step": 9666
+    },
+    {
+      "epoch": 0.154672,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.521612903225807e-05,
+      "loss": 0.1767,
+      "step": 9667
+    },
+    {
+      "epoch": 0.154688,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.521451612903227e-05,
+      "loss": 0.1791,
+      "step": 9668
+    },
+    {
+      "epoch": 0.154704,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.521290322580645e-05,
+      "loss": 0.1398,
+      "step": 9669
+    },
+    {
+      "epoch": 0.15472,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.521129032258065e-05,
+      "loss": 0.1317,
+      "step": 9670
+    },
+    {
+      "epoch": 0.154736,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.520967741935484e-05,
+      "loss": 0.1501,
+      "step": 9671
+    },
+    {
+      "epoch": 0.154752,
+      "grad_norm": 0.625,
+      "learning_rate": 8.520806451612904e-05,
+      "loss": 0.1763,
+      "step": 9672
+    },
+    {
+      "epoch": 0.154768,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.520645161290323e-05,
+      "loss": 0.1999,
+      "step": 9673
+    },
+    {
+      "epoch": 0.154784,
+      "grad_norm": 0.498046875,
+      "learning_rate": 8.520483870967743e-05,
+      "loss": 0.1717,
+      "step": 9674
+    },
+    {
+      "epoch": 0.1548,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.520322580645161e-05,
+      "loss": 0.1774,
+      "step": 9675
+    },
+    {
+      "epoch": 0.154816,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.520161290322581e-05,
+      "loss": 0.1868,
+      "step": 9676
+    },
+    {
+      "epoch": 0.154832,
+      "grad_norm": 0.625,
+      "learning_rate": 8.52e-05,
+      "loss": 0.1448,
+      "step": 9677
+    },
+    {
+      "epoch": 0.154848,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.51983870967742e-05,
+      "loss": 0.1543,
+      "step": 9678
+    },
+    {
+      "epoch": 0.154864,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.519677419354838e-05,
+      "loss": 0.1789,
+      "step": 9679
+    },
+    {
+      "epoch": 0.15488,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.519516129032258e-05,
+      "loss": 0.1494,
+      "step": 9680
+    },
+    {
+      "epoch": 0.154896,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.519354838709678e-05,
+      "loss": 0.1372,
+      "step": 9681
+    },
+    {
+      "epoch": 0.154912,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.519193548387098e-05,
+      "loss": 0.187,
+      "step": 9682
+    },
+    {
+      "epoch": 0.154928,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.519032258064517e-05,
+      "loss": 0.2058,
+      "step": 9683
+    },
+    {
+      "epoch": 0.154944,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.518870967741935e-05,
+      "loss": 0.2087,
+      "step": 9684
+    },
+    {
+      "epoch": 0.15496,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.518709677419355e-05,
+      "loss": 0.1849,
+      "step": 9685
+    },
+    {
+      "epoch": 0.154976,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.518548387096774e-05,
+      "loss": 0.1959,
+      "step": 9686
+    },
+    {
+      "epoch": 0.154992,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.518387096774194e-05,
+      "loss": 0.2087,
+      "step": 9687
+    },
+    {
+      "epoch": 0.155008,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.518225806451613e-05,
+      "loss": 0.1807,
+      "step": 9688
+    },
+    {
+      "epoch": 0.155024,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.518064516129033e-05,
+      "loss": 0.1505,
+      "step": 9689
+    },
+    {
+      "epoch": 0.15504,
+      "grad_norm": 1.296875,
+      "learning_rate": 8.517903225806451e-05,
+      "loss": 0.1715,
+      "step": 9690
+    },
+    {
+      "epoch": 0.155056,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.517741935483871e-05,
+      "loss": 0.1768,
+      "step": 9691
+    },
+    {
+      "epoch": 0.155072,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.517580645161291e-05,
+      "loss": 0.1456,
+      "step": 9692
+    },
+    {
+      "epoch": 0.155088,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.517419354838711e-05,
+      "loss": 0.1814,
+      "step": 9693
+    },
+    {
+      "epoch": 0.155104,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.51725806451613e-05,
+      "loss": 0.1501,
+      "step": 9694
+    },
+    {
+      "epoch": 0.15512,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.51709677419355e-05,
+      "loss": 0.184,
+      "step": 9695
+    },
+    {
+      "epoch": 0.155136,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.516935483870968e-05,
+      "loss": 0.1991,
+      "step": 9696
+    },
+    {
+      "epoch": 0.155152,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.516774193548388e-05,
+      "loss": 0.1732,
+      "step": 9697
+    },
+    {
+      "epoch": 0.155168,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.516612903225807e-05,
+      "loss": 0.1617,
+      "step": 9698
+    },
+    {
+      "epoch": 0.155184,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.516451612903225e-05,
+      "loss": 0.1858,
+      "step": 9699
+    },
+    {
+      "epoch": 0.1552,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.516290322580645e-05,
+      "loss": 0.1705,
+      "step": 9700
+    },
+    {
+      "epoch": 0.155216,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.516129032258064e-05,
+      "loss": 0.1614,
+      "step": 9701
+    },
+    {
+      "epoch": 0.155232,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.515967741935484e-05,
+      "loss": 0.1679,
+      "step": 9702
+    },
+    {
+      "epoch": 0.155248,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.515806451612904e-05,
+      "loss": 0.1661,
+      "step": 9703
+    },
+    {
+      "epoch": 0.155264,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.515645161290324e-05,
+      "loss": 0.1899,
+      "step": 9704
+    },
+    {
+      "epoch": 0.15528,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.515483870967742e-05,
+      "loss": 0.14,
+      "step": 9705
+    },
+    {
+      "epoch": 0.155296,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.515322580645162e-05,
+      "loss": 0.1739,
+      "step": 9706
+    },
+    {
+      "epoch": 0.155312,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.515161290322581e-05,
+      "loss": 0.189,
+      "step": 9707
+    },
+    {
+      "epoch": 0.155328,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.515000000000001e-05,
+      "loss": 0.1325,
+      "step": 9708
+    },
+    {
+      "epoch": 0.155344,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.51483870967742e-05,
+      "loss": 0.2029,
+      "step": 9709
+    },
+    {
+      "epoch": 0.15536,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.51467741935484e-05,
+      "loss": 0.1495,
+      "step": 9710
+    },
+    {
+      "epoch": 0.155376,
+      "grad_norm": 0.75,
+      "learning_rate": 8.514516129032258e-05,
+      "loss": 0.1912,
+      "step": 9711
+    },
+    {
+      "epoch": 0.155392,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.514354838709678e-05,
+      "loss": 0.1976,
+      "step": 9712
+    },
+    {
+      "epoch": 0.155408,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.514193548387097e-05,
+      "loss": 0.1829,
+      "step": 9713
+    },
+    {
+      "epoch": 0.155424,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.514032258064515e-05,
+      "loss": 0.2139,
+      "step": 9714
+    },
+    {
+      "epoch": 0.15544,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.513870967741935e-05,
+      "loss": 0.1601,
+      "step": 9715
+    },
+    {
+      "epoch": 0.155456,
+      "grad_norm": 0.875,
+      "learning_rate": 8.513709677419355e-05,
+      "loss": 0.2173,
+      "step": 9716
+    },
+    {
+      "epoch": 0.155472,
+      "grad_norm": 1.3828125,
+      "learning_rate": 8.513548387096775e-05,
+      "loss": 0.1999,
+      "step": 9717
+    },
+    {
+      "epoch": 0.155488,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.513387096774194e-05,
+      "loss": 0.222,
+      "step": 9718
+    },
+    {
+      "epoch": 0.155504,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.513225806451614e-05,
+      "loss": 0.1812,
+      "step": 9719
+    },
+    {
+      "epoch": 0.15552,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.513064516129032e-05,
+      "loss": 0.196,
+      "step": 9720
+    },
+    {
+      "epoch": 0.155536,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.512903225806452e-05,
+      "loss": 0.1462,
+      "step": 9721
+    },
+    {
+      "epoch": 0.155552,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.512741935483871e-05,
+      "loss": 0.1752,
+      "step": 9722
+    },
+    {
+      "epoch": 0.155568,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.512580645161291e-05,
+      "loss": 0.1428,
+      "step": 9723
+    },
+    {
+      "epoch": 0.155584,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.51241935483871e-05,
+      "loss": 0.1187,
+      "step": 9724
+    },
+    {
+      "epoch": 0.1556,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.51225806451613e-05,
+      "loss": 0.1806,
+      "step": 9725
+    },
+    {
+      "epoch": 0.155616,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.512096774193548e-05,
+      "loss": 0.1949,
+      "step": 9726
+    },
+    {
+      "epoch": 0.155632,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.511935483870968e-05,
+      "loss": 0.1526,
+      "step": 9727
+    },
+    {
+      "epoch": 0.155648,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.511774193548388e-05,
+      "loss": 0.16,
+      "step": 9728
+    },
+    {
+      "epoch": 0.155664,
+      "grad_norm": 1.28125,
+      "learning_rate": 8.511612903225808e-05,
+      "loss": 0.1694,
+      "step": 9729
+    },
+    {
+      "epoch": 0.15568,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.511451612903227e-05,
+      "loss": 0.2019,
+      "step": 9730
+    },
+    {
+      "epoch": 0.155696,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.511290322580645e-05,
+      "loss": 0.1835,
+      "step": 9731
+    },
+    {
+      "epoch": 0.155712,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.511129032258065e-05,
+      "loss": 0.1889,
+      "step": 9732
+    },
+    {
+      "epoch": 0.155728,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.510967741935484e-05,
+      "loss": 0.1779,
+      "step": 9733
+    },
+    {
+      "epoch": 0.155744,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.510806451612904e-05,
+      "loss": 0.1791,
+      "step": 9734
+    },
+    {
+      "epoch": 0.15576,
+      "grad_norm": 1.3828125,
+      "learning_rate": 8.510645161290322e-05,
+      "loss": 0.2105,
+      "step": 9735
+    },
+    {
+      "epoch": 0.155776,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.510483870967742e-05,
+      "loss": 0.1457,
+      "step": 9736
+    },
+    {
+      "epoch": 0.155792,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.510322580645161e-05,
+      "loss": 0.1977,
+      "step": 9737
+    },
+    {
+      "epoch": 0.155808,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.510161290322581e-05,
+      "loss": 0.2022,
+      "step": 9738
+    },
+    {
+      "epoch": 0.155824,
+      "grad_norm": 1.2734375,
+      "learning_rate": 8.510000000000001e-05,
+      "loss": 0.1814,
+      "step": 9739
+    },
+    {
+      "epoch": 0.15584,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.509838709677421e-05,
+      "loss": 0.1598,
+      "step": 9740
+    },
+    {
+      "epoch": 0.155856,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.50967741935484e-05,
+      "loss": 0.1874,
+      "step": 9741
+    },
+    {
+      "epoch": 0.155872,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.50951612903226e-05,
+      "loss": 0.173,
+      "step": 9742
+    },
+    {
+      "epoch": 0.155888,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.509354838709678e-05,
+      "loss": 0.1377,
+      "step": 9743
+    },
+    {
+      "epoch": 0.155904,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.509193548387098e-05,
+      "loss": 0.1955,
+      "step": 9744
+    },
+    {
+      "epoch": 0.15592,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.509032258064517e-05,
+      "loss": 0.1699,
+      "step": 9745
+    },
+    {
+      "epoch": 0.155936,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.508870967741935e-05,
+      "loss": 0.1688,
+      "step": 9746
+    },
+    {
+      "epoch": 0.155952,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.508709677419355e-05,
+      "loss": 0.1671,
+      "step": 9747
+    },
+    {
+      "epoch": 0.155968,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.508548387096774e-05,
+      "loss": 0.1794,
+      "step": 9748
+    },
+    {
+      "epoch": 0.155984,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.508387096774194e-05,
+      "loss": 0.1756,
+      "step": 9749
+    },
+    {
+      "epoch": 0.156,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.508225806451612e-05,
+      "loss": 0.1527,
+      "step": 9750
+    },
+    {
+      "epoch": 0.156016,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.508064516129032e-05,
+      "loss": 0.1784,
+      "step": 9751
+    },
+    {
+      "epoch": 0.156032,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.507903225806452e-05,
+      "loss": 0.1606,
+      "step": 9752
+    },
+    {
+      "epoch": 0.156048,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.507741935483872e-05,
+      "loss": 0.1921,
+      "step": 9753
+    },
+    {
+      "epoch": 0.156064,
+      "grad_norm": 1.28125,
+      "learning_rate": 8.507580645161291e-05,
+      "loss": 0.1985,
+      "step": 9754
+    },
+    {
+      "epoch": 0.15608,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.507419354838711e-05,
+      "loss": 0.1731,
+      "step": 9755
+    },
+    {
+      "epoch": 0.156096,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.50725806451613e-05,
+      "loss": 0.1882,
+      "step": 9756
+    },
+    {
+      "epoch": 0.156112,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.50709677419355e-05,
+      "loss": 0.1772,
+      "step": 9757
+    },
+    {
+      "epoch": 0.156128,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.506935483870968e-05,
+      "loss": 0.1723,
+      "step": 9758
+    },
+    {
+      "epoch": 0.156144,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.506774193548388e-05,
+      "loss": 0.1696,
+      "step": 9759
+    },
+    {
+      "epoch": 0.15616,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.506612903225807e-05,
+      "loss": 0.1701,
+      "step": 9760
+    },
+    {
+      "epoch": 0.156176,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.506451612903225e-05,
+      "loss": 0.1732,
+      "step": 9761
+    },
+    {
+      "epoch": 0.156192,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.506290322580645e-05,
+      "loss": 0.2059,
+      "step": 9762
+    },
+    {
+      "epoch": 0.156208,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.506129032258065e-05,
+      "loss": 0.2289,
+      "step": 9763
+    },
+    {
+      "epoch": 0.156224,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.505967741935485e-05,
+      "loss": 0.154,
+      "step": 9764
+    },
+    {
+      "epoch": 0.15624,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.505806451612904e-05,
+      "loss": 0.1734,
+      "step": 9765
+    },
+    {
+      "epoch": 0.156256,
+      "grad_norm": 1.4375,
+      "learning_rate": 8.505645161290324e-05,
+      "loss": 0.1865,
+      "step": 9766
+    },
+    {
+      "epoch": 0.156272,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.505483870967742e-05,
+      "loss": 0.1799,
+      "step": 9767
+    },
+    {
+      "epoch": 0.156288,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.505322580645162e-05,
+      "loss": 0.1563,
+      "step": 9768
+    },
+    {
+      "epoch": 0.156304,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.505161290322581e-05,
+      "loss": 0.1641,
+      "step": 9769
+    },
+    {
+      "epoch": 0.15632,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.505000000000001e-05,
+      "loss": 0.1726,
+      "step": 9770
+    },
+    {
+      "epoch": 0.156336,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.50483870967742e-05,
+      "loss": 0.2057,
+      "step": 9771
+    },
+    {
+      "epoch": 0.156352,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.504677419354839e-05,
+      "loss": 0.1849,
+      "step": 9772
+    },
+    {
+      "epoch": 0.156368,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.504516129032258e-05,
+      "loss": 0.165,
+      "step": 9773
+    },
+    {
+      "epoch": 0.156384,
+      "grad_norm": 0.5,
+      "learning_rate": 8.504354838709678e-05,
+      "loss": 0.1651,
+      "step": 9774
+    },
+    {
+      "epoch": 0.1564,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.504193548387096e-05,
+      "loss": 0.1673,
+      "step": 9775
+    },
+    {
+      "epoch": 0.156416,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.504032258064516e-05,
+      "loss": 0.1337,
+      "step": 9776
+    },
+    {
+      "epoch": 0.156432,
+      "grad_norm": 1.3515625,
+      "learning_rate": 8.503870967741936e-05,
+      "loss": 0.208,
+      "step": 9777
+    },
+    {
+      "epoch": 0.156448,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.503709677419355e-05,
+      "loss": 0.2025,
+      "step": 9778
+    },
+    {
+      "epoch": 0.156464,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.503548387096775e-05,
+      "loss": 0.2053,
+      "step": 9779
+    },
+    {
+      "epoch": 0.15648,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.503387096774194e-05,
+      "loss": 0.154,
+      "step": 9780
+    },
+    {
+      "epoch": 0.156496,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.503225806451614e-05,
+      "loss": 0.159,
+      "step": 9781
+    },
+    {
+      "epoch": 0.156512,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.503064516129032e-05,
+      "loss": 0.1515,
+      "step": 9782
+    },
+    {
+      "epoch": 0.156528,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.502903225806452e-05,
+      "loss": 0.1816,
+      "step": 9783
+    },
+    {
+      "epoch": 0.156544,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.502741935483871e-05,
+      "loss": 0.1474,
+      "step": 9784
+    },
+    {
+      "epoch": 0.15656,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.502580645161291e-05,
+      "loss": 0.1578,
+      "step": 9785
+    },
+    {
+      "epoch": 0.156576,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.502419354838709e-05,
+      "loss": 0.16,
+      "step": 9786
+    },
+    {
+      "epoch": 0.156592,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.502258064516129e-05,
+      "loss": 0.1891,
+      "step": 9787
+    },
+    {
+      "epoch": 0.156608,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.502096774193549e-05,
+      "loss": 0.1899,
+      "step": 9788
+    },
+    {
+      "epoch": 0.156624,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.501935483870969e-05,
+      "loss": 0.1946,
+      "step": 9789
+    },
+    {
+      "epoch": 0.15664,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.501774193548388e-05,
+      "loss": 0.1594,
+      "step": 9790
+    },
+    {
+      "epoch": 0.156656,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.501612903225808e-05,
+      "loss": 0.1829,
+      "step": 9791
+    },
+    {
+      "epoch": 0.156672,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.501451612903226e-05,
+      "loss": 0.2085,
+      "step": 9792
+    },
+    {
+      "epoch": 0.156688,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.501290322580645e-05,
+      "loss": 0.184,
+      "step": 9793
+    },
+    {
+      "epoch": 0.156704,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.501129032258065e-05,
+      "loss": 0.1634,
+      "step": 9794
+    },
+    {
+      "epoch": 0.15672,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.500967741935484e-05,
+      "loss": 0.1782,
+      "step": 9795
+    },
+    {
+      "epoch": 0.156736,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.500806451612904e-05,
+      "loss": 0.157,
+      "step": 9796
+    },
+    {
+      "epoch": 0.156752,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.500645161290322e-05,
+      "loss": 0.1493,
+      "step": 9797
+    },
+    {
+      "epoch": 0.156768,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.500483870967742e-05,
+      "loss": 0.1719,
+      "step": 9798
+    },
+    {
+      "epoch": 0.156784,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.500322580645162e-05,
+      "loss": 0.1394,
+      "step": 9799
+    },
+    {
+      "epoch": 0.1568,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.500161290322582e-05,
+      "loss": 0.1757,
+      "step": 9800
+    },
+    {
+      "epoch": 0.156816,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.5e-05,
+      "loss": 0.1597,
+      "step": 9801
+    },
+    {
+      "epoch": 0.156832,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.49983870967742e-05,
+      "loss": 0.205,
+      "step": 9802
+    },
+    {
+      "epoch": 0.156848,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.499677419354839e-05,
+      "loss": 0.1346,
+      "step": 9803
+    },
+    {
+      "epoch": 0.156864,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.499516129032259e-05,
+      "loss": 0.2072,
+      "step": 9804
+    },
+    {
+      "epoch": 0.15688,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.499354838709678e-05,
+      "loss": 0.1332,
+      "step": 9805
+    },
+    {
+      "epoch": 0.156896,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.499193548387098e-05,
+      "loss": 0.2004,
+      "step": 9806
+    },
+    {
+      "epoch": 0.156912,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.499032258064516e-05,
+      "loss": 0.1454,
+      "step": 9807
+    },
+    {
+      "epoch": 0.156928,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.498870967741935e-05,
+      "loss": 0.1731,
+      "step": 9808
+    },
+    {
+      "epoch": 0.156944,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.498709677419355e-05,
+      "loss": 0.1639,
+      "step": 9809
+    },
+    {
+      "epoch": 0.15696,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.498548387096774e-05,
+      "loss": 0.1486,
+      "step": 9810
+    },
+    {
+      "epoch": 0.156976,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.498387096774193e-05,
+      "loss": 0.1439,
+      "step": 9811
+    },
+    {
+      "epoch": 0.156992,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.498225806451613e-05,
+      "loss": 0.1962,
+      "step": 9812
+    },
+    {
+      "epoch": 0.157008,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.498064516129033e-05,
+      "loss": 0.1723,
+      "step": 9813
+    },
+    {
+      "epoch": 0.157024,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.497903225806452e-05,
+      "loss": 0.1839,
+      "step": 9814
+    },
+    {
+      "epoch": 0.15704,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.497741935483872e-05,
+      "loss": 0.1562,
+      "step": 9815
+    },
+    {
+      "epoch": 0.157056,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.49758064516129e-05,
+      "loss": 0.2088,
+      "step": 9816
+    },
+    {
+      "epoch": 0.157072,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.49741935483871e-05,
+      "loss": 0.2046,
+      "step": 9817
+    },
+    {
+      "epoch": 0.157088,
+      "grad_norm": 1.3828125,
+      "learning_rate": 8.497258064516129e-05,
+      "loss": 0.1853,
+      "step": 9818
+    },
+    {
+      "epoch": 0.157104,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.497096774193549e-05,
+      "loss": 0.1346,
+      "step": 9819
+    },
+    {
+      "epoch": 0.15712,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.496935483870968e-05,
+      "loss": 0.163,
+      "step": 9820
+    },
+    {
+      "epoch": 0.157136,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.496774193548388e-05,
+      "loss": 0.195,
+      "step": 9821
+    },
+    {
+      "epoch": 0.157152,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.496612903225806e-05,
+      "loss": 0.1568,
+      "step": 9822
+    },
+    {
+      "epoch": 0.157168,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.496451612903226e-05,
+      "loss": 0.2109,
+      "step": 9823
+    },
+    {
+      "epoch": 0.157184,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.496290322580646e-05,
+      "loss": 0.2136,
+      "step": 9824
+    },
+    {
+      "epoch": 0.1572,
+      "grad_norm": 1.234375,
+      "learning_rate": 8.496129032258065e-05,
+      "loss": 0.165,
+      "step": 9825
+    },
+    {
+      "epoch": 0.157216,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.495967741935485e-05,
+      "loss": 0.1321,
+      "step": 9826
+    },
+    {
+      "epoch": 0.157232,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.495806451612903e-05,
+      "loss": 0.2078,
+      "step": 9827
+    },
+    {
+      "epoch": 0.157248,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.495645161290323e-05,
+      "loss": 0.1688,
+      "step": 9828
+    },
+    {
+      "epoch": 0.157264,
+      "grad_norm": 0.625,
+      "learning_rate": 8.495483870967742e-05,
+      "loss": 0.1525,
+      "step": 9829
+    },
+    {
+      "epoch": 0.15728,
+      "grad_norm": 0.875,
+      "learning_rate": 8.495322580645162e-05,
+      "loss": 0.2195,
+      "step": 9830
+    },
+    {
+      "epoch": 0.157296,
+      "grad_norm": 0.482421875,
+      "learning_rate": 8.49516129032258e-05,
+      "loss": 0.1677,
+      "step": 9831
+    },
+    {
+      "epoch": 0.157312,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.495e-05,
+      "loss": 0.1651,
+      "step": 9832
+    },
+    {
+      "epoch": 0.157328,
+      "grad_norm": 0.474609375,
+      "learning_rate": 8.494838709677419e-05,
+      "loss": 0.143,
+      "step": 9833
+    },
+    {
+      "epoch": 0.157344,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.494677419354839e-05,
+      "loss": 0.1718,
+      "step": 9834
+    },
+    {
+      "epoch": 0.15736,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.494516129032259e-05,
+      "loss": 0.1832,
+      "step": 9835
+    },
+    {
+      "epoch": 0.157376,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.494354838709678e-05,
+      "loss": 0.2117,
+      "step": 9836
+    },
+    {
+      "epoch": 0.157392,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.494193548387098e-05,
+      "loss": 0.1623,
+      "step": 9837
+    },
+    {
+      "epoch": 0.157408,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.494032258064518e-05,
+      "loss": 0.1903,
+      "step": 9838
+    },
+    {
+      "epoch": 0.157424,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.493870967741936e-05,
+      "loss": 0.1833,
+      "step": 9839
+    },
+    {
+      "epoch": 0.15744,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.493709677419355e-05,
+      "loss": 0.1907,
+      "step": 9840
+    },
+    {
+      "epoch": 0.157456,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.493548387096775e-05,
+      "loss": 0.1733,
+      "step": 9841
+    },
+    {
+      "epoch": 0.157472,
+      "grad_norm": 1.28125,
+      "learning_rate": 8.493387096774193e-05,
+      "loss": 0.1851,
+      "step": 9842
+    },
+    {
+      "epoch": 0.157488,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.493225806451613e-05,
+      "loss": 0.1483,
+      "step": 9843
+    },
+    {
+      "epoch": 0.157504,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.493064516129032e-05,
+      "loss": 0.1764,
+      "step": 9844
+    },
+    {
+      "epoch": 0.15752,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.492903225806452e-05,
+      "loss": 0.2074,
+      "step": 9845
+    },
+    {
+      "epoch": 0.157536,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.49274193548387e-05,
+      "loss": 0.1537,
+      "step": 9846
+    },
+    {
+      "epoch": 0.157552,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.49258064516129e-05,
+      "loss": 0.1799,
+      "step": 9847
+    },
+    {
+      "epoch": 0.157568,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.49241935483871e-05,
+      "loss": 0.1857,
+      "step": 9848
+    },
+    {
+      "epoch": 0.157584,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.49225806451613e-05,
+      "loss": 0.1671,
+      "step": 9849
+    },
+    {
+      "epoch": 0.1576,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.492096774193549e-05,
+      "loss": 0.1439,
+      "step": 9850
+    },
+    {
+      "epoch": 0.157616,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.491935483870969e-05,
+      "loss": 0.182,
+      "step": 9851
+    },
+    {
+      "epoch": 0.157632,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.491774193548388e-05,
+      "loss": 0.181,
+      "step": 9852
+    },
+    {
+      "epoch": 0.157648,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.491612903225808e-05,
+      "loss": 0.1832,
+      "step": 9853
+    },
+    {
+      "epoch": 0.157664,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.491451612903226e-05,
+      "loss": 0.1635,
+      "step": 9854
+    },
+    {
+      "epoch": 0.15768,
+      "grad_norm": 1.296875,
+      "learning_rate": 8.491290322580645e-05,
+      "loss": 0.168,
+      "step": 9855
+    },
+    {
+      "epoch": 0.157696,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.491129032258065e-05,
+      "loss": 0.1574,
+      "step": 9856
+    },
+    {
+      "epoch": 0.157712,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.490967741935483e-05,
+      "loss": 0.189,
+      "step": 9857
+    },
+    {
+      "epoch": 0.157728,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.490806451612903e-05,
+      "loss": 0.2058,
+      "step": 9858
+    },
+    {
+      "epoch": 0.157744,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.490645161290323e-05,
+      "loss": 0.1703,
+      "step": 9859
+    },
+    {
+      "epoch": 0.15776,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.490483870967743e-05,
+      "loss": 0.1857,
+      "step": 9860
+    },
+    {
+      "epoch": 0.157776,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.490322580645162e-05,
+      "loss": 0.1961,
+      "step": 9861
+    },
+    {
+      "epoch": 0.157792,
+      "grad_norm": 1.390625,
+      "learning_rate": 8.490161290322582e-05,
+      "loss": 0.1957,
+      "step": 9862
+    },
+    {
+      "epoch": 0.157808,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.49e-05,
+      "loss": 0.1903,
+      "step": 9863
+    },
+    {
+      "epoch": 0.157824,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.48983870967742e-05,
+      "loss": 0.1672,
+      "step": 9864
+    },
+    {
+      "epoch": 0.15784,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.489677419354839e-05,
+      "loss": 0.1703,
+      "step": 9865
+    },
+    {
+      "epoch": 0.157856,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.489516129032259e-05,
+      "loss": 0.1522,
+      "step": 9866
+    },
+    {
+      "epoch": 0.157872,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.489354838709678e-05,
+      "loss": 0.1819,
+      "step": 9867
+    },
+    {
+      "epoch": 0.157888,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.489193548387097e-05,
+      "loss": 0.1624,
+      "step": 9868
+    },
+    {
+      "epoch": 0.157904,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.489032258064516e-05,
+      "loss": 0.1589,
+      "step": 9869
+    },
+    {
+      "epoch": 0.15792,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.488870967741935e-05,
+      "loss": 0.1658,
+      "step": 9870
+    },
+    {
+      "epoch": 0.157936,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.488709677419355e-05,
+      "loss": 0.1868,
+      "step": 9871
+    },
+    {
+      "epoch": 0.157952,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.488548387096775e-05,
+      "loss": 0.1527,
+      "step": 9872
+    },
+    {
+      "epoch": 0.157968,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.488387096774195e-05,
+      "loss": 0.1833,
+      "step": 9873
+    },
+    {
+      "epoch": 0.157984,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.488225806451613e-05,
+      "loss": 0.1641,
+      "step": 9874
+    },
+    {
+      "epoch": 0.158,
+      "grad_norm": 0.625,
+      "learning_rate": 8.488064516129033e-05,
+      "loss": 0.1595,
+      "step": 9875
+    },
+    {
+      "epoch": 0.158016,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.487903225806452e-05,
+      "loss": 0.1498,
+      "step": 9876
+    },
+    {
+      "epoch": 0.158032,
+      "grad_norm": 1.3984375,
+      "learning_rate": 8.487741935483872e-05,
+      "loss": 0.1999,
+      "step": 9877
+    },
+    {
+      "epoch": 0.158048,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.48758064516129e-05,
+      "loss": 0.1641,
+      "step": 9878
+    },
+    {
+      "epoch": 0.158064,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.48741935483871e-05,
+      "loss": 0.1988,
+      "step": 9879
+    },
+    {
+      "epoch": 0.15808,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.487258064516129e-05,
+      "loss": 0.1327,
+      "step": 9880
+    },
+    {
+      "epoch": 0.158096,
+      "grad_norm": 0.625,
+      "learning_rate": 8.487096774193549e-05,
+      "loss": 0.1702,
+      "step": 9881
+    },
+    {
+      "epoch": 0.158112,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.486935483870967e-05,
+      "loss": 0.1555,
+      "step": 9882
+    },
+    {
+      "epoch": 0.158128,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.486774193548387e-05,
+      "loss": 0.2021,
+      "step": 9883
+    },
+    {
+      "epoch": 0.158144,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.486612903225807e-05,
+      "loss": 0.1853,
+      "step": 9884
+    },
+    {
+      "epoch": 0.15816,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.486451612903227e-05,
+      "loss": 0.1643,
+      "step": 9885
+    },
+    {
+      "epoch": 0.158176,
+      "grad_norm": 1.2109375,
+      "learning_rate": 8.486290322580646e-05,
+      "loss": 0.2236,
+      "step": 9886
+    },
+    {
+      "epoch": 0.158192,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.486129032258065e-05,
+      "loss": 0.2042,
+      "step": 9887
+    },
+    {
+      "epoch": 0.158208,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.485967741935485e-05,
+      "loss": 0.1597,
+      "step": 9888
+    },
+    {
+      "epoch": 0.158224,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.485806451612903e-05,
+      "loss": 0.171,
+      "step": 9889
+    },
+    {
+      "epoch": 0.15824,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.485645161290323e-05,
+      "loss": 0.1727,
+      "step": 9890
+    },
+    {
+      "epoch": 0.158256,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.485483870967742e-05,
+      "loss": 0.1616,
+      "step": 9891
+    },
+    {
+      "epoch": 0.158272,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.485322580645162e-05,
+      "loss": 0.1868,
+      "step": 9892
+    },
+    {
+      "epoch": 0.158288,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.48516129032258e-05,
+      "loss": 0.1649,
+      "step": 9893
+    },
+    {
+      "epoch": 0.158304,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.485e-05,
+      "loss": 0.2231,
+      "step": 9894
+    },
+    {
+      "epoch": 0.15832,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.48483870967742e-05,
+      "loss": 0.1895,
+      "step": 9895
+    },
+    {
+      "epoch": 0.158336,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.48467741935484e-05,
+      "loss": 0.1601,
+      "step": 9896
+    },
+    {
+      "epoch": 0.158352,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.484516129032259e-05,
+      "loss": 0.1762,
+      "step": 9897
+    },
+    {
+      "epoch": 0.158368,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.484354838709679e-05,
+      "loss": 0.1834,
+      "step": 9898
+    },
+    {
+      "epoch": 0.158384,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.484193548387097e-05,
+      "loss": 0.1742,
+      "step": 9899
+    },
+    {
+      "epoch": 0.1584,
+      "grad_norm": 1.0,
+      "learning_rate": 8.484032258064517e-05,
+      "loss": 0.1741,
+      "step": 9900
+    },
+    {
+      "epoch": 0.158416,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.483870967741936e-05,
+      "loss": 0.1898,
+      "step": 9901
+    },
+    {
+      "epoch": 0.158432,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.483709677419355e-05,
+      "loss": 0.2124,
+      "step": 9902
+    },
+    {
+      "epoch": 0.158448,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.483548387096774e-05,
+      "loss": 0.202,
+      "step": 9903
+    },
+    {
+      "epoch": 0.158464,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.483387096774193e-05,
+      "loss": 0.1613,
+      "step": 9904
+    },
+    {
+      "epoch": 0.15848,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.483225806451613e-05,
+      "loss": 0.185,
+      "step": 9905
+    },
+    {
+      "epoch": 0.158496,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.483064516129032e-05,
+      "loss": 0.1231,
+      "step": 9906
+    },
+    {
+      "epoch": 0.158512,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.482903225806452e-05,
+      "loss": 0.1801,
+      "step": 9907
+    },
+    {
+      "epoch": 0.158528,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.482741935483872e-05,
+      "loss": 0.1788,
+      "step": 9908
+    },
+    {
+      "epoch": 0.158544,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.482580645161292e-05,
+      "loss": 0.1928,
+      "step": 9909
+    },
+    {
+      "epoch": 0.15856,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.48241935483871e-05,
+      "loss": 0.235,
+      "step": 9910
+    },
+    {
+      "epoch": 0.158576,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.48225806451613e-05,
+      "loss": 0.1836,
+      "step": 9911
+    },
+    {
+      "epoch": 0.158592,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.482096774193549e-05,
+      "loss": 0.1543,
+      "step": 9912
+    },
+    {
+      "epoch": 0.158608,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.481935483870969e-05,
+      "loss": 0.1752,
+      "step": 9913
+    },
+    {
+      "epoch": 0.158624,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.481774193548387e-05,
+      "loss": 0.1771,
+      "step": 9914
+    },
+    {
+      "epoch": 0.15864,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.481612903225807e-05,
+      "loss": 0.175,
+      "step": 9915
+    },
+    {
+      "epoch": 0.158656,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.481451612903226e-05,
+      "loss": 0.1738,
+      "step": 9916
+    },
+    {
+      "epoch": 0.158672,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.481290322580644e-05,
+      "loss": 0.1891,
+      "step": 9917
+    },
+    {
+      "epoch": 0.158688,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.481129032258064e-05,
+      "loss": 0.1731,
+      "step": 9918
+    },
+    {
+      "epoch": 0.158704,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.480967741935484e-05,
+      "loss": 0.2014,
+      "step": 9919
+    },
+    {
+      "epoch": 0.15872,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.480806451612904e-05,
+      "loss": 0.224,
+      "step": 9920
+    },
+    {
+      "epoch": 0.158736,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.480645161290323e-05,
+      "loss": 0.2052,
+      "step": 9921
+    },
+    {
+      "epoch": 0.158752,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.480483870967743e-05,
+      "loss": 0.159,
+      "step": 9922
+    },
+    {
+      "epoch": 0.158768,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.480322580645162e-05,
+      "loss": 0.1771,
+      "step": 9923
+    },
+    {
+      "epoch": 0.158784,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.480161290322582e-05,
+      "loss": 0.1555,
+      "step": 9924
+    },
+    {
+      "epoch": 0.1588,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.48e-05,
+      "loss": 0.1261,
+      "step": 9925
+    },
+    {
+      "epoch": 0.158816,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.47983870967742e-05,
+      "loss": 0.1776,
+      "step": 9926
+    },
+    {
+      "epoch": 0.158832,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.479677419354839e-05,
+      "loss": 0.1517,
+      "step": 9927
+    },
+    {
+      "epoch": 0.158848,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.479516129032259e-05,
+      "loss": 0.1574,
+      "step": 9928
+    },
+    {
+      "epoch": 0.158864,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.479354838709677e-05,
+      "loss": 0.1696,
+      "step": 9929
+    },
+    {
+      "epoch": 0.15888,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.479193548387097e-05,
+      "loss": 0.185,
+      "step": 9930
+    },
+    {
+      "epoch": 0.158896,
+      "grad_norm": 1.3125,
+      "learning_rate": 8.479032258064516e-05,
+      "loss": 0.1834,
+      "step": 9931
+    },
+    {
+      "epoch": 0.158912,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.478870967741936e-05,
+      "loss": 0.2181,
+      "step": 9932
+    },
+    {
+      "epoch": 0.158928,
+      "grad_norm": 0.625,
+      "learning_rate": 8.478709677419356e-05,
+      "loss": 0.158,
+      "step": 9933
+    },
+    {
+      "epoch": 0.158944,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.478548387096774e-05,
+      "loss": 0.1811,
+      "step": 9934
+    },
+    {
+      "epoch": 0.15896,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.478387096774194e-05,
+      "loss": 0.1813,
+      "step": 9935
+    },
+    {
+      "epoch": 0.158976,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.478225806451613e-05,
+      "loss": 0.1775,
+      "step": 9936
+    },
+    {
+      "epoch": 0.158992,
+      "grad_norm": 1.25,
+      "learning_rate": 8.478064516129033e-05,
+      "loss": 0.1635,
+      "step": 9937
+    },
+    {
+      "epoch": 0.159008,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.477903225806452e-05,
+      "loss": 0.1425,
+      "step": 9938
+    },
+    {
+      "epoch": 0.159024,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.477741935483871e-05,
+      "loss": 0.1856,
+      "step": 9939
+    },
+    {
+      "epoch": 0.15904,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.47758064516129e-05,
+      "loss": 0.1618,
+      "step": 9940
+    },
+    {
+      "epoch": 0.159056,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.47741935483871e-05,
+      "loss": 0.1781,
+      "step": 9941
+    },
+    {
+      "epoch": 0.159072,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.477258064516129e-05,
+      "loss": 0.2148,
+      "step": 9942
+    },
+    {
+      "epoch": 0.159088,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.477096774193549e-05,
+      "loss": 0.1825,
+      "step": 9943
+    },
+    {
+      "epoch": 0.159104,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.476935483870969e-05,
+      "loss": 0.192,
+      "step": 9944
+    },
+    {
+      "epoch": 0.15912,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.476774193548389e-05,
+      "loss": 0.1646,
+      "step": 9945
+    },
+    {
+      "epoch": 0.159136,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.476612903225807e-05,
+      "loss": 0.2033,
+      "step": 9946
+    },
+    {
+      "epoch": 0.159152,
+      "grad_norm": 0.625,
+      "learning_rate": 8.476451612903227e-05,
+      "loss": 0.1484,
+      "step": 9947
+    },
+    {
+      "epoch": 0.159168,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.476290322580646e-05,
+      "loss": 0.1778,
+      "step": 9948
+    },
+    {
+      "epoch": 0.159184,
+      "grad_norm": 1.125,
+      "learning_rate": 8.476129032258064e-05,
+      "loss": 0.1794,
+      "step": 9949
+    },
+    {
+      "epoch": 0.1592,
+      "grad_norm": 1.3359375,
+      "learning_rate": 8.475967741935484e-05,
+      "loss": 0.1562,
+      "step": 9950
+    },
+    {
+      "epoch": 0.159216,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.475806451612903e-05,
+      "loss": 0.1838,
+      "step": 9951
+    },
+    {
+      "epoch": 0.159232,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.475645161290323e-05,
+      "loss": 0.2122,
+      "step": 9952
+    },
+    {
+      "epoch": 0.159248,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.475483870967741e-05,
+      "loss": 0.1727,
+      "step": 9953
+    },
+    {
+      "epoch": 0.159264,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.475322580645161e-05,
+      "loss": 0.1817,
+      "step": 9954
+    },
+    {
+      "epoch": 0.15928,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.475161290322581e-05,
+      "loss": 0.1519,
+      "step": 9955
+    },
+    {
+      "epoch": 0.159296,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.475000000000001e-05,
+      "loss": 0.1737,
+      "step": 9956
+    },
+    {
+      "epoch": 0.159312,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.47483870967742e-05,
+      "loss": 0.1847,
+      "step": 9957
+    },
+    {
+      "epoch": 0.159328,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.47467741935484e-05,
+      "loss": 0.1862,
+      "step": 9958
+    },
+    {
+      "epoch": 0.159344,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.474516129032259e-05,
+      "loss": 0.1623,
+      "step": 9959
+    },
+    {
+      "epoch": 0.15936,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.474354838709678e-05,
+      "loss": 0.1328,
+      "step": 9960
+    },
+    {
+      "epoch": 0.159376,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.474193548387097e-05,
+      "loss": 0.176,
+      "step": 9961
+    },
+    {
+      "epoch": 0.159392,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.474032258064517e-05,
+      "loss": 0.1835,
+      "step": 9962
+    },
+    {
+      "epoch": 0.159408,
+      "grad_norm": 1.34375,
+      "learning_rate": 8.473870967741936e-05,
+      "loss": 0.1931,
+      "step": 9963
+    },
+    {
+      "epoch": 0.159424,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.473709677419354e-05,
+      "loss": 0.2107,
+      "step": 9964
+    },
+    {
+      "epoch": 0.15944,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.473548387096774e-05,
+      "loss": 0.1828,
+      "step": 9965
+    },
+    {
+      "epoch": 0.159456,
+      "grad_norm": 1.5390625,
+      "learning_rate": 8.473387096774193e-05,
+      "loss": 0.1877,
+      "step": 9966
+    },
+    {
+      "epoch": 0.159472,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.473225806451613e-05,
+      "loss": 0.1795,
+      "step": 9967
+    },
+    {
+      "epoch": 0.159488,
+      "grad_norm": 1.484375,
+      "learning_rate": 8.473064516129033e-05,
+      "loss": 0.2221,
+      "step": 9968
+    },
+    {
+      "epoch": 0.159504,
+      "grad_norm": 1.25,
+      "learning_rate": 8.472903225806453e-05,
+      "loss": 0.2102,
+      "step": 9969
+    },
+    {
+      "epoch": 0.15952,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.472741935483871e-05,
+      "loss": 0.2197,
+      "step": 9970
+    },
+    {
+      "epoch": 0.159536,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.472580645161291e-05,
+      "loss": 0.1891,
+      "step": 9971
+    },
+    {
+      "epoch": 0.159552,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.47241935483871e-05,
+      "loss": 0.1592,
+      "step": 9972
+    },
+    {
+      "epoch": 0.159568,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.47225806451613e-05,
+      "loss": 0.1806,
+      "step": 9973
+    },
+    {
+      "epoch": 0.159584,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.472096774193548e-05,
+      "loss": 0.219,
+      "step": 9974
+    },
+    {
+      "epoch": 0.1596,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.471935483870968e-05,
+      "loss": 0.1412,
+      "step": 9975
+    },
+    {
+      "epoch": 0.159616,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.471774193548387e-05,
+      "loss": 0.2158,
+      "step": 9976
+    },
+    {
+      "epoch": 0.159632,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.471612903225807e-05,
+      "loss": 0.2148,
+      "step": 9977
+    },
+    {
+      "epoch": 0.159648,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.471451612903226e-05,
+      "loss": 0.1761,
+      "step": 9978
+    },
+    {
+      "epoch": 0.159664,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.471290322580646e-05,
+      "loss": 0.1866,
+      "step": 9979
+    },
+    {
+      "epoch": 0.15968,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.471129032258066e-05,
+      "loss": 0.1909,
+      "step": 9980
+    },
+    {
+      "epoch": 0.159696,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.470967741935484e-05,
+      "loss": 0.1926,
+      "step": 9981
+    },
+    {
+      "epoch": 0.159712,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.470806451612904e-05,
+      "loss": 0.1859,
+      "step": 9982
+    },
+    {
+      "epoch": 0.159728,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.470645161290323e-05,
+      "loss": 0.1779,
+      "step": 9983
+    },
+    {
+      "epoch": 0.159744,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.470483870967743e-05,
+      "loss": 0.1924,
+      "step": 9984
+    },
+    {
+      "epoch": 0.15976,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.470322580645161e-05,
+      "loss": 0.1887,
+      "step": 9985
+    },
+    {
+      "epoch": 0.159776,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.470161290322581e-05,
+      "loss": 0.1576,
+      "step": 9986
+    },
+    {
+      "epoch": 0.159792,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.47e-05,
+      "loss": 0.1616,
+      "step": 9987
+    },
+    {
+      "epoch": 0.159808,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.46983870967742e-05,
+      "loss": 0.1961,
+      "step": 9988
+    },
+    {
+      "epoch": 0.159824,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.469677419354838e-05,
+      "loss": 0.1271,
+      "step": 9989
+    },
+    {
+      "epoch": 0.15984,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.469516129032258e-05,
+      "loss": 0.2011,
+      "step": 9990
+    },
+    {
+      "epoch": 0.159856,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.469354838709678e-05,
+      "loss": 0.1812,
+      "step": 9991
+    },
+    {
+      "epoch": 0.159872,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.469193548387098e-05,
+      "loss": 0.1784,
+      "step": 9992
+    },
+    {
+      "epoch": 0.159888,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.469032258064517e-05,
+      "loss": 0.1955,
+      "step": 9993
+    },
+    {
+      "epoch": 0.159904,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.468870967741936e-05,
+      "loss": 0.1454,
+      "step": 9994
+    },
+    {
+      "epoch": 0.15992,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.468709677419356e-05,
+      "loss": 0.1489,
+      "step": 9995
+    },
+    {
+      "epoch": 0.159936,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.468548387096774e-05,
+      "loss": 0.1969,
+      "step": 9996
+    },
+    {
+      "epoch": 0.159952,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.468387096774194e-05,
+      "loss": 0.18,
+      "step": 9997
+    },
+    {
+      "epoch": 0.159968,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.468225806451613e-05,
+      "loss": 0.1922,
+      "step": 9998
+    },
+    {
+      "epoch": 0.159984,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.468064516129033e-05,
+      "loss": 0.1384,
+      "step": 9999
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.467903225806451e-05,
+      "loss": 0.1513,
+      "step": 10000
+    },
+    {
+      "epoch": 0.160016,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.467741935483871e-05,
+      "loss": 0.2192,
+      "step": 10001
+    },
+    {
+      "epoch": 0.160032,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.46758064516129e-05,
+      "loss": 0.1666,
+      "step": 10002
+    },
+    {
+      "epoch": 0.160048,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.46741935483871e-05,
+      "loss": 0.163,
+      "step": 10003
+    },
+    {
+      "epoch": 0.160064,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.46725806451613e-05,
+      "loss": 0.1745,
+      "step": 10004
+    },
+    {
+      "epoch": 0.16008,
+      "grad_norm": 1.25,
+      "learning_rate": 8.46709677419355e-05,
+      "loss": 0.1664,
+      "step": 10005
+    },
+    {
+      "epoch": 0.160096,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.466935483870968e-05,
+      "loss": 0.1792,
+      "step": 10006
+    },
+    {
+      "epoch": 0.160112,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.466774193548388e-05,
+      "loss": 0.1854,
+      "step": 10007
+    },
+    {
+      "epoch": 0.160128,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.466612903225807e-05,
+      "loss": 0.224,
+      "step": 10008
+    },
+    {
+      "epoch": 0.160144,
+      "grad_norm": 0.875,
+      "learning_rate": 8.466451612903227e-05,
+      "loss": 0.1796,
+      "step": 10009
+    },
+    {
+      "epoch": 0.16016,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.466290322580645e-05,
+      "loss": 0.1587,
+      "step": 10010
+    },
+    {
+      "epoch": 0.160176,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.466129032258064e-05,
+      "loss": 0.1471,
+      "step": 10011
+    },
+    {
+      "epoch": 0.160192,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.465967741935484e-05,
+      "loss": 0.1549,
+      "step": 10012
+    },
+    {
+      "epoch": 0.160208,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.465806451612903e-05,
+      "loss": 0.15,
+      "step": 10013
+    },
+    {
+      "epoch": 0.160224,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.465645161290323e-05,
+      "loss": 0.1419,
+      "step": 10014
+    },
+    {
+      "epoch": 0.16024,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.465483870967743e-05,
+      "loss": 0.1734,
+      "step": 10015
+    },
+    {
+      "epoch": 0.160256,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.465322580645163e-05,
+      "loss": 0.2227,
+      "step": 10016
+    },
+    {
+      "epoch": 0.160272,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.465161290322581e-05,
+      "loss": 0.168,
+      "step": 10017
+    },
+    {
+      "epoch": 0.160288,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.465000000000001e-05,
+      "loss": 0.1786,
+      "step": 10018
+    },
+    {
+      "epoch": 0.160304,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.46483870967742e-05,
+      "loss": 0.1992,
+      "step": 10019
+    },
+    {
+      "epoch": 0.16032,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.46467741935484e-05,
+      "loss": 0.1346,
+      "step": 10020
+    },
+    {
+      "epoch": 0.160336,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.464516129032258e-05,
+      "loss": 0.1665,
+      "step": 10021
+    },
+    {
+      "epoch": 0.160352,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.464354838709678e-05,
+      "loss": 0.1648,
+      "step": 10022
+    },
+    {
+      "epoch": 0.160368,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.464193548387097e-05,
+      "loss": 0.1606,
+      "step": 10023
+    },
+    {
+      "epoch": 0.160384,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.464032258064517e-05,
+      "loss": 0.1671,
+      "step": 10024
+    },
+    {
+      "epoch": 0.1604,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.463870967741935e-05,
+      "loss": 0.1958,
+      "step": 10025
+    },
+    {
+      "epoch": 0.160416,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.463709677419355e-05,
+      "loss": 0.1497,
+      "step": 10026
+    },
+    {
+      "epoch": 0.160432,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.463548387096774e-05,
+      "loss": 0.1521,
+      "step": 10027
+    },
+    {
+      "epoch": 0.160448,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.463387096774194e-05,
+      "loss": 0.1691,
+      "step": 10028
+    },
+    {
+      "epoch": 0.160464,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.463225806451614e-05,
+      "loss": 0.2089,
+      "step": 10029
+    },
+    {
+      "epoch": 0.16048,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.463064516129033e-05,
+      "loss": 0.1751,
+      "step": 10030
+    },
+    {
+      "epoch": 0.160496,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.462903225806453e-05,
+      "loss": 0.176,
+      "step": 10031
+    },
+    {
+      "epoch": 0.160512,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.462741935483871e-05,
+      "loss": 0.1556,
+      "step": 10032
+    },
+    {
+      "epoch": 0.160528,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.462580645161291e-05,
+      "loss": 0.2183,
+      "step": 10033
+    },
+    {
+      "epoch": 0.160544,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.46241935483871e-05,
+      "loss": 0.1296,
+      "step": 10034
+    },
+    {
+      "epoch": 0.16056,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.46225806451613e-05,
+      "loss": 0.1619,
+      "step": 10035
+    },
+    {
+      "epoch": 0.160576,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.462096774193548e-05,
+      "loss": 0.1927,
+      "step": 10036
+    },
+    {
+      "epoch": 0.160592,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.461935483870968e-05,
+      "loss": 0.1181,
+      "step": 10037
+    },
+    {
+      "epoch": 0.160608,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.461774193548387e-05,
+      "loss": 0.1768,
+      "step": 10038
+    },
+    {
+      "epoch": 0.160624,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.461612903225807e-05,
+      "loss": 0.1829,
+      "step": 10039
+    },
+    {
+      "epoch": 0.16064,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.461451612903227e-05,
+      "loss": 0.1679,
+      "step": 10040
+    },
+    {
+      "epoch": 0.160656,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.461290322580645e-05,
+      "loss": 0.1607,
+      "step": 10041
+    },
+    {
+      "epoch": 0.160672,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.461129032258065e-05,
+      "loss": 0.1901,
+      "step": 10042
+    },
+    {
+      "epoch": 0.160688,
+      "grad_norm": 1.3671875,
+      "learning_rate": 8.460967741935484e-05,
+      "loss": 0.1857,
+      "step": 10043
+    },
+    {
+      "epoch": 0.160704,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.460806451612904e-05,
+      "loss": 0.1899,
+      "step": 10044
+    },
+    {
+      "epoch": 0.16072,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.460645161290322e-05,
+      "loss": 0.183,
+      "step": 10045
+    },
+    {
+      "epoch": 0.160736,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.460483870967742e-05,
+      "loss": 0.1737,
+      "step": 10046
+    },
+    {
+      "epoch": 0.160752,
+      "grad_norm": 1.484375,
+      "learning_rate": 8.460322580645161e-05,
+      "loss": 0.1793,
+      "step": 10047
+    },
+    {
+      "epoch": 0.160768,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.460161290322581e-05,
+      "loss": 0.2115,
+      "step": 10048
+    },
+    {
+      "epoch": 0.160784,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.46e-05,
+      "loss": 0.2055,
+      "step": 10049
+    },
+    {
+      "epoch": 0.1608,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.45983870967742e-05,
+      "loss": 0.2125,
+      "step": 10050
+    },
+    {
+      "epoch": 0.160816,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.45967741935484e-05,
+      "loss": 0.2065,
+      "step": 10051
+    },
+    {
+      "epoch": 0.160832,
+      "grad_norm": 1.234375,
+      "learning_rate": 8.45951612903226e-05,
+      "loss": 0.154,
+      "step": 10052
+    },
+    {
+      "epoch": 0.160848,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.459354838709678e-05,
+      "loss": 0.1707,
+      "step": 10053
+    },
+    {
+      "epoch": 0.160864,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.459193548387098e-05,
+      "loss": 0.1624,
+      "step": 10054
+    },
+    {
+      "epoch": 0.16088,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.459032258064517e-05,
+      "loss": 0.1874,
+      "step": 10055
+    },
+    {
+      "epoch": 0.160896,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.458870967741937e-05,
+      "loss": 0.1506,
+      "step": 10056
+    },
+    {
+      "epoch": 0.160912,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.458709677419355e-05,
+      "loss": 0.1647,
+      "step": 10057
+    },
+    {
+      "epoch": 0.160928,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.458548387096774e-05,
+      "loss": 0.1818,
+      "step": 10058
+    },
+    {
+      "epoch": 0.160944,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.458387096774194e-05,
+      "loss": 0.1777,
+      "step": 10059
+    },
+    {
+      "epoch": 0.16096,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.458225806451612e-05,
+      "loss": 0.1494,
+      "step": 10060
+    },
+    {
+      "epoch": 0.160976,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.458064516129032e-05,
+      "loss": 0.1486,
+      "step": 10061
+    },
+    {
+      "epoch": 0.160992,
+      "grad_norm": 0.5,
+      "learning_rate": 8.457903225806451e-05,
+      "loss": 0.1359,
+      "step": 10062
+    },
+    {
+      "epoch": 0.161008,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.457741935483871e-05,
+      "loss": 0.1714,
+      "step": 10063
+    },
+    {
+      "epoch": 0.161024,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.457580645161291e-05,
+      "loss": 0.1499,
+      "step": 10064
+    },
+    {
+      "epoch": 0.16104,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.457419354838711e-05,
+      "loss": 0.164,
+      "step": 10065
+    },
+    {
+      "epoch": 0.161056,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.45725806451613e-05,
+      "loss": 0.1674,
+      "step": 10066
+    },
+    {
+      "epoch": 0.161072,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.45709677419355e-05,
+      "loss": 0.2042,
+      "step": 10067
+    },
+    {
+      "epoch": 0.161088,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.456935483870968e-05,
+      "loss": 0.1662,
+      "step": 10068
+    },
+    {
+      "epoch": 0.161104,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.456774193548388e-05,
+      "loss": 0.134,
+      "step": 10069
+    },
+    {
+      "epoch": 0.16112,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.456612903225807e-05,
+      "loss": 0.171,
+      "step": 10070
+    },
+    {
+      "epoch": 0.161136,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.456451612903227e-05,
+      "loss": 0.1659,
+      "step": 10071
+    },
+    {
+      "epoch": 0.161152,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.456290322580645e-05,
+      "loss": 0.1536,
+      "step": 10072
+    },
+    {
+      "epoch": 0.161168,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.456129032258064e-05,
+      "loss": 0.1904,
+      "step": 10073
+    },
+    {
+      "epoch": 0.161184,
+      "grad_norm": 0.46484375,
+      "learning_rate": 8.455967741935484e-05,
+      "loss": 0.1321,
+      "step": 10074
+    },
+    {
+      "epoch": 0.1612,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.455806451612904e-05,
+      "loss": 0.1772,
+      "step": 10075
+    },
+    {
+      "epoch": 0.161216,
+      "grad_norm": 1.3046875,
+      "learning_rate": 8.455645161290324e-05,
+      "loss": 0.2016,
+      "step": 10076
+    },
+    {
+      "epoch": 0.161232,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.455483870967742e-05,
+      "loss": 0.1577,
+      "step": 10077
+    },
+    {
+      "epoch": 0.161248,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.455322580645162e-05,
+      "loss": 0.195,
+      "step": 10078
+    },
+    {
+      "epoch": 0.161264,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.455161290322581e-05,
+      "loss": 0.1289,
+      "step": 10079
+    },
+    {
+      "epoch": 0.16128,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.455000000000001e-05,
+      "loss": 0.2014,
+      "step": 10080
+    },
+    {
+      "epoch": 0.161296,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.45483870967742e-05,
+      "loss": 0.1599,
+      "step": 10081
+    },
+    {
+      "epoch": 0.161312,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.45467741935484e-05,
+      "loss": 0.1417,
+      "step": 10082
+    },
+    {
+      "epoch": 0.161328,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.454516129032258e-05,
+      "loss": 0.1499,
+      "step": 10083
+    },
+    {
+      "epoch": 0.161344,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.454354838709678e-05,
+      "loss": 0.1875,
+      "step": 10084
+    },
+    {
+      "epoch": 0.16136,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.454193548387097e-05,
+      "loss": 0.2064,
+      "step": 10085
+    },
+    {
+      "epoch": 0.161376,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.454032258064517e-05,
+      "loss": 0.1932,
+      "step": 10086
+    },
+    {
+      "epoch": 0.161392,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.453870967741937e-05,
+      "loss": 0.1526,
+      "step": 10087
+    },
+    {
+      "epoch": 0.161408,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.453709677419355e-05,
+      "loss": 0.1532,
+      "step": 10088
+    },
+    {
+      "epoch": 0.161424,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.453548387096775e-05,
+      "loss": 0.1597,
+      "step": 10089
+    },
+    {
+      "epoch": 0.16144,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.453387096774194e-05,
+      "loss": 0.1911,
+      "step": 10090
+    },
+    {
+      "epoch": 0.161456,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.453225806451614e-05,
+      "loss": 0.1732,
+      "step": 10091
+    },
+    {
+      "epoch": 0.161472,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.453064516129032e-05,
+      "loss": 0.1528,
+      "step": 10092
+    },
+    {
+      "epoch": 0.161488,
+      "grad_norm": 0.474609375,
+      "learning_rate": 8.452903225806452e-05,
+      "loss": 0.18,
+      "step": 10093
+    },
+    {
+      "epoch": 0.161504,
+      "grad_norm": 0.75,
+      "learning_rate": 8.452741935483871e-05,
+      "loss": 0.1835,
+      "step": 10094
+    },
+    {
+      "epoch": 0.16152,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.452580645161291e-05,
+      "loss": 0.1713,
+      "step": 10095
+    },
+    {
+      "epoch": 0.161536,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.45241935483871e-05,
+      "loss": 0.1692,
+      "step": 10096
+    },
+    {
+      "epoch": 0.161552,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.45225806451613e-05,
+      "loss": 0.1703,
+      "step": 10097
+    },
+    {
+      "epoch": 0.161568,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.452096774193548e-05,
+      "loss": 0.1936,
+      "step": 10098
+    },
+    {
+      "epoch": 0.161584,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.451935483870968e-05,
+      "loss": 0.1532,
+      "step": 10099
+    },
+    {
+      "epoch": 0.1616,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.451774193548388e-05,
+      "loss": 0.1697,
+      "step": 10100
+    },
+    {
+      "epoch": 0.161616,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.451612903225808e-05,
+      "loss": 0.1527,
+      "step": 10101
+    },
+    {
+      "epoch": 0.161632,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.451451612903227e-05,
+      "loss": 0.1527,
+      "step": 10102
+    },
+    {
+      "epoch": 0.161648,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.451290322580645e-05,
+      "loss": 0.1595,
+      "step": 10103
+    },
+    {
+      "epoch": 0.161664,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.451129032258065e-05,
+      "loss": 0.1767,
+      "step": 10104
+    },
+    {
+      "epoch": 0.16168,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.450967741935484e-05,
+      "loss": 0.1397,
+      "step": 10105
+    },
+    {
+      "epoch": 0.161696,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.450806451612904e-05,
+      "loss": 0.1543,
+      "step": 10106
+    },
+    {
+      "epoch": 0.161712,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.450645161290322e-05,
+      "loss": 0.1937,
+      "step": 10107
+    },
+    {
+      "epoch": 0.161728,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.450483870967742e-05,
+      "loss": 0.1764,
+      "step": 10108
+    },
+    {
+      "epoch": 0.161744,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.450322580645161e-05,
+      "loss": 0.2017,
+      "step": 10109
+    },
+    {
+      "epoch": 0.16176,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.450161290322581e-05,
+      "loss": 0.1655,
+      "step": 10110
+    },
+    {
+      "epoch": 0.161776,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.450000000000001e-05,
+      "loss": 0.182,
+      "step": 10111
+    },
+    {
+      "epoch": 0.161792,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.449838709677421e-05,
+      "loss": 0.1916,
+      "step": 10112
+    },
+    {
+      "epoch": 0.161808,
+      "grad_norm": 0.75,
+      "learning_rate": 8.449677419354839e-05,
+      "loss": 0.1465,
+      "step": 10113
+    },
+    {
+      "epoch": 0.161824,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.449516129032259e-05,
+      "loss": 0.1734,
+      "step": 10114
+    },
+    {
+      "epoch": 0.16184,
+      "grad_norm": 1.265625,
+      "learning_rate": 8.449354838709678e-05,
+      "loss": 0.1759,
+      "step": 10115
+    },
+    {
+      "epoch": 0.161856,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.449193548387098e-05,
+      "loss": 0.1771,
+      "step": 10116
+    },
+    {
+      "epoch": 0.161872,
+      "grad_norm": 1.3359375,
+      "learning_rate": 8.449032258064516e-05,
+      "loss": 0.1826,
+      "step": 10117
+    },
+    {
+      "epoch": 0.161888,
+      "grad_norm": 0.625,
+      "learning_rate": 8.448870967741936e-05,
+      "loss": 0.1442,
+      "step": 10118
+    },
+    {
+      "epoch": 0.161904,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.448709677419355e-05,
+      "loss": 0.1747,
+      "step": 10119
+    },
+    {
+      "epoch": 0.16192,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.448548387096774e-05,
+      "loss": 0.1598,
+      "step": 10120
+    },
+    {
+      "epoch": 0.161936,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.448387096774194e-05,
+      "loss": 0.1501,
+      "step": 10121
+    },
+    {
+      "epoch": 0.161952,
+      "grad_norm": 0.75,
+      "learning_rate": 8.448225806451612e-05,
+      "loss": 0.1429,
+      "step": 10122
+    },
+    {
+      "epoch": 0.161968,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.448064516129032e-05,
+      "loss": 0.1655,
+      "step": 10123
+    },
+    {
+      "epoch": 0.161984,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.447903225806452e-05,
+      "loss": 0.1579,
+      "step": 10124
+    },
+    {
+      "epoch": 0.162,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.447741935483872e-05,
+      "loss": 0.151,
+      "step": 10125
+    },
+    {
+      "epoch": 0.162016,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.447580645161291e-05,
+      "loss": 0.1578,
+      "step": 10126
+    },
+    {
+      "epoch": 0.162032,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.44741935483871e-05,
+      "loss": 0.1643,
+      "step": 10127
+    },
+    {
+      "epoch": 0.162048,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.447258064516129e-05,
+      "loss": 0.1963,
+      "step": 10128
+    },
+    {
+      "epoch": 0.162064,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.447096774193549e-05,
+      "loss": 0.1876,
+      "step": 10129
+    },
+    {
+      "epoch": 0.16208,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.446935483870968e-05,
+      "loss": 0.1775,
+      "step": 10130
+    },
+    {
+      "epoch": 0.162096,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.446774193548388e-05,
+      "loss": 0.1942,
+      "step": 10131
+    },
+    {
+      "epoch": 0.162112,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.446612903225806e-05,
+      "loss": 0.1854,
+      "step": 10132
+    },
+    {
+      "epoch": 0.162128,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.446451612903226e-05,
+      "loss": 0.183,
+      "step": 10133
+    },
+    {
+      "epoch": 0.162144,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.446290322580645e-05,
+      "loss": 0.1744,
+      "step": 10134
+    },
+    {
+      "epoch": 0.16216,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.446129032258065e-05,
+      "loss": 0.1806,
+      "step": 10135
+    },
+    {
+      "epoch": 0.162176,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.445967741935485e-05,
+      "loss": 0.1368,
+      "step": 10136
+    },
+    {
+      "epoch": 0.162192,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.445806451612904e-05,
+      "loss": 0.1804,
+      "step": 10137
+    },
+    {
+      "epoch": 0.162208,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.445645161290323e-05,
+      "loss": 0.1556,
+      "step": 10138
+    },
+    {
+      "epoch": 0.162224,
+      "grad_norm": 1.2109375,
+      "learning_rate": 8.445483870967742e-05,
+      "loss": 0.2275,
+      "step": 10139
+    },
+    {
+      "epoch": 0.16224,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.445322580645162e-05,
+      "loss": 0.1994,
+      "step": 10140
+    },
+    {
+      "epoch": 0.162256,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.44516129032258e-05,
+      "loss": 0.1819,
+      "step": 10141
+    },
+    {
+      "epoch": 0.162272,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.445e-05,
+      "loss": 0.1631,
+      "step": 10142
+    },
+    {
+      "epoch": 0.162288,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.444838709677419e-05,
+      "loss": 0.1804,
+      "step": 10143
+    },
+    {
+      "epoch": 0.162304,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.444677419354839e-05,
+      "loss": 0.1413,
+      "step": 10144
+    },
+    {
+      "epoch": 0.16232,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.444516129032258e-05,
+      "loss": 0.1744,
+      "step": 10145
+    },
+    {
+      "epoch": 0.162336,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.444354838709678e-05,
+      "loss": 0.1668,
+      "step": 10146
+    },
+    {
+      "epoch": 0.162352,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.444193548387098e-05,
+      "loss": 0.1537,
+      "step": 10147
+    },
+    {
+      "epoch": 0.162368,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.444032258064518e-05,
+      "loss": 0.1829,
+      "step": 10148
+    },
+    {
+      "epoch": 0.162384,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.443870967741936e-05,
+      "loss": 0.1815,
+      "step": 10149
+    },
+    {
+      "epoch": 0.1624,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.443709677419355e-05,
+      "loss": 0.1417,
+      "step": 10150
+    },
+    {
+      "epoch": 0.162416,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.443548387096775e-05,
+      "loss": 0.1884,
+      "step": 10151
+    },
+    {
+      "epoch": 0.162432,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.443387096774193e-05,
+      "loss": 0.1768,
+      "step": 10152
+    },
+    {
+      "epoch": 0.162448,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.443225806451613e-05,
+      "loss": 0.2014,
+      "step": 10153
+    },
+    {
+      "epoch": 0.162464,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.443064516129032e-05,
+      "loss": 0.1633,
+      "step": 10154
+    },
+    {
+      "epoch": 0.16248,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.442903225806452e-05,
+      "loss": 0.1764,
+      "step": 10155
+    },
+    {
+      "epoch": 0.162496,
+      "grad_norm": 1.2578125,
+      "learning_rate": 8.44274193548387e-05,
+      "loss": 0.1967,
+      "step": 10156
+    },
+    {
+      "epoch": 0.162512,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.44258064516129e-05,
+      "loss": 0.1696,
+      "step": 10157
+    },
+    {
+      "epoch": 0.162528,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.442419354838709e-05,
+      "loss": 0.1765,
+      "step": 10158
+    },
+    {
+      "epoch": 0.162544,
+      "grad_norm": 1.359375,
+      "learning_rate": 8.442258064516129e-05,
+      "loss": 0.1848,
+      "step": 10159
+    },
+    {
+      "epoch": 0.16256,
+      "grad_norm": 1.390625,
+      "learning_rate": 8.442096774193549e-05,
+      "loss": 0.1661,
+      "step": 10160
+    },
+    {
+      "epoch": 0.162576,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.441935483870969e-05,
+      "loss": 0.1486,
+      "step": 10161
+    },
+    {
+      "epoch": 0.162592,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.441774193548388e-05,
+      "loss": 0.1395,
+      "step": 10162
+    },
+    {
+      "epoch": 0.162608,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.441612903225808e-05,
+      "loss": 0.1802,
+      "step": 10163
+    },
+    {
+      "epoch": 0.162624,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.441451612903226e-05,
+      "loss": 0.1753,
+      "step": 10164
+    },
+    {
+      "epoch": 0.16264,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.441290322580646e-05,
+      "loss": 0.1686,
+      "step": 10165
+    },
+    {
+      "epoch": 0.162656,
+      "grad_norm": 1.46875,
+      "learning_rate": 8.441129032258065e-05,
+      "loss": 0.1958,
+      "step": 10166
+    },
+    {
+      "epoch": 0.162672,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.440967741935483e-05,
+      "loss": 0.1903,
+      "step": 10167
+    },
+    {
+      "epoch": 0.162688,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.440806451612903e-05,
+      "loss": 0.1802,
+      "step": 10168
+    },
+    {
+      "epoch": 0.162704,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.440645161290322e-05,
+      "loss": 0.191,
+      "step": 10169
+    },
+    {
+      "epoch": 0.16272,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.440483870967742e-05,
+      "loss": 0.1558,
+      "step": 10170
+    },
+    {
+      "epoch": 0.162736,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.440322580645162e-05,
+      "loss": 0.18,
+      "step": 10171
+    },
+    {
+      "epoch": 0.162752,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.440161290322582e-05,
+      "loss": 0.1982,
+      "step": 10172
+    },
+    {
+      "epoch": 0.162768,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.44e-05,
+      "loss": 0.1273,
+      "step": 10173
+    },
+    {
+      "epoch": 0.162784,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.43983870967742e-05,
+      "loss": 0.2195,
+      "step": 10174
+    },
+    {
+      "epoch": 0.1628,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.439677419354839e-05,
+      "loss": 0.145,
+      "step": 10175
+    },
+    {
+      "epoch": 0.162816,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.439516129032259e-05,
+      "loss": 0.2055,
+      "step": 10176
+    },
+    {
+      "epoch": 0.162832,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.439354838709678e-05,
+      "loss": 0.1698,
+      "step": 10177
+    },
+    {
+      "epoch": 0.162848,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.439193548387098e-05,
+      "loss": 0.1714,
+      "step": 10178
+    },
+    {
+      "epoch": 0.162864,
+      "grad_norm": 1.515625,
+      "learning_rate": 8.439032258064516e-05,
+      "loss": 0.1661,
+      "step": 10179
+    },
+    {
+      "epoch": 0.16288,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.438870967741936e-05,
+      "loss": 0.1824,
+      "step": 10180
+    },
+    {
+      "epoch": 0.162896,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.438709677419355e-05,
+      "loss": 0.202,
+      "step": 10181
+    },
+    {
+      "epoch": 0.162912,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.438548387096775e-05,
+      "loss": 0.179,
+      "step": 10182
+    },
+    {
+      "epoch": 0.162928,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.438387096774193e-05,
+      "loss": 0.1602,
+      "step": 10183
+    },
+    {
+      "epoch": 0.162944,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.438225806451613e-05,
+      "loss": 0.2001,
+      "step": 10184
+    },
+    {
+      "epoch": 0.16296,
+      "grad_norm": 1.296875,
+      "learning_rate": 8.438064516129033e-05,
+      "loss": 0.1438,
+      "step": 10185
+    },
+    {
+      "epoch": 0.162976,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.437903225806452e-05,
+      "loss": 0.1438,
+      "step": 10186
+    },
+    {
+      "epoch": 0.162992,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.437741935483872e-05,
+      "loss": 0.1651,
+      "step": 10187
+    },
+    {
+      "epoch": 0.163008,
+      "grad_norm": 0.875,
+      "learning_rate": 8.43758064516129e-05,
+      "loss": 0.1449,
+      "step": 10188
+    },
+    {
+      "epoch": 0.163024,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.43741935483871e-05,
+      "loss": 0.1365,
+      "step": 10189
+    },
+    {
+      "epoch": 0.16304,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.437258064516129e-05,
+      "loss": 0.222,
+      "step": 10190
+    },
+    {
+      "epoch": 0.163056,
+      "grad_norm": 0.875,
+      "learning_rate": 8.437096774193549e-05,
+      "loss": 0.1644,
+      "step": 10191
+    },
+    {
+      "epoch": 0.163072,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.436935483870968e-05,
+      "loss": 0.1729,
+      "step": 10192
+    },
+    {
+      "epoch": 0.163088,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.436774193548388e-05,
+      "loss": 0.1561,
+      "step": 10193
+    },
+    {
+      "epoch": 0.163104,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.436612903225806e-05,
+      "loss": 0.1448,
+      "step": 10194
+    },
+    {
+      "epoch": 0.16312,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.436451612903226e-05,
+      "loss": 0.1918,
+      "step": 10195
+    },
+    {
+      "epoch": 0.163136,
+      "grad_norm": 1.2734375,
+      "learning_rate": 8.436290322580646e-05,
+      "loss": 0.2224,
+      "step": 10196
+    },
+    {
+      "epoch": 0.163152,
+      "grad_norm": 0.625,
+      "learning_rate": 8.436129032258065e-05,
+      "loss": 0.1679,
+      "step": 10197
+    },
+    {
+      "epoch": 0.163168,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.435967741935485e-05,
+      "loss": 0.1771,
+      "step": 10198
+    },
+    {
+      "epoch": 0.163184,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.435806451612903e-05,
+      "loss": 0.169,
+      "step": 10199
+    },
+    {
+      "epoch": 0.1632,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.435645161290323e-05,
+      "loss": 0.1656,
+      "step": 10200
+    },
+    {
+      "epoch": 0.163216,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.435483870967742e-05,
+      "loss": 0.1557,
+      "step": 10201
+    },
+    {
+      "epoch": 0.163232,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.435322580645162e-05,
+      "loss": 0.1831,
+      "step": 10202
+    },
+    {
+      "epoch": 0.163248,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.43516129032258e-05,
+      "loss": 0.1963,
+      "step": 10203
+    },
+    {
+      "epoch": 0.163264,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.435e-05,
+      "loss": 0.1927,
+      "step": 10204
+    },
+    {
+      "epoch": 0.16328,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.434838709677419e-05,
+      "loss": 0.223,
+      "step": 10205
+    },
+    {
+      "epoch": 0.163296,
+      "grad_norm": 0.75,
+      "learning_rate": 8.434677419354839e-05,
+      "loss": 0.1966,
+      "step": 10206
+    },
+    {
+      "epoch": 0.163312,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.434516129032259e-05,
+      "loss": 0.167,
+      "step": 10207
+    },
+    {
+      "epoch": 0.163328,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.434354838709679e-05,
+      "loss": 0.1664,
+      "step": 10208
+    },
+    {
+      "epoch": 0.163344,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.434193548387097e-05,
+      "loss": 0.1589,
+      "step": 10209
+    },
+    {
+      "epoch": 0.16336,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.434032258064517e-05,
+      "loss": 0.1901,
+      "step": 10210
+    },
+    {
+      "epoch": 0.163376,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.433870967741936e-05,
+      "loss": 0.155,
+      "step": 10211
+    },
+    {
+      "epoch": 0.163392,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.433709677419355e-05,
+      "loss": 0.2178,
+      "step": 10212
+    },
+    {
+      "epoch": 0.163408,
+      "grad_norm": 1.359375,
+      "learning_rate": 8.433548387096775e-05,
+      "loss": 0.1984,
+      "step": 10213
+    },
+    {
+      "epoch": 0.163424,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.433387096774193e-05,
+      "loss": 0.167,
+      "step": 10214
+    },
+    {
+      "epoch": 0.16344,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.433225806451613e-05,
+      "loss": 0.1796,
+      "step": 10215
+    },
+    {
+      "epoch": 0.163456,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.433064516129032e-05,
+      "loss": 0.2058,
+      "step": 10216
+    },
+    {
+      "epoch": 0.163472,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.432903225806452e-05,
+      "loss": 0.1945,
+      "step": 10217
+    },
+    {
+      "epoch": 0.163488,
+      "grad_norm": 1.90625,
+      "learning_rate": 8.43274193548387e-05,
+      "loss": 0.1981,
+      "step": 10218
+    },
+    {
+      "epoch": 0.163504,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.43258064516129e-05,
+      "loss": 0.1985,
+      "step": 10219
+    },
+    {
+      "epoch": 0.16352,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.43241935483871e-05,
+      "loss": 0.1574,
+      "step": 10220
+    },
+    {
+      "epoch": 0.163536,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.43225806451613e-05,
+      "loss": 0.1565,
+      "step": 10221
+    },
+    {
+      "epoch": 0.163552,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.432096774193549e-05,
+      "loss": 0.155,
+      "step": 10222
+    },
+    {
+      "epoch": 0.163568,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.431935483870969e-05,
+      "loss": 0.1854,
+      "step": 10223
+    },
+    {
+      "epoch": 0.163584,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.431774193548387e-05,
+      "loss": 0.156,
+      "step": 10224
+    },
+    {
+      "epoch": 0.1636,
+      "grad_norm": 0.75,
+      "learning_rate": 8.431612903225807e-05,
+      "loss": 0.177,
+      "step": 10225
+    },
+    {
+      "epoch": 0.163616,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.431451612903226e-05,
+      "loss": 0.1848,
+      "step": 10226
+    },
+    {
+      "epoch": 0.163632,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.431290322580646e-05,
+      "loss": 0.1269,
+      "step": 10227
+    },
+    {
+      "epoch": 0.163648,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.431129032258065e-05,
+      "loss": 0.1797,
+      "step": 10228
+    },
+    {
+      "epoch": 0.163664,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.430967741935483e-05,
+      "loss": 0.1796,
+      "step": 10229
+    },
+    {
+      "epoch": 0.16368,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.430806451612903e-05,
+      "loss": 0.1789,
+      "step": 10230
+    },
+    {
+      "epoch": 0.163696,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.430645161290323e-05,
+      "loss": 0.1559,
+      "step": 10231
+    },
+    {
+      "epoch": 0.163712,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.430483870967743e-05,
+      "loss": 0.204,
+      "step": 10232
+    },
+    {
+      "epoch": 0.163728,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.430322580645162e-05,
+      "loss": 0.1949,
+      "step": 10233
+    },
+    {
+      "epoch": 0.163744,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.430161290322582e-05,
+      "loss": 0.2028,
+      "step": 10234
+    },
+    {
+      "epoch": 0.16376,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.43e-05,
+      "loss": 0.1869,
+      "step": 10235
+    },
+    {
+      "epoch": 0.163776,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.42983870967742e-05,
+      "loss": 0.1779,
+      "step": 10236
+    },
+    {
+      "epoch": 0.163792,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.429677419354839e-05,
+      "loss": 0.1647,
+      "step": 10237
+    },
+    {
+      "epoch": 0.163808,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.429516129032259e-05,
+      "loss": 0.1699,
+      "step": 10238
+    },
+    {
+      "epoch": 0.163824,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.429354838709677e-05,
+      "loss": 0.1412,
+      "step": 10239
+    },
+    {
+      "epoch": 0.16384,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.429193548387097e-05,
+      "loss": 0.1529,
+      "step": 10240
+    },
+    {
+      "epoch": 0.163856,
+      "grad_norm": 1.6640625,
+      "learning_rate": 8.429032258064516e-05,
+      "loss": 0.2218,
+      "step": 10241
+    },
+    {
+      "epoch": 0.163872,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.428870967741936e-05,
+      "loss": 0.1452,
+      "step": 10242
+    },
+    {
+      "epoch": 0.163888,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.428709677419356e-05,
+      "loss": 0.1153,
+      "step": 10243
+    },
+    {
+      "epoch": 0.163904,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.428548387096775e-05,
+      "loss": 0.1749,
+      "step": 10244
+    },
+    {
+      "epoch": 0.16392,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.428387096774194e-05,
+      "loss": 0.1845,
+      "step": 10245
+    },
+    {
+      "epoch": 0.163936,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.428225806451613e-05,
+      "loss": 0.1443,
+      "step": 10246
+    },
+    {
+      "epoch": 0.163952,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.428064516129033e-05,
+      "loss": 0.1848,
+      "step": 10247
+    },
+    {
+      "epoch": 0.163968,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.427903225806452e-05,
+      "loss": 0.1379,
+      "step": 10248
+    },
+    {
+      "epoch": 0.163984,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.427741935483872e-05,
+      "loss": 0.1768,
+      "step": 10249
+    },
+    {
+      "epoch": 0.164,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.42758064516129e-05,
+      "loss": 0.1748,
+      "step": 10250
+    },
+    {
+      "epoch": 0.164016,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.42741935483871e-05,
+      "loss": 0.1655,
+      "step": 10251
+    },
+    {
+      "epoch": 0.164032,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.427258064516129e-05,
+      "loss": 0.2116,
+      "step": 10252
+    },
+    {
+      "epoch": 0.164048,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.427096774193549e-05,
+      "loss": 0.1682,
+      "step": 10253
+    },
+    {
+      "epoch": 0.164064,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.426935483870967e-05,
+      "loss": 0.1591,
+      "step": 10254
+    },
+    {
+      "epoch": 0.16408,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.426774193548387e-05,
+      "loss": 0.1582,
+      "step": 10255
+    },
+    {
+      "epoch": 0.164096,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.426612903225807e-05,
+      "loss": 0.1573,
+      "step": 10256
+    },
+    {
+      "epoch": 0.164112,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.426451612903227e-05,
+      "loss": 0.1492,
+      "step": 10257
+    },
+    {
+      "epoch": 0.164128,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.426290322580646e-05,
+      "loss": 0.1717,
+      "step": 10258
+    },
+    {
+      "epoch": 0.164144,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.426129032258064e-05,
+      "loss": 0.1448,
+      "step": 10259
+    },
+    {
+      "epoch": 0.16416,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.425967741935484e-05,
+      "loss": 0.175,
+      "step": 10260
+    },
+    {
+      "epoch": 0.164176,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.425806451612903e-05,
+      "loss": 0.1602,
+      "step": 10261
+    },
+    {
+      "epoch": 0.164192,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.425645161290323e-05,
+      "loss": 0.1666,
+      "step": 10262
+    },
+    {
+      "epoch": 0.164208,
+      "grad_norm": 0.95703125,
+      "learning_rate": 8.425483870967742e-05,
+      "loss": 0.1557,
+      "step": 10263
+    },
+    {
+      "epoch": 0.164224,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.425322580645162e-05,
+      "loss": 0.1413,
+      "step": 10264
+    },
+    {
+      "epoch": 0.16424,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.42516129032258e-05,
+      "loss": 0.1962,
+      "step": 10265
+    },
+    {
+      "epoch": 0.164256,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.425e-05,
+      "loss": 0.1926,
+      "step": 10266
+    },
+    {
+      "epoch": 0.164272,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.42483870967742e-05,
+      "loss": 0.1802,
+      "step": 10267
+    },
+    {
+      "epoch": 0.164288,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.42467741935484e-05,
+      "loss": 0.178,
+      "step": 10268
+    },
+    {
+      "epoch": 0.164304,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.424516129032259e-05,
+      "loss": 0.1802,
+      "step": 10269
+    },
+    {
+      "epoch": 0.16432,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.424354838709679e-05,
+      "loss": 0.2014,
+      "step": 10270
+    },
+    {
+      "epoch": 0.164336,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.424193548387097e-05,
+      "loss": 0.2218,
+      "step": 10271
+    },
+    {
+      "epoch": 0.164352,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.424032258064517e-05,
+      "loss": 0.1948,
+      "step": 10272
+    },
+    {
+      "epoch": 0.164368,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.423870967741936e-05,
+      "loss": 0.1719,
+      "step": 10273
+    },
+    {
+      "epoch": 0.164384,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.423709677419356e-05,
+      "loss": 0.1755,
+      "step": 10274
+    },
+    {
+      "epoch": 0.1644,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.423548387096774e-05,
+      "loss": 0.1359,
+      "step": 10275
+    },
+    {
+      "epoch": 0.164416,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.423387096774193e-05,
+      "loss": 0.143,
+      "step": 10276
+    },
+    {
+      "epoch": 0.164432,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.423225806451613e-05,
+      "loss": 0.1464,
+      "step": 10277
+    },
+    {
+      "epoch": 0.164448,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.423064516129033e-05,
+      "loss": 0.185,
+      "step": 10278
+    },
+    {
+      "epoch": 0.164464,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.422903225806452e-05,
+      "loss": 0.1876,
+      "step": 10279
+    },
+    {
+      "epoch": 0.16448,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.422741935483871e-05,
+      "loss": 0.1526,
+      "step": 10280
+    },
+    {
+      "epoch": 0.164496,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.422580645161291e-05,
+      "loss": 0.1723,
+      "step": 10281
+    },
+    {
+      "epoch": 0.164512,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.42241935483871e-05,
+      "loss": 0.162,
+      "step": 10282
+    },
+    {
+      "epoch": 0.164528,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.42225806451613e-05,
+      "loss": 0.1549,
+      "step": 10283
+    },
+    {
+      "epoch": 0.164544,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.422096774193549e-05,
+      "loss": 0.1475,
+      "step": 10284
+    },
+    {
+      "epoch": 0.16456,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.421935483870969e-05,
+      "loss": 0.1381,
+      "step": 10285
+    },
+    {
+      "epoch": 0.164576,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.421774193548387e-05,
+      "loss": 0.1983,
+      "step": 10286
+    },
+    {
+      "epoch": 0.164592,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.421612903225807e-05,
+      "loss": 0.1898,
+      "step": 10287
+    },
+    {
+      "epoch": 0.164608,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.421451612903226e-05,
+      "loss": 0.1476,
+      "step": 10288
+    },
+    {
+      "epoch": 0.164624,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.421290322580646e-05,
+      "loss": 0.1574,
+      "step": 10289
+    },
+    {
+      "epoch": 0.16464,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.421129032258064e-05,
+      "loss": 0.1561,
+      "step": 10290
+    },
+    {
+      "epoch": 0.164656,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.420967741935484e-05,
+      "loss": 0.1735,
+      "step": 10291
+    },
+    {
+      "epoch": 0.164672,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.420806451612904e-05,
+      "loss": 0.2002,
+      "step": 10292
+    },
+    {
+      "epoch": 0.164688,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.420645161290323e-05,
+      "loss": 0.193,
+      "step": 10293
+    },
+    {
+      "epoch": 0.164704,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.420483870967743e-05,
+      "loss": 0.1849,
+      "step": 10294
+    },
+    {
+      "epoch": 0.16472,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.420322580645161e-05,
+      "loss": 0.2058,
+      "step": 10295
+    },
+    {
+      "epoch": 0.164736,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.420161290322581e-05,
+      "loss": 0.2284,
+      "step": 10296
+    },
+    {
+      "epoch": 0.164752,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.42e-05,
+      "loss": 0.1537,
+      "step": 10297
+    },
+    {
+      "epoch": 0.164768,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.41983870967742e-05,
+      "loss": 0.1174,
+      "step": 10298
+    },
+    {
+      "epoch": 0.164784,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.419677419354839e-05,
+      "loss": 0.1688,
+      "step": 10299
+    },
+    {
+      "epoch": 0.1648,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.419516129032259e-05,
+      "loss": 0.1495,
+      "step": 10300
+    },
+    {
+      "epoch": 0.164816,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.419354838709677e-05,
+      "loss": 0.1584,
+      "step": 10301
+    },
+    {
+      "epoch": 0.164832,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.419193548387097e-05,
+      "loss": 0.1767,
+      "step": 10302
+    },
+    {
+      "epoch": 0.164848,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.419032258064517e-05,
+      "loss": 0.2113,
+      "step": 10303
+    },
+    {
+      "epoch": 0.164864,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.418870967741937e-05,
+      "loss": 0.2141,
+      "step": 10304
+    },
+    {
+      "epoch": 0.16488,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.418709677419356e-05,
+      "loss": 0.204,
+      "step": 10305
+    },
+    {
+      "epoch": 0.164896,
+      "grad_norm": 1.3984375,
+      "learning_rate": 8.418548387096774e-05,
+      "loss": 0.1646,
+      "step": 10306
+    },
+    {
+      "epoch": 0.164912,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.418387096774194e-05,
+      "loss": 0.1911,
+      "step": 10307
+    },
+    {
+      "epoch": 0.164928,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.418225806451613e-05,
+      "loss": 0.1773,
+      "step": 10308
+    },
+    {
+      "epoch": 0.164944,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.418064516129033e-05,
+      "loss": 0.1238,
+      "step": 10309
+    },
+    {
+      "epoch": 0.16496,
+      "grad_norm": 1.34375,
+      "learning_rate": 8.417903225806451e-05,
+      "loss": 0.1783,
+      "step": 10310
+    },
+    {
+      "epoch": 0.164976,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.417741935483871e-05,
+      "loss": 0.1682,
+      "step": 10311
+    },
+    {
+      "epoch": 0.164992,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.41758064516129e-05,
+      "loss": 0.1522,
+      "step": 10312
+    },
+    {
+      "epoch": 0.165008,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.41741935483871e-05,
+      "loss": 0.1896,
+      "step": 10313
+    },
+    {
+      "epoch": 0.165024,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.417258064516129e-05,
+      "loss": 0.2244,
+      "step": 10314
+    },
+    {
+      "epoch": 0.16504,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.417096774193549e-05,
+      "loss": 0.1785,
+      "step": 10315
+    },
+    {
+      "epoch": 0.165056,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.416935483870968e-05,
+      "loss": 0.1776,
+      "step": 10316
+    },
+    {
+      "epoch": 0.165072,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.416774193548388e-05,
+      "loss": 0.1462,
+      "step": 10317
+    },
+    {
+      "epoch": 0.165088,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.416612903225807e-05,
+      "loss": 0.1647,
+      "step": 10318
+    },
+    {
+      "epoch": 0.165104,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.416451612903227e-05,
+      "loss": 0.1546,
+      "step": 10319
+    },
+    {
+      "epoch": 0.16512,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.416290322580646e-05,
+      "loss": 0.1561,
+      "step": 10320
+    },
+    {
+      "epoch": 0.165136,
+      "grad_norm": 0.49609375,
+      "learning_rate": 8.416129032258064e-05,
+      "loss": 0.127,
+      "step": 10321
+    },
+    {
+      "epoch": 0.165152,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.415967741935484e-05,
+      "loss": 0.1435,
+      "step": 10322
+    },
+    {
+      "epoch": 0.165168,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.415806451612903e-05,
+      "loss": 0.1674,
+      "step": 10323
+    },
+    {
+      "epoch": 0.165184,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.415645161290323e-05,
+      "loss": 0.2004,
+      "step": 10324
+    },
+    {
+      "epoch": 0.1652,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.415483870967741e-05,
+      "loss": 0.1781,
+      "step": 10325
+    },
+    {
+      "epoch": 0.165216,
+      "grad_norm": 1.25,
+      "learning_rate": 8.415322580645161e-05,
+      "loss": 0.1955,
+      "step": 10326
+    },
+    {
+      "epoch": 0.165232,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.415161290322581e-05,
+      "loss": 0.1946,
+      "step": 10327
+    },
+    {
+      "epoch": 0.165248,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.415000000000001e-05,
+      "loss": 0.1715,
+      "step": 10328
+    },
+    {
+      "epoch": 0.165264,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.41483870967742e-05,
+      "loss": 0.1755,
+      "step": 10329
+    },
+    {
+      "epoch": 0.16528,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.41467741935484e-05,
+      "loss": 0.1904,
+      "step": 10330
+    },
+    {
+      "epoch": 0.165296,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.414516129032258e-05,
+      "loss": 0.1505,
+      "step": 10331
+    },
+    {
+      "epoch": 0.165312,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.414354838709678e-05,
+      "loss": 0.1793,
+      "step": 10332
+    },
+    {
+      "epoch": 0.165328,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.414193548387097e-05,
+      "loss": 0.1447,
+      "step": 10333
+    },
+    {
+      "epoch": 0.165344,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.414032258064517e-05,
+      "loss": 0.1225,
+      "step": 10334
+    },
+    {
+      "epoch": 0.16536,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.413870967741936e-05,
+      "loss": 0.1992,
+      "step": 10335
+    },
+    {
+      "epoch": 0.165376,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.413709677419356e-05,
+      "loss": 0.1922,
+      "step": 10336
+    },
+    {
+      "epoch": 0.165392,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.413548387096774e-05,
+      "loss": 0.1687,
+      "step": 10337
+    },
+    {
+      "epoch": 0.165408,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.413387096774194e-05,
+      "loss": 0.1598,
+      "step": 10338
+    },
+    {
+      "epoch": 0.165424,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.413225806451614e-05,
+      "loss": 0.1707,
+      "step": 10339
+    },
+    {
+      "epoch": 0.16544,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.413064516129033e-05,
+      "loss": 0.172,
+      "step": 10340
+    },
+    {
+      "epoch": 0.165456,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.412903225806453e-05,
+      "loss": 0.1794,
+      "step": 10341
+    },
+    {
+      "epoch": 0.165472,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.412741935483871e-05,
+      "loss": 0.1658,
+      "step": 10342
+    },
+    {
+      "epoch": 0.165488,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.412580645161291e-05,
+      "loss": 0.1989,
+      "step": 10343
+    },
+    {
+      "epoch": 0.165504,
+      "grad_norm": 0.439453125,
+      "learning_rate": 8.41241935483871e-05,
+      "loss": 0.171,
+      "step": 10344
+    },
+    {
+      "epoch": 0.16552,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.41225806451613e-05,
+      "loss": 0.1899,
+      "step": 10345
+    },
+    {
+      "epoch": 0.165536,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.412096774193548e-05,
+      "loss": 0.1822,
+      "step": 10346
+    },
+    {
+      "epoch": 0.165552,
+      "grad_norm": 0.75,
+      "learning_rate": 8.411935483870968e-05,
+      "loss": 0.2002,
+      "step": 10347
+    },
+    {
+      "epoch": 0.165568,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.411774193548387e-05,
+      "loss": 0.1768,
+      "step": 10348
+    },
+    {
+      "epoch": 0.165584,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.411612903225807e-05,
+      "loss": 0.1812,
+      "step": 10349
+    },
+    {
+      "epoch": 0.1656,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.411451612903226e-05,
+      "loss": 0.151,
+      "step": 10350
+    },
+    {
+      "epoch": 0.165616,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.411290322580645e-05,
+      "loss": 0.1662,
+      "step": 10351
+    },
+    {
+      "epoch": 0.165632,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.411129032258065e-05,
+      "loss": 0.166,
+      "step": 10352
+    },
+    {
+      "epoch": 0.165648,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.410967741935484e-05,
+      "loss": 0.168,
+      "step": 10353
+    },
+    {
+      "epoch": 0.165664,
+      "grad_norm": 1.546875,
+      "learning_rate": 8.410806451612904e-05,
+      "loss": 0.2214,
+      "step": 10354
+    },
+    {
+      "epoch": 0.16568,
+      "grad_norm": 1.4296875,
+      "learning_rate": 8.410645161290323e-05,
+      "loss": 0.1621,
+      "step": 10355
+    },
+    {
+      "epoch": 0.165696,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.410483870967743e-05,
+      "loss": 0.1486,
+      "step": 10356
+    },
+    {
+      "epoch": 0.165712,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.410322580645161e-05,
+      "loss": 0.145,
+      "step": 10357
+    },
+    {
+      "epoch": 0.165728,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.410161290322581e-05,
+      "loss": 0.1496,
+      "step": 10358
+    },
+    {
+      "epoch": 0.165744,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.41e-05,
+      "loss": 0.2014,
+      "step": 10359
+    },
+    {
+      "epoch": 0.16576,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.40983870967742e-05,
+      "loss": 0.182,
+      "step": 10360
+    },
+    {
+      "epoch": 0.165776,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.409677419354838e-05,
+      "loss": 0.1493,
+      "step": 10361
+    },
+    {
+      "epoch": 0.165792,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.409516129032258e-05,
+      "loss": 0.2648,
+      "step": 10362
+    },
+    {
+      "epoch": 0.165808,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.409354838709678e-05,
+      "loss": 0.1781,
+      "step": 10363
+    },
+    {
+      "epoch": 0.165824,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.409193548387098e-05,
+      "loss": 0.1863,
+      "step": 10364
+    },
+    {
+      "epoch": 0.16584,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.409032258064517e-05,
+      "loss": 0.2481,
+      "step": 10365
+    },
+    {
+      "epoch": 0.165856,
+      "grad_norm": 1.0,
+      "learning_rate": 8.408870967741937e-05,
+      "loss": 0.2011,
+      "step": 10366
+    },
+    {
+      "epoch": 0.165872,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.408709677419355e-05,
+      "loss": 0.1471,
+      "step": 10367
+    },
+    {
+      "epoch": 0.165888,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.408548387096774e-05,
+      "loss": 0.1557,
+      "step": 10368
+    },
+    {
+      "epoch": 0.165904,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.408387096774194e-05,
+      "loss": 0.1386,
+      "step": 10369
+    },
+    {
+      "epoch": 0.16592,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.408225806451613e-05,
+      "loss": 0.1846,
+      "step": 10370
+    },
+    {
+      "epoch": 0.165936,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.408064516129033e-05,
+      "loss": 0.1774,
+      "step": 10371
+    },
+    {
+      "epoch": 0.165952,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.407903225806451e-05,
+      "loss": 0.1923,
+      "step": 10372
+    },
+    {
+      "epoch": 0.165968,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.407741935483871e-05,
+      "loss": 0.1528,
+      "step": 10373
+    },
+    {
+      "epoch": 0.165984,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.40758064516129e-05,
+      "loss": 0.1996,
+      "step": 10374
+    },
+    {
+      "epoch": 0.166,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.40741935483871e-05,
+      "loss": 0.1481,
+      "step": 10375
+    },
+    {
+      "epoch": 0.166016,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.40725806451613e-05,
+      "loss": 0.1636,
+      "step": 10376
+    },
+    {
+      "epoch": 0.166032,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.40709677419355e-05,
+      "loss": 0.1561,
+      "step": 10377
+    },
+    {
+      "epoch": 0.166048,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.406935483870968e-05,
+      "loss": 0.1671,
+      "step": 10378
+    },
+    {
+      "epoch": 0.166064,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.406774193548388e-05,
+      "loss": 0.152,
+      "step": 10379
+    },
+    {
+      "epoch": 0.16608,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.406612903225807e-05,
+      "loss": 0.1802,
+      "step": 10380
+    },
+    {
+      "epoch": 0.166096,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.406451612903227e-05,
+      "loss": 0.2034,
+      "step": 10381
+    },
+    {
+      "epoch": 0.166112,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.406290322580645e-05,
+      "loss": 0.17,
+      "step": 10382
+    },
+    {
+      "epoch": 0.166128,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.406129032258065e-05,
+      "loss": 0.2071,
+      "step": 10383
+    },
+    {
+      "epoch": 0.166144,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.405967741935484e-05,
+      "loss": 0.1784,
+      "step": 10384
+    },
+    {
+      "epoch": 0.16616,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.405806451612903e-05,
+      "loss": 0.1643,
+      "step": 10385
+    },
+    {
+      "epoch": 0.166176,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.405645161290323e-05,
+      "loss": 0.1595,
+      "step": 10386
+    },
+    {
+      "epoch": 0.166192,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.405483870967742e-05,
+      "loss": 0.1614,
+      "step": 10387
+    },
+    {
+      "epoch": 0.166208,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.405322580645162e-05,
+      "loss": 0.1853,
+      "step": 10388
+    },
+    {
+      "epoch": 0.166224,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.405161290322581e-05,
+      "loss": 0.1714,
+      "step": 10389
+    },
+    {
+      "epoch": 0.16624,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.405000000000001e-05,
+      "loss": 0.1781,
+      "step": 10390
+    },
+    {
+      "epoch": 0.166256,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.40483870967742e-05,
+      "loss": 0.1813,
+      "step": 10391
+    },
+    {
+      "epoch": 0.166272,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.40467741935484e-05,
+      "loss": 0.1281,
+      "step": 10392
+    },
+    {
+      "epoch": 0.166288,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.404516129032258e-05,
+      "loss": 0.1504,
+      "step": 10393
+    },
+    {
+      "epoch": 0.166304,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.404354838709678e-05,
+      "loss": 0.1396,
+      "step": 10394
+    },
+    {
+      "epoch": 0.16632,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.404193548387097e-05,
+      "loss": 0.1583,
+      "step": 10395
+    },
+    {
+      "epoch": 0.166336,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.404032258064517e-05,
+      "loss": 0.1534,
+      "step": 10396
+    },
+    {
+      "epoch": 0.166352,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.403870967741935e-05,
+      "loss": 0.1615,
+      "step": 10397
+    },
+    {
+      "epoch": 0.166368,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.403709677419355e-05,
+      "loss": 0.187,
+      "step": 10398
+    },
+    {
+      "epoch": 0.166384,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.403548387096775e-05,
+      "loss": 0.1725,
+      "step": 10399
+    },
+    {
+      "epoch": 0.1664,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.403387096774194e-05,
+      "loss": 0.1803,
+      "step": 10400
+    },
+    {
+      "epoch": 0.166416,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.403225806451614e-05,
+      "loss": 0.1789,
+      "step": 10401
+    },
+    {
+      "epoch": 0.166432,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.403064516129032e-05,
+      "loss": 0.1455,
+      "step": 10402
+    },
+    {
+      "epoch": 0.166448,
+      "grad_norm": 1.328125,
+      "learning_rate": 8.402903225806452e-05,
+      "loss": 0.19,
+      "step": 10403
+    },
+    {
+      "epoch": 0.166464,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.402741935483871e-05,
+      "loss": 0.1834,
+      "step": 10404
+    },
+    {
+      "epoch": 0.16648,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.402580645161291e-05,
+      "loss": 0.1537,
+      "step": 10405
+    },
+    {
+      "epoch": 0.166496,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.40241935483871e-05,
+      "loss": 0.1575,
+      "step": 10406
+    },
+    {
+      "epoch": 0.166512,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.40225806451613e-05,
+      "loss": 0.1746,
+      "step": 10407
+    },
+    {
+      "epoch": 0.166528,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.402096774193548e-05,
+      "loss": 0.2106,
+      "step": 10408
+    },
+    {
+      "epoch": 0.166544,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.401935483870968e-05,
+      "loss": 0.1668,
+      "step": 10409
+    },
+    {
+      "epoch": 0.16656,
+      "grad_norm": 1.4375,
+      "learning_rate": 8.401774193548387e-05,
+      "loss": 0.1566,
+      "step": 10410
+    },
+    {
+      "epoch": 0.166576,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.401612903225807e-05,
+      "loss": 0.1869,
+      "step": 10411
+    },
+    {
+      "epoch": 0.166592,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.401451612903227e-05,
+      "loss": 0.1642,
+      "step": 10412
+    },
+    {
+      "epoch": 0.166608,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.401290322580647e-05,
+      "loss": 0.155,
+      "step": 10413
+    },
+    {
+      "epoch": 0.166624,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.401129032258065e-05,
+      "loss": 0.1933,
+      "step": 10414
+    },
+    {
+      "epoch": 0.16664,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.400967741935484e-05,
+      "loss": 0.1916,
+      "step": 10415
+    },
+    {
+      "epoch": 0.166656,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.400806451612904e-05,
+      "loss": 0.1586,
+      "step": 10416
+    },
+    {
+      "epoch": 0.166672,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.400645161290322e-05,
+      "loss": 0.1743,
+      "step": 10417
+    },
+    {
+      "epoch": 0.166688,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.400483870967742e-05,
+      "loss": 0.1532,
+      "step": 10418
+    },
+    {
+      "epoch": 0.166704,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.400322580645161e-05,
+      "loss": 0.1835,
+      "step": 10419
+    },
+    {
+      "epoch": 0.16672,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.400161290322581e-05,
+      "loss": 0.1848,
+      "step": 10420
+    },
+    {
+      "epoch": 0.166736,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.4e-05,
+      "loss": 0.1614,
+      "step": 10421
+    },
+    {
+      "epoch": 0.166752,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.39983870967742e-05,
+      "loss": 0.1962,
+      "step": 10422
+    },
+    {
+      "epoch": 0.166768,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.39967741935484e-05,
+      "loss": 0.1611,
+      "step": 10423
+    },
+    {
+      "epoch": 0.166784,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.39951612903226e-05,
+      "loss": 0.1588,
+      "step": 10424
+    },
+    {
+      "epoch": 0.1668,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.399354838709678e-05,
+      "loss": 0.1768,
+      "step": 10425
+    },
+    {
+      "epoch": 0.166816,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.399193548387098e-05,
+      "loss": 0.1536,
+      "step": 10426
+    },
+    {
+      "epoch": 0.166832,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.399032258064517e-05,
+      "loss": 0.1951,
+      "step": 10427
+    },
+    {
+      "epoch": 0.166848,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.398870967741937e-05,
+      "loss": 0.1734,
+      "step": 10428
+    },
+    {
+      "epoch": 0.166864,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.398709677419355e-05,
+      "loss": 0.1423,
+      "step": 10429
+    },
+    {
+      "epoch": 0.16688,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.398548387096774e-05,
+      "loss": 0.1507,
+      "step": 10430
+    },
+    {
+      "epoch": 0.166896,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.398387096774194e-05,
+      "loss": 0.1837,
+      "step": 10431
+    },
+    {
+      "epoch": 0.166912,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.398225806451612e-05,
+      "loss": 0.1681,
+      "step": 10432
+    },
+    {
+      "epoch": 0.166928,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.398064516129032e-05,
+      "loss": 0.1916,
+      "step": 10433
+    },
+    {
+      "epoch": 0.166944,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.397903225806452e-05,
+      "loss": 0.195,
+      "step": 10434
+    },
+    {
+      "epoch": 0.16696,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.397741935483871e-05,
+      "loss": 0.1745,
+      "step": 10435
+    },
+    {
+      "epoch": 0.166976,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.397580645161291e-05,
+      "loss": 0.2014,
+      "step": 10436
+    },
+    {
+      "epoch": 0.166992,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.397419354838711e-05,
+      "loss": 0.1466,
+      "step": 10437
+    },
+    {
+      "epoch": 0.167008,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.39725806451613e-05,
+      "loss": 0.1829,
+      "step": 10438
+    },
+    {
+      "epoch": 0.167024,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.39709677419355e-05,
+      "loss": 0.1839,
+      "step": 10439
+    },
+    {
+      "epoch": 0.16704,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.396935483870968e-05,
+      "loss": 0.1633,
+      "step": 10440
+    },
+    {
+      "epoch": 0.167056,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.396774193548388e-05,
+      "loss": 0.1883,
+      "step": 10441
+    },
+    {
+      "epoch": 0.167072,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.396612903225807e-05,
+      "loss": 0.1984,
+      "step": 10442
+    },
+    {
+      "epoch": 0.167088,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.396451612903227e-05,
+      "loss": 0.1775,
+      "step": 10443
+    },
+    {
+      "epoch": 0.167104,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.396290322580645e-05,
+      "loss": 0.1837,
+      "step": 10444
+    },
+    {
+      "epoch": 0.16712,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.396129032258065e-05,
+      "loss": 0.1759,
+      "step": 10445
+    },
+    {
+      "epoch": 0.167136,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.395967741935484e-05,
+      "loss": 0.1569,
+      "step": 10446
+    },
+    {
+      "epoch": 0.167152,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.395806451612904e-05,
+      "loss": 0.1195,
+      "step": 10447
+    },
+    {
+      "epoch": 0.167168,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.395645161290324e-05,
+      "loss": 0.1663,
+      "step": 10448
+    },
+    {
+      "epoch": 0.167184,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.395483870967742e-05,
+      "loss": 0.2034,
+      "step": 10449
+    },
+    {
+      "epoch": 0.1672,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.395322580645162e-05,
+      "loss": 0.1938,
+      "step": 10450
+    },
+    {
+      "epoch": 0.167216,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.395161290322581e-05,
+      "loss": 0.1858,
+      "step": 10451
+    },
+    {
+      "epoch": 0.167232,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.395000000000001e-05,
+      "loss": 0.2338,
+      "step": 10452
+    },
+    {
+      "epoch": 0.167248,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.39483870967742e-05,
+      "loss": 0.1835,
+      "step": 10453
+    },
+    {
+      "epoch": 0.167264,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.39467741935484e-05,
+      "loss": 0.1697,
+      "step": 10454
+    },
+    {
+      "epoch": 0.16728,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.394516129032258e-05,
+      "loss": 0.1681,
+      "step": 10455
+    },
+    {
+      "epoch": 0.167296,
+      "grad_norm": 0.625,
+      "learning_rate": 8.394354838709678e-05,
+      "loss": 0.1598,
+      "step": 10456
+    },
+    {
+      "epoch": 0.167312,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.394193548387097e-05,
+      "loss": 0.1854,
+      "step": 10457
+    },
+    {
+      "epoch": 0.167328,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.394032258064516e-05,
+      "loss": 0.1736,
+      "step": 10458
+    },
+    {
+      "epoch": 0.167344,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.393870967741936e-05,
+      "loss": 0.1554,
+      "step": 10459
+    },
+    {
+      "epoch": 0.16736,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.393709677419356e-05,
+      "loss": 0.1349,
+      "step": 10460
+    },
+    {
+      "epoch": 0.167376,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.393548387096775e-05,
+      "loss": 0.1823,
+      "step": 10461
+    },
+    {
+      "epoch": 0.167392,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.393387096774194e-05,
+      "loss": 0.1524,
+      "step": 10462
+    },
+    {
+      "epoch": 0.167408,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.393225806451614e-05,
+      "loss": 0.1193,
+      "step": 10463
+    },
+    {
+      "epoch": 0.167424,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.393064516129032e-05,
+      "loss": 0.1683,
+      "step": 10464
+    },
+    {
+      "epoch": 0.16744,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.392903225806452e-05,
+      "loss": 0.1826,
+      "step": 10465
+    },
+    {
+      "epoch": 0.167456,
+      "grad_norm": 1.3515625,
+      "learning_rate": 8.392741935483871e-05,
+      "loss": 0.2114,
+      "step": 10466
+    },
+    {
+      "epoch": 0.167472,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.392580645161291e-05,
+      "loss": 0.123,
+      "step": 10467
+    },
+    {
+      "epoch": 0.167488,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.39241935483871e-05,
+      "loss": 0.1845,
+      "step": 10468
+    },
+    {
+      "epoch": 0.167504,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.392258064516129e-05,
+      "loss": 0.1477,
+      "step": 10469
+    },
+    {
+      "epoch": 0.16752,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.392096774193548e-05,
+      "loss": 0.1251,
+      "step": 10470
+    },
+    {
+      "epoch": 0.167536,
+      "grad_norm": 1.328125,
+      "learning_rate": 8.391935483870968e-05,
+      "loss": 0.202,
+      "step": 10471
+    },
+    {
+      "epoch": 0.167552,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.391774193548388e-05,
+      "loss": 0.217,
+      "step": 10472
+    },
+    {
+      "epoch": 0.167568,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.391612903225808e-05,
+      "loss": 0.198,
+      "step": 10473
+    },
+    {
+      "epoch": 0.167584,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.391451612903226e-05,
+      "loss": 0.1659,
+      "step": 10474
+    },
+    {
+      "epoch": 0.1676,
+      "grad_norm": 0.625,
+      "learning_rate": 8.391290322580646e-05,
+      "loss": 0.1855,
+      "step": 10475
+    },
+    {
+      "epoch": 0.167616,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.391129032258065e-05,
+      "loss": 0.142,
+      "step": 10476
+    },
+    {
+      "epoch": 0.167632,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.390967741935484e-05,
+      "loss": 0.1465,
+      "step": 10477
+    },
+    {
+      "epoch": 0.167648,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.390806451612904e-05,
+      "loss": 0.1748,
+      "step": 10478
+    },
+    {
+      "epoch": 0.167664,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.390645161290322e-05,
+      "loss": 0.1978,
+      "step": 10479
+    },
+    {
+      "epoch": 0.16768,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.390483870967742e-05,
+      "loss": 0.137,
+      "step": 10480
+    },
+    {
+      "epoch": 0.167696,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.390322580645161e-05,
+      "loss": 0.1644,
+      "step": 10481
+    },
+    {
+      "epoch": 0.167712,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.390161290322581e-05,
+      "loss": 0.2018,
+      "step": 10482
+    },
+    {
+      "epoch": 0.167728,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.39e-05,
+      "loss": 0.1609,
+      "step": 10483
+    },
+    {
+      "epoch": 0.167744,
+      "grad_norm": 0.95703125,
+      "learning_rate": 8.38983870967742e-05,
+      "loss": 0.187,
+      "step": 10484
+    },
+    {
+      "epoch": 0.16776,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.389677419354839e-05,
+      "loss": 0.1521,
+      "step": 10485
+    },
+    {
+      "epoch": 0.167776,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.389516129032259e-05,
+      "loss": 0.1702,
+      "step": 10486
+    },
+    {
+      "epoch": 0.167792,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.389354838709678e-05,
+      "loss": 0.1836,
+      "step": 10487
+    },
+    {
+      "epoch": 0.167808,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.389193548387098e-05,
+      "loss": 0.2259,
+      "step": 10488
+    },
+    {
+      "epoch": 0.167824,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.389032258064516e-05,
+      "loss": 0.1566,
+      "step": 10489
+    },
+    {
+      "epoch": 0.16784,
+      "grad_norm": 0.625,
+      "learning_rate": 8.388870967741936e-05,
+      "loss": 0.1746,
+      "step": 10490
+    },
+    {
+      "epoch": 0.167856,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.388709677419355e-05,
+      "loss": 0.1681,
+      "step": 10491
+    },
+    {
+      "epoch": 0.167872,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.388548387096774e-05,
+      "loss": 0.174,
+      "step": 10492
+    },
+    {
+      "epoch": 0.167888,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.388387096774194e-05,
+      "loss": 0.1953,
+      "step": 10493
+    },
+    {
+      "epoch": 0.167904,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.388225806451613e-05,
+      "loss": 0.2138,
+      "step": 10494
+    },
+    {
+      "epoch": 0.16792,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.388064516129033e-05,
+      "loss": 0.1789,
+      "step": 10495
+    },
+    {
+      "epoch": 0.167936,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.387903225806452e-05,
+      "loss": 0.1961,
+      "step": 10496
+    },
+    {
+      "epoch": 0.167952,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.387741935483872e-05,
+      "loss": 0.1418,
+      "step": 10497
+    },
+    {
+      "epoch": 0.167968,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.38758064516129e-05,
+      "loss": 0.2006,
+      "step": 10498
+    },
+    {
+      "epoch": 0.167984,
+      "grad_norm": 1.0,
+      "learning_rate": 8.38741935483871e-05,
+      "loss": 0.197,
+      "step": 10499
+    },
+    {
+      "epoch": 0.168,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.387258064516129e-05,
+      "loss": 0.144,
+      "step": 10500
+    },
+    {
+      "epoch": 0.168016,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.387096774193549e-05,
+      "loss": 0.1826,
+      "step": 10501
+    },
+    {
+      "epoch": 0.168032,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.386935483870968e-05,
+      "loss": 0.1719,
+      "step": 10502
+    },
+    {
+      "epoch": 0.168048,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.386774193548388e-05,
+      "loss": 0.2011,
+      "step": 10503
+    },
+    {
+      "epoch": 0.168064,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.386612903225806e-05,
+      "loss": 0.1606,
+      "step": 10504
+    },
+    {
+      "epoch": 0.16808,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.386451612903226e-05,
+      "loss": 0.1738,
+      "step": 10505
+    },
+    {
+      "epoch": 0.168096,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.386290322580645e-05,
+      "loss": 0.1914,
+      "step": 10506
+    },
+    {
+      "epoch": 0.168112,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.386129032258065e-05,
+      "loss": 0.1562,
+      "step": 10507
+    },
+    {
+      "epoch": 0.168128,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.385967741935485e-05,
+      "loss": 0.1874,
+      "step": 10508
+    },
+    {
+      "epoch": 0.168144,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.385806451612903e-05,
+      "loss": 0.1412,
+      "step": 10509
+    },
+    {
+      "epoch": 0.16816,
+      "grad_norm": 1.0,
+      "learning_rate": 8.385645161290323e-05,
+      "loss": 0.1823,
+      "step": 10510
+    },
+    {
+      "epoch": 0.168176,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.385483870967742e-05,
+      "loss": 0.196,
+      "step": 10511
+    },
+    {
+      "epoch": 0.168192,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.385322580645162e-05,
+      "loss": 0.1708,
+      "step": 10512
+    },
+    {
+      "epoch": 0.168208,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.38516129032258e-05,
+      "loss": 0.1669,
+      "step": 10513
+    },
+    {
+      "epoch": 0.168224,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.385e-05,
+      "loss": 0.2004,
+      "step": 10514
+    },
+    {
+      "epoch": 0.16824,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.384838709677419e-05,
+      "loss": 0.1563,
+      "step": 10515
+    },
+    {
+      "epoch": 0.168256,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.384677419354839e-05,
+      "loss": 0.1695,
+      "step": 10516
+    },
+    {
+      "epoch": 0.168272,
+      "grad_norm": 1.3203125,
+      "learning_rate": 8.384516129032258e-05,
+      "loss": 0.2057,
+      "step": 10517
+    },
+    {
+      "epoch": 0.168288,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.384354838709678e-05,
+      "loss": 0.1836,
+      "step": 10518
+    },
+    {
+      "epoch": 0.168304,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.384193548387098e-05,
+      "loss": 0.1845,
+      "step": 10519
+    },
+    {
+      "epoch": 0.16832,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.384032258064518e-05,
+      "loss": 0.1573,
+      "step": 10520
+    },
+    {
+      "epoch": 0.168336,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.383870967741936e-05,
+      "loss": 0.1811,
+      "step": 10521
+    },
+    {
+      "epoch": 0.168352,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.383709677419356e-05,
+      "loss": 0.1788,
+      "step": 10522
+    },
+    {
+      "epoch": 0.168368,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.383548387096775e-05,
+      "loss": 0.1895,
+      "step": 10523
+    },
+    {
+      "epoch": 0.168384,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.383387096774193e-05,
+      "loss": 0.191,
+      "step": 10524
+    },
+    {
+      "epoch": 0.1684,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.383225806451613e-05,
+      "loss": 0.1614,
+      "step": 10525
+    },
+    {
+      "epoch": 0.168416,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.383064516129032e-05,
+      "loss": 0.1556,
+      "step": 10526
+    },
+    {
+      "epoch": 0.168432,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.382903225806452e-05,
+      "loss": 0.1721,
+      "step": 10527
+    },
+    {
+      "epoch": 0.168448,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.38274193548387e-05,
+      "loss": 0.1543,
+      "step": 10528
+    },
+    {
+      "epoch": 0.168464,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.38258064516129e-05,
+      "loss": 0.1487,
+      "step": 10529
+    },
+    {
+      "epoch": 0.16848,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.38241935483871e-05,
+      "loss": 0.1396,
+      "step": 10530
+    },
+    {
+      "epoch": 0.168496,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.382258064516129e-05,
+      "loss": 0.2166,
+      "step": 10531
+    },
+    {
+      "epoch": 0.168512,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.382096774193549e-05,
+      "loss": 0.1317,
+      "step": 10532
+    },
+    {
+      "epoch": 0.168528,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.381935483870969e-05,
+      "loss": 0.1935,
+      "step": 10533
+    },
+    {
+      "epoch": 0.168544,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.381774193548388e-05,
+      "loss": 0.185,
+      "step": 10534
+    },
+    {
+      "epoch": 0.16856,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.381612903225808e-05,
+      "loss": 0.1765,
+      "step": 10535
+    },
+    {
+      "epoch": 0.168576,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.381451612903226e-05,
+      "loss": 0.2253,
+      "step": 10536
+    },
+    {
+      "epoch": 0.168592,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.381290322580646e-05,
+      "loss": 0.1387,
+      "step": 10537
+    },
+    {
+      "epoch": 0.168608,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.381129032258065e-05,
+      "loss": 0.1395,
+      "step": 10538
+    },
+    {
+      "epoch": 0.168624,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.380967741935483e-05,
+      "loss": 0.167,
+      "step": 10539
+    },
+    {
+      "epoch": 0.16864,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.380806451612903e-05,
+      "loss": 0.1875,
+      "step": 10540
+    },
+    {
+      "epoch": 0.168656,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.380645161290322e-05,
+      "loss": 0.17,
+      "step": 10541
+    },
+    {
+      "epoch": 0.168672,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.380483870967742e-05,
+      "loss": 0.1564,
+      "step": 10542
+    },
+    {
+      "epoch": 0.168688,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.380322580645162e-05,
+      "loss": 0.172,
+      "step": 10543
+    },
+    {
+      "epoch": 0.168704,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.380161290322582e-05,
+      "loss": 0.2634,
+      "step": 10544
+    },
+    {
+      "epoch": 0.16872,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.38e-05,
+      "loss": 0.1657,
+      "step": 10545
+    },
+    {
+      "epoch": 0.168736,
+      "grad_norm": 1.390625,
+      "learning_rate": 8.37983870967742e-05,
+      "loss": 0.22,
+      "step": 10546
+    },
+    {
+      "epoch": 0.168752,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.379677419354839e-05,
+      "loss": 0.1659,
+      "step": 10547
+    },
+    {
+      "epoch": 0.168768,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.379516129032259e-05,
+      "loss": 0.1746,
+      "step": 10548
+    },
+    {
+      "epoch": 0.168784,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.379354838709678e-05,
+      "loss": 0.2318,
+      "step": 10549
+    },
+    {
+      "epoch": 0.1688,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.379193548387098e-05,
+      "loss": 0.2091,
+      "step": 10550
+    },
+    {
+      "epoch": 0.168816,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.379032258064516e-05,
+      "loss": 0.1515,
+      "step": 10551
+    },
+    {
+      "epoch": 0.168832,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.378870967741936e-05,
+      "loss": 0.2189,
+      "step": 10552
+    },
+    {
+      "epoch": 0.168848,
+      "grad_norm": 1.3359375,
+      "learning_rate": 8.378709677419355e-05,
+      "loss": 0.1343,
+      "step": 10553
+    },
+    {
+      "epoch": 0.168864,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.378548387096775e-05,
+      "loss": 0.1677,
+      "step": 10554
+    },
+    {
+      "epoch": 0.16888,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.378387096774195e-05,
+      "loss": 0.1742,
+      "step": 10555
+    },
+    {
+      "epoch": 0.168896,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.378225806451613e-05,
+      "loss": 0.2001,
+      "step": 10556
+    },
+    {
+      "epoch": 0.168912,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.378064516129033e-05,
+      "loss": 0.157,
+      "step": 10557
+    },
+    {
+      "epoch": 0.168928,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.377903225806452e-05,
+      "loss": 0.1896,
+      "step": 10558
+    },
+    {
+      "epoch": 0.168944,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.377741935483872e-05,
+      "loss": 0.1721,
+      "step": 10559
+    },
+    {
+      "epoch": 0.16896,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.37758064516129e-05,
+      "loss": 0.1658,
+      "step": 10560
+    },
+    {
+      "epoch": 0.168976,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.37741935483871e-05,
+      "loss": 0.1677,
+      "step": 10561
+    },
+    {
+      "epoch": 0.168992,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.377258064516129e-05,
+      "loss": 0.1955,
+      "step": 10562
+    },
+    {
+      "epoch": 0.169008,
+      "grad_norm": 0.458984375,
+      "learning_rate": 8.377096774193549e-05,
+      "loss": 0.1471,
+      "step": 10563
+    },
+    {
+      "epoch": 0.169024,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.376935483870968e-05,
+      "loss": 0.1641,
+      "step": 10564
+    },
+    {
+      "epoch": 0.16904,
+      "grad_norm": 1.390625,
+      "learning_rate": 8.376774193548387e-05,
+      "loss": 0.2091,
+      "step": 10565
+    },
+    {
+      "epoch": 0.169056,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.376612903225806e-05,
+      "loss": 0.2098,
+      "step": 10566
+    },
+    {
+      "epoch": 0.169072,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.376451612903226e-05,
+      "loss": 0.1745,
+      "step": 10567
+    },
+    {
+      "epoch": 0.169088,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.376290322580646e-05,
+      "loss": 0.1793,
+      "step": 10568
+    },
+    {
+      "epoch": 0.169104,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.376129032258066e-05,
+      "loss": 0.1634,
+      "step": 10569
+    },
+    {
+      "epoch": 0.16912,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.375967741935485e-05,
+      "loss": 0.2077,
+      "step": 10570
+    },
+    {
+      "epoch": 0.169136,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.375806451612903e-05,
+      "loss": 0.1521,
+      "step": 10571
+    },
+    {
+      "epoch": 0.169152,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.375645161290323e-05,
+      "loss": 0.1995,
+      "step": 10572
+    },
+    {
+      "epoch": 0.169168,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.375483870967742e-05,
+      "loss": 0.1272,
+      "step": 10573
+    },
+    {
+      "epoch": 0.169184,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.375322580645162e-05,
+      "loss": 0.1555,
+      "step": 10574
+    },
+    {
+      "epoch": 0.1692,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.37516129032258e-05,
+      "loss": 0.2227,
+      "step": 10575
+    },
+    {
+      "epoch": 0.169216,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.375e-05,
+      "loss": 0.1942,
+      "step": 10576
+    },
+    {
+      "epoch": 0.169232,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.374838709677419e-05,
+      "loss": 0.221,
+      "step": 10577
+    },
+    {
+      "epoch": 0.169248,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.374677419354839e-05,
+      "loss": 0.1607,
+      "step": 10578
+    },
+    {
+      "epoch": 0.169264,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.374516129032259e-05,
+      "loss": 0.1736,
+      "step": 10579
+    },
+    {
+      "epoch": 0.16928,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.374354838709679e-05,
+      "loss": 0.1823,
+      "step": 10580
+    },
+    {
+      "epoch": 0.169296,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.374193548387097e-05,
+      "loss": 0.1656,
+      "step": 10581
+    },
+    {
+      "epoch": 0.169312,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.374032258064517e-05,
+      "loss": 0.1345,
+      "step": 10582
+    },
+    {
+      "epoch": 0.169328,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.373870967741936e-05,
+      "loss": 0.1426,
+      "step": 10583
+    },
+    {
+      "epoch": 0.169344,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.373709677419356e-05,
+      "loss": 0.2093,
+      "step": 10584
+    },
+    {
+      "epoch": 0.16936,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.373548387096775e-05,
+      "loss": 0.1497,
+      "step": 10585
+    },
+    {
+      "epoch": 0.169376,
+      "grad_norm": 1.2734375,
+      "learning_rate": 8.373387096774193e-05,
+      "loss": 0.1516,
+      "step": 10586
+    },
+    {
+      "epoch": 0.169392,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.373225806451613e-05,
+      "loss": 0.1776,
+      "step": 10587
+    },
+    {
+      "epoch": 0.169408,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.373064516129032e-05,
+      "loss": 0.2166,
+      "step": 10588
+    },
+    {
+      "epoch": 0.169424,
+      "grad_norm": 1.2734375,
+      "learning_rate": 8.372903225806452e-05,
+      "loss": 0.2034,
+      "step": 10589
+    },
+    {
+      "epoch": 0.16944,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.372741935483872e-05,
+      "loss": 0.1902,
+      "step": 10590
+    },
+    {
+      "epoch": 0.169456,
+      "grad_norm": 1.3828125,
+      "learning_rate": 8.372580645161292e-05,
+      "loss": 0.2083,
+      "step": 10591
+    },
+    {
+      "epoch": 0.169472,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.37241935483871e-05,
+      "loss": 0.1749,
+      "step": 10592
+    },
+    {
+      "epoch": 0.169488,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.37225806451613e-05,
+      "loss": 0.1303,
+      "step": 10593
+    },
+    {
+      "epoch": 0.169504,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.372096774193549e-05,
+      "loss": 0.1794,
+      "step": 10594
+    },
+    {
+      "epoch": 0.16952,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.371935483870969e-05,
+      "loss": 0.1846,
+      "step": 10595
+    },
+    {
+      "epoch": 0.169536,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.371774193548387e-05,
+      "loss": 0.1822,
+      "step": 10596
+    },
+    {
+      "epoch": 0.169552,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.371612903225807e-05,
+      "loss": 0.1463,
+      "step": 10597
+    },
+    {
+      "epoch": 0.169568,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.371451612903226e-05,
+      "loss": 0.1447,
+      "step": 10598
+    },
+    {
+      "epoch": 0.169584,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.371290322580646e-05,
+      "loss": 0.1676,
+      "step": 10599
+    },
+    {
+      "epoch": 0.1696,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.371129032258064e-05,
+      "loss": 0.1722,
+      "step": 10600
+    },
+    {
+      "epoch": 0.169616,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.370967741935483e-05,
+      "loss": 0.1987,
+      "step": 10601
+    },
+    {
+      "epoch": 0.169632,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.370806451612903e-05,
+      "loss": 0.1929,
+      "step": 10602
+    },
+    {
+      "epoch": 0.169648,
+      "grad_norm": 0.75,
+      "learning_rate": 8.370645161290323e-05,
+      "loss": 0.1824,
+      "step": 10603
+    },
+    {
+      "epoch": 0.169664,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.370483870967743e-05,
+      "loss": 0.1742,
+      "step": 10604
+    },
+    {
+      "epoch": 0.16968,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.370322580645162e-05,
+      "loss": 0.1874,
+      "step": 10605
+    },
+    {
+      "epoch": 0.169696,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.370161290322582e-05,
+      "loss": 0.2146,
+      "step": 10606
+    },
+    {
+      "epoch": 0.169712,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.37e-05,
+      "loss": 0.1682,
+      "step": 10607
+    },
+    {
+      "epoch": 0.169728,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.36983870967742e-05,
+      "loss": 0.1566,
+      "step": 10608
+    },
+    {
+      "epoch": 0.169744,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.369677419354839e-05,
+      "loss": 0.1803,
+      "step": 10609
+    },
+    {
+      "epoch": 0.16976,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.369516129032259e-05,
+      "loss": 0.1735,
+      "step": 10610
+    },
+    {
+      "epoch": 0.169776,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.369354838709677e-05,
+      "loss": 0.1667,
+      "step": 10611
+    },
+    {
+      "epoch": 0.169792,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.369193548387097e-05,
+      "loss": 0.1857,
+      "step": 10612
+    },
+    {
+      "epoch": 0.169808,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.369032258064516e-05,
+      "loss": 0.1873,
+      "step": 10613
+    },
+    {
+      "epoch": 0.169824,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.368870967741936e-05,
+      "loss": 0.168,
+      "step": 10614
+    },
+    {
+      "epoch": 0.16984,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.368709677419356e-05,
+      "loss": 0.1609,
+      "step": 10615
+    },
+    {
+      "epoch": 0.169856,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.368548387096776e-05,
+      "loss": 0.1304,
+      "step": 10616
+    },
+    {
+      "epoch": 0.169872,
+      "grad_norm": 0.482421875,
+      "learning_rate": 8.368387096774194e-05,
+      "loss": 0.1653,
+      "step": 10617
+    },
+    {
+      "epoch": 0.169888,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.368225806451613e-05,
+      "loss": 0.1131,
+      "step": 10618
+    },
+    {
+      "epoch": 0.169904,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.368064516129033e-05,
+      "loss": 0.1775,
+      "step": 10619
+    },
+    {
+      "epoch": 0.16992,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.367903225806452e-05,
+      "loss": 0.1556,
+      "step": 10620
+    },
+    {
+      "epoch": 0.169936,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.367741935483872e-05,
+      "loss": 0.1683,
+      "step": 10621
+    },
+    {
+      "epoch": 0.169952,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.36758064516129e-05,
+      "loss": 0.1539,
+      "step": 10622
+    },
+    {
+      "epoch": 0.169968,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.36741935483871e-05,
+      "loss": 0.1967,
+      "step": 10623
+    },
+    {
+      "epoch": 0.169984,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.367258064516129e-05,
+      "loss": 0.1984,
+      "step": 10624
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.367096774193549e-05,
+      "loss": 0.1903,
+      "step": 10625
+    },
+    {
+      "epoch": 0.170016,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.366935483870967e-05,
+      "loss": 0.1907,
+      "step": 10626
+    },
+    {
+      "epoch": 0.170032,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.366774193548387e-05,
+      "loss": 0.1885,
+      "step": 10627
+    },
+    {
+      "epoch": 0.170048,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.366612903225807e-05,
+      "loss": 0.1763,
+      "step": 10628
+    },
+    {
+      "epoch": 0.170064,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.366451612903227e-05,
+      "loss": 0.1988,
+      "step": 10629
+    },
+    {
+      "epoch": 0.17008,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.366290322580646e-05,
+      "loss": 0.1883,
+      "step": 10630
+    },
+    {
+      "epoch": 0.170096,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.366129032258066e-05,
+      "loss": 0.1247,
+      "step": 10631
+    },
+    {
+      "epoch": 0.170112,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.365967741935484e-05,
+      "loss": 0.1452,
+      "step": 10632
+    },
+    {
+      "epoch": 0.170128,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.365806451612903e-05,
+      "loss": 0.195,
+      "step": 10633
+    },
+    {
+      "epoch": 0.170144,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.365645161290323e-05,
+      "loss": 0.1772,
+      "step": 10634
+    },
+    {
+      "epoch": 0.17016,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.365483870967742e-05,
+      "loss": 0.1565,
+      "step": 10635
+    },
+    {
+      "epoch": 0.170176,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.365322580645161e-05,
+      "loss": 0.1528,
+      "step": 10636
+    },
+    {
+      "epoch": 0.170192,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.36516129032258e-05,
+      "loss": 0.1667,
+      "step": 10637
+    },
+    {
+      "epoch": 0.170208,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.365e-05,
+      "loss": 0.2129,
+      "step": 10638
+    },
+    {
+      "epoch": 0.170224,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.36483870967742e-05,
+      "loss": 0.1564,
+      "step": 10639
+    },
+    {
+      "epoch": 0.17024,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.36467741935484e-05,
+      "loss": 0.2154,
+      "step": 10640
+    },
+    {
+      "epoch": 0.170256,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.364516129032259e-05,
+      "loss": 0.1271,
+      "step": 10641
+    },
+    {
+      "epoch": 0.170272,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.364354838709679e-05,
+      "loss": 0.1695,
+      "step": 10642
+    },
+    {
+      "epoch": 0.170288,
+      "grad_norm": 1.3046875,
+      "learning_rate": 8.364193548387097e-05,
+      "loss": 0.1582,
+      "step": 10643
+    },
+    {
+      "epoch": 0.170304,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.364032258064517e-05,
+      "loss": 0.1859,
+      "step": 10644
+    },
+    {
+      "epoch": 0.17032,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.363870967741936e-05,
+      "loss": 0.1542,
+      "step": 10645
+    },
+    {
+      "epoch": 0.170336,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.363709677419356e-05,
+      "loss": 0.1568,
+      "step": 10646
+    },
+    {
+      "epoch": 0.170352,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.363548387096774e-05,
+      "loss": 0.1564,
+      "step": 10647
+    },
+    {
+      "epoch": 0.170368,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.363387096774193e-05,
+      "loss": 0.1509,
+      "step": 10648
+    },
+    {
+      "epoch": 0.170384,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.363225806451613e-05,
+      "loss": 0.1886,
+      "step": 10649
+    },
+    {
+      "epoch": 0.1704,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.363064516129033e-05,
+      "loss": 0.1702,
+      "step": 10650
+    },
+    {
+      "epoch": 0.170416,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.362903225806453e-05,
+      "loss": 0.1364,
+      "step": 10651
+    },
+    {
+      "epoch": 0.170432,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.362741935483871e-05,
+      "loss": 0.1443,
+      "step": 10652
+    },
+    {
+      "epoch": 0.170448,
+      "grad_norm": 1.5546875,
+      "learning_rate": 8.362580645161291e-05,
+      "loss": 0.1608,
+      "step": 10653
+    },
+    {
+      "epoch": 0.170464,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.36241935483871e-05,
+      "loss": 0.1686,
+      "step": 10654
+    },
+    {
+      "epoch": 0.17048,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.36225806451613e-05,
+      "loss": 0.1971,
+      "step": 10655
+    },
+    {
+      "epoch": 0.170496,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.362096774193549e-05,
+      "loss": 0.1898,
+      "step": 10656
+    },
+    {
+      "epoch": 0.170512,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.361935483870968e-05,
+      "loss": 0.175,
+      "step": 10657
+    },
+    {
+      "epoch": 0.170528,
+      "grad_norm": 1.25,
+      "learning_rate": 8.361774193548387e-05,
+      "loss": 0.1764,
+      "step": 10658
+    },
+    {
+      "epoch": 0.170544,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.361612903225807e-05,
+      "loss": 0.1981,
+      "step": 10659
+    },
+    {
+      "epoch": 0.17056,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.361451612903226e-05,
+      "loss": 0.1799,
+      "step": 10660
+    },
+    {
+      "epoch": 0.170576,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.361290322580646e-05,
+      "loss": 0.1737,
+      "step": 10661
+    },
+    {
+      "epoch": 0.170592,
+      "grad_norm": 1.4375,
+      "learning_rate": 8.361129032258064e-05,
+      "loss": 0.1691,
+      "step": 10662
+    },
+    {
+      "epoch": 0.170608,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.360967741935484e-05,
+      "loss": 0.1614,
+      "step": 10663
+    },
+    {
+      "epoch": 0.170624,
+      "grad_norm": 1.28125,
+      "learning_rate": 8.360806451612904e-05,
+      "loss": 0.1737,
+      "step": 10664
+    },
+    {
+      "epoch": 0.17064,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.360645161290323e-05,
+      "loss": 0.2298,
+      "step": 10665
+    },
+    {
+      "epoch": 0.170656,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.360483870967743e-05,
+      "loss": 0.1808,
+      "step": 10666
+    },
+    {
+      "epoch": 0.170672,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.360322580645161e-05,
+      "loss": 0.1628,
+      "step": 10667
+    },
+    {
+      "epoch": 0.170688,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.360161290322581e-05,
+      "loss": 0.2042,
+      "step": 10668
+    },
+    {
+      "epoch": 0.170704,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.36e-05,
+      "loss": 0.1678,
+      "step": 10669
+    },
+    {
+      "epoch": 0.17072,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.35983870967742e-05,
+      "loss": 0.1855,
+      "step": 10670
+    },
+    {
+      "epoch": 0.170736,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.359677419354838e-05,
+      "loss": 0.1678,
+      "step": 10671
+    },
+    {
+      "epoch": 0.170752,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.359516129032258e-05,
+      "loss": 0.1838,
+      "step": 10672
+    },
+    {
+      "epoch": 0.170768,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.359354838709677e-05,
+      "loss": 0.1385,
+      "step": 10673
+    },
+    {
+      "epoch": 0.170784,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.359193548387097e-05,
+      "loss": 0.1742,
+      "step": 10674
+    },
+    {
+      "epoch": 0.1708,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.359032258064517e-05,
+      "loss": 0.1421,
+      "step": 10675
+    },
+    {
+      "epoch": 0.170816,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.358870967741937e-05,
+      "loss": 0.152,
+      "step": 10676
+    },
+    {
+      "epoch": 0.170832,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.358709677419356e-05,
+      "loss": 0.1759,
+      "step": 10677
+    },
+    {
+      "epoch": 0.170848,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.358548387096776e-05,
+      "loss": 0.1669,
+      "step": 10678
+    },
+    {
+      "epoch": 0.170864,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.358387096774194e-05,
+      "loss": 0.216,
+      "step": 10679
+    },
+    {
+      "epoch": 0.17088,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.358225806451613e-05,
+      "loss": 0.1658,
+      "step": 10680
+    },
+    {
+      "epoch": 0.170896,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.358064516129033e-05,
+      "loss": 0.1747,
+      "step": 10681
+    },
+    {
+      "epoch": 0.170912,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.357903225806451e-05,
+      "loss": 0.2073,
+      "step": 10682
+    },
+    {
+      "epoch": 0.170928,
+      "grad_norm": 1.3671875,
+      "learning_rate": 8.357741935483871e-05,
+      "loss": 0.1968,
+      "step": 10683
+    },
+    {
+      "epoch": 0.170944,
+      "grad_norm": 1.25,
+      "learning_rate": 8.35758064516129e-05,
+      "loss": 0.1685,
+      "step": 10684
+    },
+    {
+      "epoch": 0.17096,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.35741935483871e-05,
+      "loss": 0.1618,
+      "step": 10685
+    },
+    {
+      "epoch": 0.170976,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.35725806451613e-05,
+      "loss": 0.1423,
+      "step": 10686
+    },
+    {
+      "epoch": 0.170992,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.357096774193548e-05,
+      "loss": 0.1647,
+      "step": 10687
+    },
+    {
+      "epoch": 0.171008,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.356935483870968e-05,
+      "loss": 0.1629,
+      "step": 10688
+    },
+    {
+      "epoch": 0.171024,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.356774193548388e-05,
+      "loss": 0.1753,
+      "step": 10689
+    },
+    {
+      "epoch": 0.17104,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.356612903225807e-05,
+      "loss": 0.1621,
+      "step": 10690
+    },
+    {
+      "epoch": 0.171056,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.356451612903227e-05,
+      "loss": 0.1503,
+      "step": 10691
+    },
+    {
+      "epoch": 0.171072,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.356290322580646e-05,
+      "loss": 0.1912,
+      "step": 10692
+    },
+    {
+      "epoch": 0.171088,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.356129032258065e-05,
+      "loss": 0.1848,
+      "step": 10693
+    },
+    {
+      "epoch": 0.171104,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.355967741935484e-05,
+      "loss": 0.1968,
+      "step": 10694
+    },
+    {
+      "epoch": 0.17112,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.355806451612903e-05,
+      "loss": 0.1532,
+      "step": 10695
+    },
+    {
+      "epoch": 0.171136,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.355645161290323e-05,
+      "loss": 0.1496,
+      "step": 10696
+    },
+    {
+      "epoch": 0.171152,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.355483870967741e-05,
+      "loss": 0.1959,
+      "step": 10697
+    },
+    {
+      "epoch": 0.171168,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.355322580645161e-05,
+      "loss": 0.1709,
+      "step": 10698
+    },
+    {
+      "epoch": 0.171184,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.355161290322581e-05,
+      "loss": 0.1282,
+      "step": 10699
+    },
+    {
+      "epoch": 0.1712,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.355000000000001e-05,
+      "loss": 0.1739,
+      "step": 10700
+    },
+    {
+      "epoch": 0.171216,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.35483870967742e-05,
+      "loss": 0.1291,
+      "step": 10701
+    },
+    {
+      "epoch": 0.171232,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.35467741935484e-05,
+      "loss": 0.1737,
+      "step": 10702
+    },
+    {
+      "epoch": 0.171248,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.354516129032258e-05,
+      "loss": 0.2048,
+      "step": 10703
+    },
+    {
+      "epoch": 0.171264,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.354354838709678e-05,
+      "loss": 0.1995,
+      "step": 10704
+    },
+    {
+      "epoch": 0.17128,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.354193548387097e-05,
+      "loss": 0.1261,
+      "step": 10705
+    },
+    {
+      "epoch": 0.171296,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.354032258064517e-05,
+      "loss": 0.1607,
+      "step": 10706
+    },
+    {
+      "epoch": 0.171312,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.353870967741935e-05,
+      "loss": 0.1814,
+      "step": 10707
+    },
+    {
+      "epoch": 0.171328,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.353709677419355e-05,
+      "loss": 0.1863,
+      "step": 10708
+    },
+    {
+      "epoch": 0.171344,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.353548387096774e-05,
+      "loss": 0.1558,
+      "step": 10709
+    },
+    {
+      "epoch": 0.17136,
+      "grad_norm": 1.3203125,
+      "learning_rate": 8.353387096774194e-05,
+      "loss": 0.1777,
+      "step": 10710
+    },
+    {
+      "epoch": 0.171376,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.353225806451614e-05,
+      "loss": 0.1753,
+      "step": 10711
+    },
+    {
+      "epoch": 0.171392,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.353064516129033e-05,
+      "loss": 0.1998,
+      "step": 10712
+    },
+    {
+      "epoch": 0.171408,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.352903225806453e-05,
+      "loss": 0.1761,
+      "step": 10713
+    },
+    {
+      "epoch": 0.171424,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.352741935483871e-05,
+      "loss": 0.1898,
+      "step": 10714
+    },
+    {
+      "epoch": 0.17144,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.352580645161291e-05,
+      "loss": 0.1882,
+      "step": 10715
+    },
+    {
+      "epoch": 0.171456,
+      "grad_norm": 1.25,
+      "learning_rate": 8.35241935483871e-05,
+      "loss": 0.1929,
+      "step": 10716
+    },
+    {
+      "epoch": 0.171472,
+      "grad_norm": 1.3671875,
+      "learning_rate": 8.35225806451613e-05,
+      "loss": 0.1816,
+      "step": 10717
+    },
+    {
+      "epoch": 0.171488,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.352096774193548e-05,
+      "loss": 0.1597,
+      "step": 10718
+    },
+    {
+      "epoch": 0.171504,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.351935483870968e-05,
+      "loss": 0.1601,
+      "step": 10719
+    },
+    {
+      "epoch": 0.17152,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.351774193548387e-05,
+      "loss": 0.1679,
+      "step": 10720
+    },
+    {
+      "epoch": 0.171536,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.351612903225807e-05,
+      "loss": 0.1546,
+      "step": 10721
+    },
+    {
+      "epoch": 0.171552,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.351451612903225e-05,
+      "loss": 0.2147,
+      "step": 10722
+    },
+    {
+      "epoch": 0.171568,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.351290322580645e-05,
+      "loss": 0.2524,
+      "step": 10723
+    },
+    {
+      "epoch": 0.171584,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.351129032258065e-05,
+      "loss": 0.1365,
+      "step": 10724
+    },
+    {
+      "epoch": 0.1716,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.350967741935485e-05,
+      "loss": 0.1485,
+      "step": 10725
+    },
+    {
+      "epoch": 0.171616,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.350806451612904e-05,
+      "loss": 0.2016,
+      "step": 10726
+    },
+    {
+      "epoch": 0.171632,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.350645161290323e-05,
+      "loss": 0.1516,
+      "step": 10727
+    },
+    {
+      "epoch": 0.171648,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.350483870967742e-05,
+      "loss": 0.1887,
+      "step": 10728
+    },
+    {
+      "epoch": 0.171664,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.350322580645161e-05,
+      "loss": 0.1635,
+      "step": 10729
+    },
+    {
+      "epoch": 0.17168,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.350161290322581e-05,
+      "loss": 0.1442,
+      "step": 10730
+    },
+    {
+      "epoch": 0.171696,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.35e-05,
+      "loss": 0.1817,
+      "step": 10731
+    },
+    {
+      "epoch": 0.171712,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.34983870967742e-05,
+      "loss": 0.1508,
+      "step": 10732
+    },
+    {
+      "epoch": 0.171728,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.349677419354838e-05,
+      "loss": 0.1992,
+      "step": 10733
+    },
+    {
+      "epoch": 0.171744,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.349516129032258e-05,
+      "loss": 0.1947,
+      "step": 10734
+    },
+    {
+      "epoch": 0.17176,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.349354838709678e-05,
+      "loss": 0.1488,
+      "step": 10735
+    },
+    {
+      "epoch": 0.171776,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.349193548387098e-05,
+      "loss": 0.1838,
+      "step": 10736
+    },
+    {
+      "epoch": 0.171792,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.349032258064517e-05,
+      "loss": 0.1885,
+      "step": 10737
+    },
+    {
+      "epoch": 0.171808,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.348870967741937e-05,
+      "loss": 0.1667,
+      "step": 10738
+    },
+    {
+      "epoch": 0.171824,
+      "grad_norm": 0.498046875,
+      "learning_rate": 8.348709677419355e-05,
+      "loss": 0.1473,
+      "step": 10739
+    },
+    {
+      "epoch": 0.17184,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.348548387096775e-05,
+      "loss": 0.1919,
+      "step": 10740
+    },
+    {
+      "epoch": 0.171856,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.348387096774194e-05,
+      "loss": 0.1827,
+      "step": 10741
+    },
+    {
+      "epoch": 0.171872,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.348225806451612e-05,
+      "loss": 0.1441,
+      "step": 10742
+    },
+    {
+      "epoch": 0.171888,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.348064516129032e-05,
+      "loss": 0.1869,
+      "step": 10743
+    },
+    {
+      "epoch": 0.171904,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.347903225806451e-05,
+      "loss": 0.2025,
+      "step": 10744
+    },
+    {
+      "epoch": 0.17192,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.347741935483871e-05,
+      "loss": 0.187,
+      "step": 10745
+    },
+    {
+      "epoch": 0.171936,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.347580645161291e-05,
+      "loss": 0.1933,
+      "step": 10746
+    },
+    {
+      "epoch": 0.171952,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.347419354838711e-05,
+      "loss": 0.1886,
+      "step": 10747
+    },
+    {
+      "epoch": 0.171968,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.34725806451613e-05,
+      "loss": 0.1631,
+      "step": 10748
+    },
+    {
+      "epoch": 0.171984,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.34709677419355e-05,
+      "loss": 0.1413,
+      "step": 10749
+    },
+    {
+      "epoch": 0.172,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.346935483870968e-05,
+      "loss": 0.1829,
+      "step": 10750
+    },
+    {
+      "epoch": 0.172016,
+      "grad_norm": 1.4375,
+      "learning_rate": 8.346774193548388e-05,
+      "loss": 0.1851,
+      "step": 10751
+    },
+    {
+      "epoch": 0.172032,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.346612903225807e-05,
+      "loss": 0.1616,
+      "step": 10752
+    },
+    {
+      "epoch": 0.172048,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.346451612903227e-05,
+      "loss": 0.2046,
+      "step": 10753
+    },
+    {
+      "epoch": 0.172064,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.346290322580645e-05,
+      "loss": 0.1848,
+      "step": 10754
+    },
+    {
+      "epoch": 0.17208,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.346129032258065e-05,
+      "loss": 0.1636,
+      "step": 10755
+    },
+    {
+      "epoch": 0.172096,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.345967741935484e-05,
+      "loss": 0.1676,
+      "step": 10756
+    },
+    {
+      "epoch": 0.172112,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.345806451612902e-05,
+      "loss": 0.1748,
+      "step": 10757
+    },
+    {
+      "epoch": 0.172128,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.345645161290322e-05,
+      "loss": 0.2084,
+      "step": 10758
+    },
+    {
+      "epoch": 0.172144,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.345483870967742e-05,
+      "loss": 0.1526,
+      "step": 10759
+    },
+    {
+      "epoch": 0.17216,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.345322580645162e-05,
+      "loss": 0.1509,
+      "step": 10760
+    },
+    {
+      "epoch": 0.172176,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.345161290322581e-05,
+      "loss": 0.1813,
+      "step": 10761
+    },
+    {
+      "epoch": 0.172192,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.345000000000001e-05,
+      "loss": 0.138,
+      "step": 10762
+    },
+    {
+      "epoch": 0.172208,
+      "grad_norm": 0.875,
+      "learning_rate": 8.34483870967742e-05,
+      "loss": 0.1662,
+      "step": 10763
+    },
+    {
+      "epoch": 0.172224,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.34467741935484e-05,
+      "loss": 0.1371,
+      "step": 10764
+    },
+    {
+      "epoch": 0.17224,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.344516129032258e-05,
+      "loss": 0.1515,
+      "step": 10765
+    },
+    {
+      "epoch": 0.172256,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.344354838709678e-05,
+      "loss": 0.1903,
+      "step": 10766
+    },
+    {
+      "epoch": 0.172272,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.344193548387097e-05,
+      "loss": 0.1575,
+      "step": 10767
+    },
+    {
+      "epoch": 0.172288,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.344032258064517e-05,
+      "loss": 0.1938,
+      "step": 10768
+    },
+    {
+      "epoch": 0.172304,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.343870967741935e-05,
+      "loss": 0.157,
+      "step": 10769
+    },
+    {
+      "epoch": 0.17232,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.343709677419355e-05,
+      "loss": 0.1659,
+      "step": 10770
+    },
+    {
+      "epoch": 0.172336,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.343548387096775e-05,
+      "loss": 0.1594,
+      "step": 10771
+    },
+    {
+      "epoch": 0.172352,
+      "grad_norm": 1.546875,
+      "learning_rate": 8.343387096774195e-05,
+      "loss": 0.2062,
+      "step": 10772
+    },
+    {
+      "epoch": 0.172368,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.343225806451614e-05,
+      "loss": 0.1351,
+      "step": 10773
+    },
+    {
+      "epoch": 0.172384,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.343064516129032e-05,
+      "loss": 0.1855,
+      "step": 10774
+    },
+    {
+      "epoch": 0.1724,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.342903225806452e-05,
+      "loss": 0.159,
+      "step": 10775
+    },
+    {
+      "epoch": 0.172416,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.342741935483871e-05,
+      "loss": 0.182,
+      "step": 10776
+    },
+    {
+      "epoch": 0.172432,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.342580645161291e-05,
+      "loss": 0.194,
+      "step": 10777
+    },
+    {
+      "epoch": 0.172448,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.34241935483871e-05,
+      "loss": 0.1496,
+      "step": 10778
+    },
+    {
+      "epoch": 0.172464,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.34225806451613e-05,
+      "loss": 0.2232,
+      "step": 10779
+    },
+    {
+      "epoch": 0.17248,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.342096774193548e-05,
+      "loss": 0.1964,
+      "step": 10780
+    },
+    {
+      "epoch": 0.172496,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.341935483870968e-05,
+      "loss": 0.1559,
+      "step": 10781
+    },
+    {
+      "epoch": 0.172512,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.341774193548388e-05,
+      "loss": 0.1689,
+      "step": 10782
+    },
+    {
+      "epoch": 0.172528,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.341612903225807e-05,
+      "loss": 0.1986,
+      "step": 10783
+    },
+    {
+      "epoch": 0.172544,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.341451612903227e-05,
+      "loss": 0.1936,
+      "step": 10784
+    },
+    {
+      "epoch": 0.17256,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.341290322580647e-05,
+      "loss": 0.1602,
+      "step": 10785
+    },
+    {
+      "epoch": 0.172576,
+      "grad_norm": 1.265625,
+      "learning_rate": 8.341129032258065e-05,
+      "loss": 0.1603,
+      "step": 10786
+    },
+    {
+      "epoch": 0.172592,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.340967741935485e-05,
+      "loss": 0.1623,
+      "step": 10787
+    },
+    {
+      "epoch": 0.172608,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.340806451612904e-05,
+      "loss": 0.1717,
+      "step": 10788
+    },
+    {
+      "epoch": 0.172624,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.340645161290322e-05,
+      "loss": 0.1711,
+      "step": 10789
+    },
+    {
+      "epoch": 0.17264,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.340483870967742e-05,
+      "loss": 0.161,
+      "step": 10790
+    },
+    {
+      "epoch": 0.172656,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.340322580645161e-05,
+      "loss": 0.1728,
+      "step": 10791
+    },
+    {
+      "epoch": 0.172672,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.340161290322581e-05,
+      "loss": 0.1595,
+      "step": 10792
+    },
+    {
+      "epoch": 0.172688,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.34e-05,
+      "loss": 0.1695,
+      "step": 10793
+    },
+    {
+      "epoch": 0.172704,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.33983870967742e-05,
+      "loss": 0.1826,
+      "step": 10794
+    },
+    {
+      "epoch": 0.17272,
+      "grad_norm": 1.125,
+      "learning_rate": 8.33967741935484e-05,
+      "loss": 0.1837,
+      "step": 10795
+    },
+    {
+      "epoch": 0.172736,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.339516129032259e-05,
+      "loss": 0.1586,
+      "step": 10796
+    },
+    {
+      "epoch": 0.172752,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.339354838709678e-05,
+      "loss": 0.1699,
+      "step": 10797
+    },
+    {
+      "epoch": 0.172768,
+      "grad_norm": 1.3359375,
+      "learning_rate": 8.339193548387098e-05,
+      "loss": 0.1844,
+      "step": 10798
+    },
+    {
+      "epoch": 0.172784,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.339032258064516e-05,
+      "loss": 0.1582,
+      "step": 10799
+    },
+    {
+      "epoch": 0.1728,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.338870967741936e-05,
+      "loss": 0.1668,
+      "step": 10800
+    },
+    {
+      "epoch": 0.172816,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.338709677419355e-05,
+      "loss": 0.1743,
+      "step": 10801
+    },
+    {
+      "epoch": 0.172832,
+      "grad_norm": 0.875,
+      "learning_rate": 8.338548387096775e-05,
+      "loss": 0.1629,
+      "step": 10802
+    },
+    {
+      "epoch": 0.172848,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.338387096774194e-05,
+      "loss": 0.1641,
+      "step": 10803
+    },
+    {
+      "epoch": 0.172864,
+      "grad_norm": 1.234375,
+      "learning_rate": 8.338225806451612e-05,
+      "loss": 0.1891,
+      "step": 10804
+    },
+    {
+      "epoch": 0.17288,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.338064516129032e-05,
+      "loss": 0.1829,
+      "step": 10805
+    },
+    {
+      "epoch": 0.172896,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.337903225806452e-05,
+      "loss": 0.1508,
+      "step": 10806
+    },
+    {
+      "epoch": 0.172912,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.337741935483872e-05,
+      "loss": 0.1857,
+      "step": 10807
+    },
+    {
+      "epoch": 0.172928,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.337580645161291e-05,
+      "loss": 0.1383,
+      "step": 10808
+    },
+    {
+      "epoch": 0.172944,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.337419354838711e-05,
+      "loss": 0.1839,
+      "step": 10809
+    },
+    {
+      "epoch": 0.17296,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.337258064516129e-05,
+      "loss": 0.1959,
+      "step": 10810
+    },
+    {
+      "epoch": 0.172976,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.337096774193549e-05,
+      "loss": 0.1586,
+      "step": 10811
+    },
+    {
+      "epoch": 0.172992,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.336935483870968e-05,
+      "loss": 0.1669,
+      "step": 10812
+    },
+    {
+      "epoch": 0.173008,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.336774193548388e-05,
+      "loss": 0.166,
+      "step": 10813
+    },
+    {
+      "epoch": 0.173024,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.336612903225806e-05,
+      "loss": 0.1279,
+      "step": 10814
+    },
+    {
+      "epoch": 0.17304,
+      "grad_norm": 1.2890625,
+      "learning_rate": 8.336451612903226e-05,
+      "loss": 0.1783,
+      "step": 10815
+    },
+    {
+      "epoch": 0.173056,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.336290322580645e-05,
+      "loss": 0.1938,
+      "step": 10816
+    },
+    {
+      "epoch": 0.173072,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.336129032258065e-05,
+      "loss": 0.1601,
+      "step": 10817
+    },
+    {
+      "epoch": 0.173088,
+      "grad_norm": 1.4140625,
+      "learning_rate": 8.335967741935484e-05,
+      "loss": 0.1819,
+      "step": 10818
+    },
+    {
+      "epoch": 0.173104,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.335806451612904e-05,
+      "loss": 0.1609,
+      "step": 10819
+    },
+    {
+      "epoch": 0.17312,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.335645161290324e-05,
+      "loss": 0.1788,
+      "step": 10820
+    },
+    {
+      "epoch": 0.173136,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.335483870967742e-05,
+      "loss": 0.1501,
+      "step": 10821
+    },
+    {
+      "epoch": 0.173152,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.335322580645162e-05,
+      "loss": 0.1769,
+      "step": 10822
+    },
+    {
+      "epoch": 0.173168,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.335161290322581e-05,
+      "loss": 0.1721,
+      "step": 10823
+    },
+    {
+      "epoch": 0.173184,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.335e-05,
+      "loss": 0.1772,
+      "step": 10824
+    },
+    {
+      "epoch": 0.1732,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.334838709677419e-05,
+      "loss": 0.1823,
+      "step": 10825
+    },
+    {
+      "epoch": 0.173216,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.334677419354839e-05,
+      "loss": 0.153,
+      "step": 10826
+    },
+    {
+      "epoch": 0.173232,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.334516129032258e-05,
+      "loss": 0.1663,
+      "step": 10827
+    },
+    {
+      "epoch": 0.173248,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.334354838709678e-05,
+      "loss": 0.1416,
+      "step": 10828
+    },
+    {
+      "epoch": 0.173264,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.334193548387096e-05,
+      "loss": 0.159,
+      "step": 10829
+    },
+    {
+      "epoch": 0.17328,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.334032258064516e-05,
+      "loss": 0.1786,
+      "step": 10830
+    },
+    {
+      "epoch": 0.173296,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.333870967741936e-05,
+      "loss": 0.1761,
+      "step": 10831
+    },
+    {
+      "epoch": 0.173312,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.333709677419356e-05,
+      "loss": 0.1854,
+      "step": 10832
+    },
+    {
+      "epoch": 0.173328,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.333548387096775e-05,
+      "loss": 0.1799,
+      "step": 10833
+    },
+    {
+      "epoch": 0.173344,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.333387096774195e-05,
+      "loss": 0.1664,
+      "step": 10834
+    },
+    {
+      "epoch": 0.17336,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.333225806451613e-05,
+      "loss": 0.2205,
+      "step": 10835
+    },
+    {
+      "epoch": 0.173376,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.333064516129032e-05,
+      "loss": 0.1741,
+      "step": 10836
+    },
+    {
+      "epoch": 0.173392,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.332903225806452e-05,
+      "loss": 0.1999,
+      "step": 10837
+    },
+    {
+      "epoch": 0.173408,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.33274193548387e-05,
+      "loss": 0.1665,
+      "step": 10838
+    },
+    {
+      "epoch": 0.173424,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.33258064516129e-05,
+      "loss": 0.2031,
+      "step": 10839
+    },
+    {
+      "epoch": 0.17344,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.332419354838709e-05,
+      "loss": 0.1868,
+      "step": 10840
+    },
+    {
+      "epoch": 0.173456,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.332258064516129e-05,
+      "loss": 0.1523,
+      "step": 10841
+    },
+    {
+      "epoch": 0.173472,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.332096774193549e-05,
+      "loss": 0.1797,
+      "step": 10842
+    },
+    {
+      "epoch": 0.173488,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.331935483870969e-05,
+      "loss": 0.207,
+      "step": 10843
+    },
+    {
+      "epoch": 0.173504,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.331774193548388e-05,
+      "loss": 0.1473,
+      "step": 10844
+    },
+    {
+      "epoch": 0.17352,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.331612903225808e-05,
+      "loss": 0.1672,
+      "step": 10845
+    },
+    {
+      "epoch": 0.173536,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.331451612903226e-05,
+      "loss": 0.1735,
+      "step": 10846
+    },
+    {
+      "epoch": 0.173552,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.331290322580646e-05,
+      "loss": 0.1823,
+      "step": 10847
+    },
+    {
+      "epoch": 0.173568,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.331129032258065e-05,
+      "loss": 0.1572,
+      "step": 10848
+    },
+    {
+      "epoch": 0.173584,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.330967741935485e-05,
+      "loss": 0.1406,
+      "step": 10849
+    },
+    {
+      "epoch": 0.1736,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.330806451612903e-05,
+      "loss": 0.1578,
+      "step": 10850
+    },
+    {
+      "epoch": 0.173616,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.330645161290322e-05,
+      "loss": 0.1779,
+      "step": 10851
+    },
+    {
+      "epoch": 0.173632,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.330483870967742e-05,
+      "loss": 0.1376,
+      "step": 10852
+    },
+    {
+      "epoch": 0.173648,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.33032258064516e-05,
+      "loss": 0.2204,
+      "step": 10853
+    },
+    {
+      "epoch": 0.173664,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.33016129032258e-05,
+      "loss": 0.1347,
+      "step": 10854
+    },
+    {
+      "epoch": 0.17368,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.33e-05,
+      "loss": 0.1947,
+      "step": 10855
+    },
+    {
+      "epoch": 0.173696,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.32983870967742e-05,
+      "loss": 0.1469,
+      "step": 10856
+    },
+    {
+      "epoch": 0.173712,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.329677419354839e-05,
+      "loss": 0.1429,
+      "step": 10857
+    },
+    {
+      "epoch": 0.173728,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.329516129032259e-05,
+      "loss": 0.183,
+      "step": 10858
+    },
+    {
+      "epoch": 0.173744,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.329354838709678e-05,
+      "loss": 0.1667,
+      "step": 10859
+    },
+    {
+      "epoch": 0.17376,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.329193548387098e-05,
+      "loss": 0.1846,
+      "step": 10860
+    },
+    {
+      "epoch": 0.173776,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.329032258064516e-05,
+      "loss": 0.1897,
+      "step": 10861
+    },
+    {
+      "epoch": 0.173792,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.328870967741936e-05,
+      "loss": 0.1493,
+      "step": 10862
+    },
+    {
+      "epoch": 0.173808,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.328709677419355e-05,
+      "loss": 0.2025,
+      "step": 10863
+    },
+    {
+      "epoch": 0.173824,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.328548387096775e-05,
+      "loss": 0.1373,
+      "step": 10864
+    },
+    {
+      "epoch": 0.17384,
+      "grad_norm": 1.34375,
+      "learning_rate": 8.328387096774193e-05,
+      "loss": 0.1949,
+      "step": 10865
+    },
+    {
+      "epoch": 0.173856,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.328225806451613e-05,
+      "loss": 0.1515,
+      "step": 10866
+    },
+    {
+      "epoch": 0.173872,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.328064516129033e-05,
+      "loss": 0.1733,
+      "step": 10867
+    },
+    {
+      "epoch": 0.173888,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.327903225806452e-05,
+      "loss": 0.1693,
+      "step": 10868
+    },
+    {
+      "epoch": 0.173904,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.327741935483872e-05,
+      "loss": 0.1535,
+      "step": 10869
+    },
+    {
+      "epoch": 0.17392,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.32758064516129e-05,
+      "loss": 0.1519,
+      "step": 10870
+    },
+    {
+      "epoch": 0.173936,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.32741935483871e-05,
+      "loss": 0.1821,
+      "step": 10871
+    },
+    {
+      "epoch": 0.173952,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.327258064516129e-05,
+      "loss": 0.1837,
+      "step": 10872
+    },
+    {
+      "epoch": 0.173968,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.327096774193549e-05,
+      "loss": 0.1658,
+      "step": 10873
+    },
+    {
+      "epoch": 0.173984,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.326935483870968e-05,
+      "loss": 0.1602,
+      "step": 10874
+    },
+    {
+      "epoch": 0.174,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.326774193548388e-05,
+      "loss": 0.2205,
+      "step": 10875
+    },
+    {
+      "epoch": 0.174016,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.326612903225806e-05,
+      "loss": 0.1806,
+      "step": 10876
+    },
+    {
+      "epoch": 0.174032,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.326451612903226e-05,
+      "loss": 0.1626,
+      "step": 10877
+    },
+    {
+      "epoch": 0.174048,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.326290322580645e-05,
+      "loss": 0.1395,
+      "step": 10878
+    },
+    {
+      "epoch": 0.174064,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.326129032258065e-05,
+      "loss": 0.1493,
+      "step": 10879
+    },
+    {
+      "epoch": 0.17408,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.325967741935485e-05,
+      "loss": 0.2002,
+      "step": 10880
+    },
+    {
+      "epoch": 0.174096,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.325806451612905e-05,
+      "loss": 0.1222,
+      "step": 10881
+    },
+    {
+      "epoch": 0.174112,
+      "grad_norm": 1.125,
+      "learning_rate": 8.325645161290323e-05,
+      "loss": 0.1918,
+      "step": 10882
+    },
+    {
+      "epoch": 0.174128,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.325483870967742e-05,
+      "loss": 0.1962,
+      "step": 10883
+    },
+    {
+      "epoch": 0.174144,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.325322580645162e-05,
+      "loss": 0.2016,
+      "step": 10884
+    },
+    {
+      "epoch": 0.17416,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.32516129032258e-05,
+      "loss": 0.1931,
+      "step": 10885
+    },
+    {
+      "epoch": 0.174176,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.325e-05,
+      "loss": 0.1606,
+      "step": 10886
+    },
+    {
+      "epoch": 0.174192,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.324838709677419e-05,
+      "loss": 0.1697,
+      "step": 10887
+    },
+    {
+      "epoch": 0.174208,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.324677419354839e-05,
+      "loss": 0.1254,
+      "step": 10888
+    },
+    {
+      "epoch": 0.174224,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.324516129032258e-05,
+      "loss": 0.1442,
+      "step": 10889
+    },
+    {
+      "epoch": 0.17424,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.324354838709678e-05,
+      "loss": 0.1514,
+      "step": 10890
+    },
+    {
+      "epoch": 0.174256,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.324193548387098e-05,
+      "loss": 0.1867,
+      "step": 10891
+    },
+    {
+      "epoch": 0.174272,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.324032258064517e-05,
+      "loss": 0.1994,
+      "step": 10892
+    },
+    {
+      "epoch": 0.174288,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.323870967741936e-05,
+      "loss": 0.1903,
+      "step": 10893
+    },
+    {
+      "epoch": 0.174304,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.323709677419356e-05,
+      "loss": 0.1214,
+      "step": 10894
+    },
+    {
+      "epoch": 0.17432,
+      "grad_norm": 1.0,
+      "learning_rate": 8.323548387096775e-05,
+      "loss": 0.1769,
+      "step": 10895
+    },
+    {
+      "epoch": 0.174336,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.323387096774195e-05,
+      "loss": 0.1665,
+      "step": 10896
+    },
+    {
+      "epoch": 0.174352,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.323225806451613e-05,
+      "loss": 0.1722,
+      "step": 10897
+    },
+    {
+      "epoch": 0.174368,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.323064516129032e-05,
+      "loss": 0.1866,
+      "step": 10898
+    },
+    {
+      "epoch": 0.174384,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.322903225806452e-05,
+      "loss": 0.1626,
+      "step": 10899
+    },
+    {
+      "epoch": 0.1744,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.32274193548387e-05,
+      "loss": 0.2086,
+      "step": 10900
+    },
+    {
+      "epoch": 0.174416,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.32258064516129e-05,
+      "loss": 0.1739,
+      "step": 10901
+    },
+    {
+      "epoch": 0.174432,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.32241935483871e-05,
+      "loss": 0.1587,
+      "step": 10902
+    },
+    {
+      "epoch": 0.174448,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.32225806451613e-05,
+      "loss": 0.1746,
+      "step": 10903
+    },
+    {
+      "epoch": 0.174464,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.322096774193549e-05,
+      "loss": 0.1253,
+      "step": 10904
+    },
+    {
+      "epoch": 0.17448,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.321935483870969e-05,
+      "loss": 0.2136,
+      "step": 10905
+    },
+    {
+      "epoch": 0.174496,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.321774193548387e-05,
+      "loss": 0.1549,
+      "step": 10906
+    },
+    {
+      "epoch": 0.174512,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.321612903225807e-05,
+      "loss": 0.1756,
+      "step": 10907
+    },
+    {
+      "epoch": 0.174528,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.321451612903226e-05,
+      "loss": 0.158,
+      "step": 10908
+    },
+    {
+      "epoch": 0.174544,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.321290322580646e-05,
+      "loss": 0.1626,
+      "step": 10909
+    },
+    {
+      "epoch": 0.17456,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.321129032258065e-05,
+      "loss": 0.1557,
+      "step": 10910
+    },
+    {
+      "epoch": 0.174576,
+      "grad_norm": 1.3046875,
+      "learning_rate": 8.320967741935485e-05,
+      "loss": 0.2365,
+      "step": 10911
+    },
+    {
+      "epoch": 0.174592,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.320806451612903e-05,
+      "loss": 0.1556,
+      "step": 10912
+    },
+    {
+      "epoch": 0.174608,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.320645161290322e-05,
+      "loss": 0.1689,
+      "step": 10913
+    },
+    {
+      "epoch": 0.174624,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.320483870967742e-05,
+      "loss": 0.198,
+      "step": 10914
+    },
+    {
+      "epoch": 0.17464,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.320322580645162e-05,
+      "loss": 0.178,
+      "step": 10915
+    },
+    {
+      "epoch": 0.174656,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.320161290322582e-05,
+      "loss": 0.1324,
+      "step": 10916
+    },
+    {
+      "epoch": 0.174672,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.32e-05,
+      "loss": 0.1954,
+      "step": 10917
+    },
+    {
+      "epoch": 0.174688,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.31983870967742e-05,
+      "loss": 0.1616,
+      "step": 10918
+    },
+    {
+      "epoch": 0.174704,
+      "grad_norm": 1.375,
+      "learning_rate": 8.319677419354839e-05,
+      "loss": 0.2221,
+      "step": 10919
+    },
+    {
+      "epoch": 0.17472,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.319516129032259e-05,
+      "loss": 0.1928,
+      "step": 10920
+    },
+    {
+      "epoch": 0.174736,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.319354838709677e-05,
+      "loss": 0.1707,
+      "step": 10921
+    },
+    {
+      "epoch": 0.174752,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.319193548387097e-05,
+      "loss": 0.1775,
+      "step": 10922
+    },
+    {
+      "epoch": 0.174768,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.319032258064516e-05,
+      "loss": 0.1773,
+      "step": 10923
+    },
+    {
+      "epoch": 0.174784,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.318870967741936e-05,
+      "loss": 0.1202,
+      "step": 10924
+    },
+    {
+      "epoch": 0.1748,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.318709677419355e-05,
+      "loss": 0.1888,
+      "step": 10925
+    },
+    {
+      "epoch": 0.174816,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.318548387096775e-05,
+      "loss": 0.1825,
+      "step": 10926
+    },
+    {
+      "epoch": 0.174832,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.318387096774195e-05,
+      "loss": 0.182,
+      "step": 10927
+    },
+    {
+      "epoch": 0.174848,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.318225806451613e-05,
+      "loss": 0.1741,
+      "step": 10928
+    },
+    {
+      "epoch": 0.174864,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.318064516129033e-05,
+      "loss": 0.1214,
+      "step": 10929
+    },
+    {
+      "epoch": 0.17488,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.317903225806452e-05,
+      "loss": 0.1584,
+      "step": 10930
+    },
+    {
+      "epoch": 0.174896,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.317741935483872e-05,
+      "loss": 0.1778,
+      "step": 10931
+    },
+    {
+      "epoch": 0.174912,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.31758064516129e-05,
+      "loss": 0.1514,
+      "step": 10932
+    },
+    {
+      "epoch": 0.174928,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.31741935483871e-05,
+      "loss": 0.1844,
+      "step": 10933
+    },
+    {
+      "epoch": 0.174944,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.317258064516129e-05,
+      "loss": 0.1713,
+      "step": 10934
+    },
+    {
+      "epoch": 0.17496,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.317096774193549e-05,
+      "loss": 0.1677,
+      "step": 10935
+    },
+    {
+      "epoch": 0.174976,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.316935483870967e-05,
+      "loss": 0.1857,
+      "step": 10936
+    },
+    {
+      "epoch": 0.174992,
+      "grad_norm": 1.359375,
+      "learning_rate": 8.316774193548387e-05,
+      "loss": 0.2426,
+      "step": 10937
+    },
+    {
+      "epoch": 0.175008,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.316612903225807e-05,
+      "loss": 0.1746,
+      "step": 10938
+    },
+    {
+      "epoch": 0.175024,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.316451612903226e-05,
+      "loss": 0.1934,
+      "step": 10939
+    },
+    {
+      "epoch": 0.17504,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.316290322580646e-05,
+      "loss": 0.1566,
+      "step": 10940
+    },
+    {
+      "epoch": 0.175056,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.316129032258066e-05,
+      "loss": 0.1756,
+      "step": 10941
+    },
+    {
+      "epoch": 0.175072,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.315967741935484e-05,
+      "loss": 0.1674,
+      "step": 10942
+    },
+    {
+      "epoch": 0.175088,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.315806451612904e-05,
+      "loss": 0.1351,
+      "step": 10943
+    },
+    {
+      "epoch": 0.175104,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.315645161290323e-05,
+      "loss": 0.1299,
+      "step": 10944
+    },
+    {
+      "epoch": 0.17512,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.315483870967742e-05,
+      "loss": 0.1771,
+      "step": 10945
+    },
+    {
+      "epoch": 0.175136,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.315322580645162e-05,
+      "loss": 0.1764,
+      "step": 10946
+    },
+    {
+      "epoch": 0.175152,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.31516129032258e-05,
+      "loss": 0.1507,
+      "step": 10947
+    },
+    {
+      "epoch": 0.175168,
+      "grad_norm": 0.95703125,
+      "learning_rate": 8.315e-05,
+      "loss": 0.1767,
+      "step": 10948
+    },
+    {
+      "epoch": 0.175184,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.314838709677419e-05,
+      "loss": 0.1566,
+      "step": 10949
+    },
+    {
+      "epoch": 0.1752,
+      "grad_norm": 1.265625,
+      "learning_rate": 8.314677419354839e-05,
+      "loss": 0.1973,
+      "step": 10950
+    },
+    {
+      "epoch": 0.175216,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.314516129032259e-05,
+      "loss": 0.1529,
+      "step": 10951
+    },
+    {
+      "epoch": 0.175232,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.314354838709679e-05,
+      "loss": 0.128,
+      "step": 10952
+    },
+    {
+      "epoch": 0.175248,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.314193548387097e-05,
+      "loss": 0.1486,
+      "step": 10953
+    },
+    {
+      "epoch": 0.175264,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.314032258064517e-05,
+      "loss": 0.1443,
+      "step": 10954
+    },
+    {
+      "epoch": 0.17528,
+      "grad_norm": 1.2890625,
+      "learning_rate": 8.313870967741936e-05,
+      "loss": 0.1861,
+      "step": 10955
+    },
+    {
+      "epoch": 0.175296,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.313709677419356e-05,
+      "loss": 0.1916,
+      "step": 10956
+    },
+    {
+      "epoch": 0.175312,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.313548387096774e-05,
+      "loss": 0.2246,
+      "step": 10957
+    },
+    {
+      "epoch": 0.175328,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.313387096774194e-05,
+      "loss": 0.2085,
+      "step": 10958
+    },
+    {
+      "epoch": 0.175344,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.313225806451613e-05,
+      "loss": 0.1835,
+      "step": 10959
+    },
+    {
+      "epoch": 0.17536,
+      "grad_norm": 1.265625,
+      "learning_rate": 8.313064516129032e-05,
+      "loss": 0.1976,
+      "step": 10960
+    },
+    {
+      "epoch": 0.175376,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.312903225806452e-05,
+      "loss": 0.1654,
+      "step": 10961
+    },
+    {
+      "epoch": 0.175392,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.312741935483872e-05,
+      "loss": 0.2187,
+      "step": 10962
+    },
+    {
+      "epoch": 0.175408,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.312580645161291e-05,
+      "loss": 0.1876,
+      "step": 10963
+    },
+    {
+      "epoch": 0.175424,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.31241935483871e-05,
+      "loss": 0.1577,
+      "step": 10964
+    },
+    {
+      "epoch": 0.17544,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.31225806451613e-05,
+      "loss": 0.1428,
+      "step": 10965
+    },
+    {
+      "epoch": 0.175456,
+      "grad_norm": 1.5390625,
+      "learning_rate": 8.312096774193549e-05,
+      "loss": 0.1878,
+      "step": 10966
+    },
+    {
+      "epoch": 0.175472,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.311935483870969e-05,
+      "loss": 0.1616,
+      "step": 10967
+    },
+    {
+      "epoch": 0.175488,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.311774193548387e-05,
+      "loss": 0.1926,
+      "step": 10968
+    },
+    {
+      "epoch": 0.175504,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.311612903225807e-05,
+      "loss": 0.1689,
+      "step": 10969
+    },
+    {
+      "epoch": 0.17552,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.311451612903226e-05,
+      "loss": 0.1712,
+      "step": 10970
+    },
+    {
+      "epoch": 0.175536,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.311290322580646e-05,
+      "loss": 0.1878,
+      "step": 10971
+    },
+    {
+      "epoch": 0.175552,
+      "grad_norm": 0.4921875,
+      "learning_rate": 8.311129032258064e-05,
+      "loss": 0.1555,
+      "step": 10972
+    },
+    {
+      "epoch": 0.175568,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.310967741935484e-05,
+      "loss": 0.1504,
+      "step": 10973
+    },
+    {
+      "epoch": 0.175584,
+      "grad_norm": 1.9375,
+      "learning_rate": 8.310806451612903e-05,
+      "loss": 0.2228,
+      "step": 10974
+    },
+    {
+      "epoch": 0.1756,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.310645161290323e-05,
+      "loss": 0.1539,
+      "step": 10975
+    },
+    {
+      "epoch": 0.175616,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.310483870967743e-05,
+      "loss": 0.2139,
+      "step": 10976
+    },
+    {
+      "epoch": 0.175632,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.310322580645161e-05,
+      "loss": 0.2088,
+      "step": 10977
+    },
+    {
+      "epoch": 0.175648,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.310161290322581e-05,
+      "loss": 0.1181,
+      "step": 10978
+    },
+    {
+      "epoch": 0.175664,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.31e-05,
+      "loss": 0.1723,
+      "step": 10979
+    },
+    {
+      "epoch": 0.17568,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.30983870967742e-05,
+      "loss": 0.195,
+      "step": 10980
+    },
+    {
+      "epoch": 0.175696,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.309677419354839e-05,
+      "loss": 0.1624,
+      "step": 10981
+    },
+    {
+      "epoch": 0.175712,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.309516129032259e-05,
+      "loss": 0.1839,
+      "step": 10982
+    },
+    {
+      "epoch": 0.175728,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.309354838709677e-05,
+      "loss": 0.1663,
+      "step": 10983
+    },
+    {
+      "epoch": 0.175744,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.309193548387097e-05,
+      "loss": 0.1979,
+      "step": 10984
+    },
+    {
+      "epoch": 0.17576,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.309032258064516e-05,
+      "loss": 0.1877,
+      "step": 10985
+    },
+    {
+      "epoch": 0.175776,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.308870967741936e-05,
+      "loss": 0.1745,
+      "step": 10986
+    },
+    {
+      "epoch": 0.175792,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.308709677419356e-05,
+      "loss": 0.1756,
+      "step": 10987
+    },
+    {
+      "epoch": 0.175808,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.308548387096776e-05,
+      "loss": 0.1744,
+      "step": 10988
+    },
+    {
+      "epoch": 0.175824,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.308387096774194e-05,
+      "loss": 0.1795,
+      "step": 10989
+    },
+    {
+      "epoch": 0.17584,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.308225806451613e-05,
+      "loss": 0.1471,
+      "step": 10990
+    },
+    {
+      "epoch": 0.175856,
+      "grad_norm": 1.0,
+      "learning_rate": 8.308064516129033e-05,
+      "loss": 0.1636,
+      "step": 10991
+    },
+    {
+      "epoch": 0.175872,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.307903225806451e-05,
+      "loss": 0.1635,
+      "step": 10992
+    },
+    {
+      "epoch": 0.175888,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.307741935483871e-05,
+      "loss": 0.1431,
+      "step": 10993
+    },
+    {
+      "epoch": 0.175904,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.30758064516129e-05,
+      "loss": 0.1753,
+      "step": 10994
+    },
+    {
+      "epoch": 0.17592,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.30741935483871e-05,
+      "loss": 0.1646,
+      "step": 10995
+    },
+    {
+      "epoch": 0.175936,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.307258064516129e-05,
+      "loss": 0.189,
+      "step": 10996
+    },
+    {
+      "epoch": 0.175952,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.307096774193549e-05,
+      "loss": 0.1847,
+      "step": 10997
+    },
+    {
+      "epoch": 0.175968,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.306935483870969e-05,
+      "loss": 0.1904,
+      "step": 10998
+    },
+    {
+      "epoch": 0.175984,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.306774193548388e-05,
+      "loss": 0.1722,
+      "step": 10999
+    },
+    {
+      "epoch": 0.176,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.306612903225807e-05,
+      "loss": 0.1695,
+      "step": 11000
+    },
+    {
+      "epoch": 0.176016,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.306451612903227e-05,
+      "loss": 0.1797,
+      "step": 11001
+    },
+    {
+      "epoch": 0.176032,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.306290322580646e-05,
+      "loss": 0.1682,
+      "step": 11002
+    },
+    {
+      "epoch": 0.176048,
+      "grad_norm": 0.400390625,
+      "learning_rate": 8.306129032258066e-05,
+      "loss": 0.1287,
+      "step": 11003
+    },
+    {
+      "epoch": 0.176064,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.305967741935484e-05,
+      "loss": 0.1563,
+      "step": 11004
+    },
+    {
+      "epoch": 0.17608,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.305806451612904e-05,
+      "loss": 0.1428,
+      "step": 11005
+    },
+    {
+      "epoch": 0.176096,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.305645161290323e-05,
+      "loss": 0.1839,
+      "step": 11006
+    },
+    {
+      "epoch": 0.176112,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.305483870967741e-05,
+      "loss": 0.1624,
+      "step": 11007
+    },
+    {
+      "epoch": 0.176128,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.305322580645161e-05,
+      "loss": 0.1438,
+      "step": 11008
+    },
+    {
+      "epoch": 0.176144,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.30516129032258e-05,
+      "loss": 0.1482,
+      "step": 11009
+    },
+    {
+      "epoch": 0.17616,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.305e-05,
+      "loss": 0.1849,
+      "step": 11010
+    },
+    {
+      "epoch": 0.176176,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.30483870967742e-05,
+      "loss": 0.1674,
+      "step": 11011
+    },
+    {
+      "epoch": 0.176192,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.30467741935484e-05,
+      "loss": 0.1493,
+      "step": 11012
+    },
+    {
+      "epoch": 0.176208,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.304516129032258e-05,
+      "loss": 0.1669,
+      "step": 11013
+    },
+    {
+      "epoch": 0.176224,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.304354838709678e-05,
+      "loss": 0.1706,
+      "step": 11014
+    },
+    {
+      "epoch": 0.17624,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.304193548387097e-05,
+      "loss": 0.1736,
+      "step": 11015
+    },
+    {
+      "epoch": 0.176256,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.304032258064517e-05,
+      "loss": 0.1479,
+      "step": 11016
+    },
+    {
+      "epoch": 0.176272,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.303870967741936e-05,
+      "loss": 0.2007,
+      "step": 11017
+    },
+    {
+      "epoch": 0.176288,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.303709677419356e-05,
+      "loss": 0.1604,
+      "step": 11018
+    },
+    {
+      "epoch": 0.176304,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.303548387096774e-05,
+      "loss": 0.2038,
+      "step": 11019
+    },
+    {
+      "epoch": 0.17632,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.303387096774194e-05,
+      "loss": 0.1933,
+      "step": 11020
+    },
+    {
+      "epoch": 0.176336,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.303225806451613e-05,
+      "loss": 0.191,
+      "step": 11021
+    },
+    {
+      "epoch": 0.176352,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.303064516129033e-05,
+      "loss": 0.1233,
+      "step": 11022
+    },
+    {
+      "epoch": 0.176368,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.302903225806453e-05,
+      "loss": 0.2136,
+      "step": 11023
+    },
+    {
+      "epoch": 0.176384,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.302741935483871e-05,
+      "loss": 0.1392,
+      "step": 11024
+    },
+    {
+      "epoch": 0.1764,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.302580645161291e-05,
+      "loss": 0.1882,
+      "step": 11025
+    },
+    {
+      "epoch": 0.176416,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.30241935483871e-05,
+      "loss": 0.1574,
+      "step": 11026
+    },
+    {
+      "epoch": 0.176432,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.30225806451613e-05,
+      "loss": 0.1887,
+      "step": 11027
+    },
+    {
+      "epoch": 0.176448,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.302096774193548e-05,
+      "loss": 0.1797,
+      "step": 11028
+    },
+    {
+      "epoch": 0.176464,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.301935483870968e-05,
+      "loss": 0.1939,
+      "step": 11029
+    },
+    {
+      "epoch": 0.17648,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.301774193548387e-05,
+      "loss": 0.1586,
+      "step": 11030
+    },
+    {
+      "epoch": 0.176496,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.301612903225807e-05,
+      "loss": 0.1359,
+      "step": 11031
+    },
+    {
+      "epoch": 0.176512,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.301451612903226e-05,
+      "loss": 0.1916,
+      "step": 11032
+    },
+    {
+      "epoch": 0.176528,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.301290322580646e-05,
+      "loss": 0.1484,
+      "step": 11033
+    },
+    {
+      "epoch": 0.176544,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.301129032258065e-05,
+      "loss": 0.1685,
+      "step": 11034
+    },
+    {
+      "epoch": 0.17656,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.300967741935484e-05,
+      "loss": 0.155,
+      "step": 11035
+    },
+    {
+      "epoch": 0.176576,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.300806451612904e-05,
+      "loss": 0.1865,
+      "step": 11036
+    },
+    {
+      "epoch": 0.176592,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.300645161290323e-05,
+      "loss": 0.1715,
+      "step": 11037
+    },
+    {
+      "epoch": 0.176608,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.300483870967743e-05,
+      "loss": 0.1543,
+      "step": 11038
+    },
+    {
+      "epoch": 0.176624,
+      "grad_norm": 1.2890625,
+      "learning_rate": 8.300322580645161e-05,
+      "loss": 0.2415,
+      "step": 11039
+    },
+    {
+      "epoch": 0.17664,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.300161290322581e-05,
+      "loss": 0.1866,
+      "step": 11040
+    },
+    {
+      "epoch": 0.176656,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.3e-05,
+      "loss": 0.1666,
+      "step": 11041
+    },
+    {
+      "epoch": 0.176672,
+      "grad_norm": 1.8203125,
+      "learning_rate": 8.29983870967742e-05,
+      "loss": 0.1549,
+      "step": 11042
+    },
+    {
+      "epoch": 0.176688,
+      "grad_norm": 1.2734375,
+      "learning_rate": 8.299677419354838e-05,
+      "loss": 0.1865,
+      "step": 11043
+    },
+    {
+      "epoch": 0.176704,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.299516129032258e-05,
+      "loss": 0.1948,
+      "step": 11044
+    },
+    {
+      "epoch": 0.17672,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.299354838709677e-05,
+      "loss": 0.2148,
+      "step": 11045
+    },
+    {
+      "epoch": 0.176736,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.299193548387097e-05,
+      "loss": 0.1957,
+      "step": 11046
+    },
+    {
+      "epoch": 0.176752,
+      "grad_norm": 0.875,
+      "learning_rate": 8.299032258064517e-05,
+      "loss": 0.1581,
+      "step": 11047
+    },
+    {
+      "epoch": 0.176768,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.298870967741937e-05,
+      "loss": 0.1816,
+      "step": 11048
+    },
+    {
+      "epoch": 0.176784,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.298709677419355e-05,
+      "loss": 0.1445,
+      "step": 11049
+    },
+    {
+      "epoch": 0.1768,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.298548387096775e-05,
+      "loss": 0.1659,
+      "step": 11050
+    },
+    {
+      "epoch": 0.176816,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.298387096774194e-05,
+      "loss": 0.1629,
+      "step": 11051
+    },
+    {
+      "epoch": 0.176832,
+      "grad_norm": 1.5,
+      "learning_rate": 8.298225806451614e-05,
+      "loss": 0.2336,
+      "step": 11052
+    },
+    {
+      "epoch": 0.176848,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.298064516129033e-05,
+      "loss": 0.1625,
+      "step": 11053
+    },
+    {
+      "epoch": 0.176864,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.297903225806451e-05,
+      "loss": 0.1358,
+      "step": 11054
+    },
+    {
+      "epoch": 0.17688,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.297741935483871e-05,
+      "loss": 0.1768,
+      "step": 11055
+    },
+    {
+      "epoch": 0.176896,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.29758064516129e-05,
+      "loss": 0.1917,
+      "step": 11056
+    },
+    {
+      "epoch": 0.176912,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.29741935483871e-05,
+      "loss": 0.1574,
+      "step": 11057
+    },
+    {
+      "epoch": 0.176928,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.29725806451613e-05,
+      "loss": 0.1741,
+      "step": 11058
+    },
+    {
+      "epoch": 0.176944,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.29709677419355e-05,
+      "loss": 0.1363,
+      "step": 11059
+    },
+    {
+      "epoch": 0.17696,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.296935483870968e-05,
+      "loss": 0.17,
+      "step": 11060
+    },
+    {
+      "epoch": 0.176976,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.296774193548388e-05,
+      "loss": 0.1553,
+      "step": 11061
+    },
+    {
+      "epoch": 0.176992,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.296612903225807e-05,
+      "loss": 0.1515,
+      "step": 11062
+    },
+    {
+      "epoch": 0.177008,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.296451612903227e-05,
+      "loss": 0.1793,
+      "step": 11063
+    },
+    {
+      "epoch": 0.177024,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.296290322580645e-05,
+      "loss": 0.1364,
+      "step": 11064
+    },
+    {
+      "epoch": 0.17704,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.296129032258065e-05,
+      "loss": 0.1641,
+      "step": 11065
+    },
+    {
+      "epoch": 0.177056,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.295967741935484e-05,
+      "loss": 0.1726,
+      "step": 11066
+    },
+    {
+      "epoch": 0.177072,
+      "grad_norm": 0.75,
+      "learning_rate": 8.295806451612904e-05,
+      "loss": 0.178,
+      "step": 11067
+    },
+    {
+      "epoch": 0.177088,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.295645161290323e-05,
+      "loss": 0.1309,
+      "step": 11068
+    },
+    {
+      "epoch": 0.177104,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.295483870967741e-05,
+      "loss": 0.1651,
+      "step": 11069
+    },
+    {
+      "epoch": 0.17712,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.295322580645161e-05,
+      "loss": 0.1471,
+      "step": 11070
+    },
+    {
+      "epoch": 0.177136,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.295161290322581e-05,
+      "loss": 0.1622,
+      "step": 11071
+    },
+    {
+      "epoch": 0.177152,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.295000000000001e-05,
+      "loss": 0.1914,
+      "step": 11072
+    },
+    {
+      "epoch": 0.177168,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.29483870967742e-05,
+      "loss": 0.2082,
+      "step": 11073
+    },
+    {
+      "epoch": 0.177184,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.29467741935484e-05,
+      "loss": 0.1259,
+      "step": 11074
+    },
+    {
+      "epoch": 0.1772,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.294516129032258e-05,
+      "loss": 0.1467,
+      "step": 11075
+    },
+    {
+      "epoch": 0.177216,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.294354838709678e-05,
+      "loss": 0.207,
+      "step": 11076
+    },
+    {
+      "epoch": 0.177232,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.294193548387097e-05,
+      "loss": 0.1806,
+      "step": 11077
+    },
+    {
+      "epoch": 0.177248,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.294032258064517e-05,
+      "loss": 0.163,
+      "step": 11078
+    },
+    {
+      "epoch": 0.177264,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.293870967741935e-05,
+      "loss": 0.1605,
+      "step": 11079
+    },
+    {
+      "epoch": 0.17728,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.293709677419355e-05,
+      "loss": 0.1561,
+      "step": 11080
+    },
+    {
+      "epoch": 0.177296,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.293548387096774e-05,
+      "loss": 0.1597,
+      "step": 11081
+    },
+    {
+      "epoch": 0.177312,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.293387096774194e-05,
+      "loss": 0.145,
+      "step": 11082
+    },
+    {
+      "epoch": 0.177328,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.293225806451614e-05,
+      "loss": 0.1834,
+      "step": 11083
+    },
+    {
+      "epoch": 0.177344,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.293064516129032e-05,
+      "loss": 0.1528,
+      "step": 11084
+    },
+    {
+      "epoch": 0.17736,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.292903225806452e-05,
+      "loss": 0.1665,
+      "step": 11085
+    },
+    {
+      "epoch": 0.177376,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.292741935483871e-05,
+      "loss": 0.142,
+      "step": 11086
+    },
+    {
+      "epoch": 0.177392,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.292580645161291e-05,
+      "loss": 0.1822,
+      "step": 11087
+    },
+    {
+      "epoch": 0.177408,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.29241935483871e-05,
+      "loss": 0.1831,
+      "step": 11088
+    },
+    {
+      "epoch": 0.177424,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.29225806451613e-05,
+      "loss": 0.151,
+      "step": 11089
+    },
+    {
+      "epoch": 0.17744,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.292096774193548e-05,
+      "loss": 0.1432,
+      "step": 11090
+    },
+    {
+      "epoch": 0.177456,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.291935483870968e-05,
+      "loss": 0.1889,
+      "step": 11091
+    },
+    {
+      "epoch": 0.177472,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.291774193548387e-05,
+      "loss": 0.2,
+      "step": 11092
+    },
+    {
+      "epoch": 0.177488,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.291612903225807e-05,
+      "loss": 0.1861,
+      "step": 11093
+    },
+    {
+      "epoch": 0.177504,
+      "grad_norm": 0.75,
+      "learning_rate": 8.291451612903227e-05,
+      "loss": 0.1639,
+      "step": 11094
+    },
+    {
+      "epoch": 0.17752,
+      "grad_norm": 1.3828125,
+      "learning_rate": 8.291290322580647e-05,
+      "loss": 0.1965,
+      "step": 11095
+    },
+    {
+      "epoch": 0.177536,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.291129032258065e-05,
+      "loss": 0.1205,
+      "step": 11096
+    },
+    {
+      "epoch": 0.177552,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.290967741935485e-05,
+      "loss": 0.212,
+      "step": 11097
+    },
+    {
+      "epoch": 0.177568,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.290806451612904e-05,
+      "loss": 0.1621,
+      "step": 11098
+    },
+    {
+      "epoch": 0.177584,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.290645161290322e-05,
+      "loss": 0.1886,
+      "step": 11099
+    },
+    {
+      "epoch": 0.1776,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.290483870967742e-05,
+      "loss": 0.1706,
+      "step": 11100
+    },
+    {
+      "epoch": 0.177616,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.290322580645161e-05,
+      "loss": 0.1471,
+      "step": 11101
+    },
+    {
+      "epoch": 0.177632,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.290161290322581e-05,
+      "loss": 0.1656,
+      "step": 11102
+    },
+    {
+      "epoch": 0.177648,
+      "grad_norm": 1.3359375,
+      "learning_rate": 8.29e-05,
+      "loss": 0.1816,
+      "step": 11103
+    },
+    {
+      "epoch": 0.177664,
+      "grad_norm": 1.2734375,
+      "learning_rate": 8.28983870967742e-05,
+      "loss": 0.156,
+      "step": 11104
+    },
+    {
+      "epoch": 0.17768,
+      "grad_norm": 1.0,
+      "learning_rate": 8.289677419354838e-05,
+      "loss": 0.1777,
+      "step": 11105
+    },
+    {
+      "epoch": 0.177696,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.289516129032258e-05,
+      "loss": 0.1211,
+      "step": 11106
+    },
+    {
+      "epoch": 0.177712,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.289354838709678e-05,
+      "loss": 0.1726,
+      "step": 11107
+    },
+    {
+      "epoch": 0.177728,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.289193548387098e-05,
+      "loss": 0.1731,
+      "step": 11108
+    },
+    {
+      "epoch": 0.177744,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.289032258064517e-05,
+      "loss": 0.1931,
+      "step": 11109
+    },
+    {
+      "epoch": 0.17776,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.288870967741937e-05,
+      "loss": 0.1832,
+      "step": 11110
+    },
+    {
+      "epoch": 0.177776,
+      "grad_norm": 1.40625,
+      "learning_rate": 8.288709677419355e-05,
+      "loss": 0.1527,
+      "step": 11111
+    },
+    {
+      "epoch": 0.177792,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.288548387096775e-05,
+      "loss": 0.1763,
+      "step": 11112
+    },
+    {
+      "epoch": 0.177808,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.288387096774194e-05,
+      "loss": 0.1448,
+      "step": 11113
+    },
+    {
+      "epoch": 0.177824,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.288225806451614e-05,
+      "loss": 0.1513,
+      "step": 11114
+    },
+    {
+      "epoch": 0.17784,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.288064516129032e-05,
+      "loss": 0.2284,
+      "step": 11115
+    },
+    {
+      "epoch": 0.177856,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.287903225806451e-05,
+      "loss": 0.2118,
+      "step": 11116
+    },
+    {
+      "epoch": 0.177872,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.287741935483871e-05,
+      "loss": 0.1555,
+      "step": 11117
+    },
+    {
+      "epoch": 0.177888,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.287580645161291e-05,
+      "loss": 0.1921,
+      "step": 11118
+    },
+    {
+      "epoch": 0.177904,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.287419354838711e-05,
+      "loss": 0.1525,
+      "step": 11119
+    },
+    {
+      "epoch": 0.17792,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.28725806451613e-05,
+      "loss": 0.1359,
+      "step": 11120
+    },
+    {
+      "epoch": 0.177936,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.28709677419355e-05,
+      "loss": 0.1603,
+      "step": 11121
+    },
+    {
+      "epoch": 0.177952,
+      "grad_norm": 0.75,
+      "learning_rate": 8.286935483870968e-05,
+      "loss": 0.1807,
+      "step": 11122
+    },
+    {
+      "epoch": 0.177968,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.286774193548388e-05,
+      "loss": 0.2083,
+      "step": 11123
+    },
+    {
+      "epoch": 0.177984,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.286612903225807e-05,
+      "loss": 0.18,
+      "step": 11124
+    },
+    {
+      "epoch": 0.178,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.286451612903227e-05,
+      "loss": 0.152,
+      "step": 11125
+    },
+    {
+      "epoch": 0.178016,
+      "grad_norm": 0.5,
+      "learning_rate": 8.286290322580645e-05,
+      "loss": 0.1605,
+      "step": 11126
+    },
+    {
+      "epoch": 0.178032,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.286129032258065e-05,
+      "loss": 0.1917,
+      "step": 11127
+    },
+    {
+      "epoch": 0.178048,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.285967741935484e-05,
+      "loss": 0.1792,
+      "step": 11128
+    },
+    {
+      "epoch": 0.178064,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.285806451612904e-05,
+      "loss": 0.1671,
+      "step": 11129
+    },
+    {
+      "epoch": 0.17808,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.285645161290322e-05,
+      "loss": 0.1663,
+      "step": 11130
+    },
+    {
+      "epoch": 0.178096,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.285483870967742e-05,
+      "loss": 0.1436,
+      "step": 11131
+    },
+    {
+      "epoch": 0.178112,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.285322580645162e-05,
+      "loss": 0.1621,
+      "step": 11132
+    },
+    {
+      "epoch": 0.178128,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.285161290322581e-05,
+      "loss": 0.2174,
+      "step": 11133
+    },
+    {
+      "epoch": 0.178144,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.285000000000001e-05,
+      "loss": 0.167,
+      "step": 11134
+    },
+    {
+      "epoch": 0.17816,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.28483870967742e-05,
+      "loss": 0.1472,
+      "step": 11135
+    },
+    {
+      "epoch": 0.178176,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.28467741935484e-05,
+      "loss": 0.168,
+      "step": 11136
+    },
+    {
+      "epoch": 0.178192,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.284516129032258e-05,
+      "loss": 0.161,
+      "step": 11137
+    },
+    {
+      "epoch": 0.178208,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.284354838709678e-05,
+      "loss": 0.1761,
+      "step": 11138
+    },
+    {
+      "epoch": 0.178224,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.284193548387097e-05,
+      "loss": 0.1798,
+      "step": 11139
+    },
+    {
+      "epoch": 0.17824,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.284032258064517e-05,
+      "loss": 0.135,
+      "step": 11140
+    },
+    {
+      "epoch": 0.178256,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.283870967741935e-05,
+      "loss": 0.1522,
+      "step": 11141
+    },
+    {
+      "epoch": 0.178272,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.283709677419355e-05,
+      "loss": 0.2054,
+      "step": 11142
+    },
+    {
+      "epoch": 0.178288,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.283548387096775e-05,
+      "loss": 0.1662,
+      "step": 11143
+    },
+    {
+      "epoch": 0.178304,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.283387096774195e-05,
+      "loss": 0.1324,
+      "step": 11144
+    },
+    {
+      "epoch": 0.17832,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.283225806451614e-05,
+      "loss": 0.1682,
+      "step": 11145
+    },
+    {
+      "epoch": 0.178336,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.283064516129032e-05,
+      "loss": 0.159,
+      "step": 11146
+    },
+    {
+      "epoch": 0.178352,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.282903225806452e-05,
+      "loss": 0.176,
+      "step": 11147
+    },
+    {
+      "epoch": 0.178368,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.282741935483871e-05,
+      "loss": 0.1637,
+      "step": 11148
+    },
+    {
+      "epoch": 0.178384,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.282580645161291e-05,
+      "loss": 0.1504,
+      "step": 11149
+    },
+    {
+      "epoch": 0.1784,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.28241935483871e-05,
+      "loss": 0.158,
+      "step": 11150
+    },
+    {
+      "epoch": 0.178416,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.28225806451613e-05,
+      "loss": 0.1671,
+      "step": 11151
+    },
+    {
+      "epoch": 0.178432,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.282096774193548e-05,
+      "loss": 0.183,
+      "step": 11152
+    },
+    {
+      "epoch": 0.178448,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.281935483870968e-05,
+      "loss": 0.179,
+      "step": 11153
+    },
+    {
+      "epoch": 0.178464,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.281774193548388e-05,
+      "loss": 0.1928,
+      "step": 11154
+    },
+    {
+      "epoch": 0.17848,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.281612903225808e-05,
+      "loss": 0.1329,
+      "step": 11155
+    },
+    {
+      "epoch": 0.178496,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.281451612903226e-05,
+      "loss": 0.1837,
+      "step": 11156
+    },
+    {
+      "epoch": 0.178512,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.281290322580646e-05,
+      "loss": 0.1887,
+      "step": 11157
+    },
+    {
+      "epoch": 0.178528,
+      "grad_norm": 1.2890625,
+      "learning_rate": 8.281129032258065e-05,
+      "loss": 0.1677,
+      "step": 11158
+    },
+    {
+      "epoch": 0.178544,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.280967741935485e-05,
+      "loss": 0.1741,
+      "step": 11159
+    },
+    {
+      "epoch": 0.17856,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.280806451612904e-05,
+      "loss": 0.1551,
+      "step": 11160
+    },
+    {
+      "epoch": 0.178576,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.280645161290324e-05,
+      "loss": 0.125,
+      "step": 11161
+    },
+    {
+      "epoch": 0.178592,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.280483870967742e-05,
+      "loss": 0.172,
+      "step": 11162
+    },
+    {
+      "epoch": 0.178608,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.280322580645161e-05,
+      "loss": 0.1735,
+      "step": 11163
+    },
+    {
+      "epoch": 0.178624,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.280161290322581e-05,
+      "loss": 0.1525,
+      "step": 11164
+    },
+    {
+      "epoch": 0.17864,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.28e-05,
+      "loss": 0.1634,
+      "step": 11165
+    },
+    {
+      "epoch": 0.178656,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.279838709677419e-05,
+      "loss": 0.1623,
+      "step": 11166
+    },
+    {
+      "epoch": 0.178672,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.279677419354839e-05,
+      "loss": 0.1322,
+      "step": 11167
+    },
+    {
+      "epoch": 0.178688,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.279516129032259e-05,
+      "loss": 0.1401,
+      "step": 11168
+    },
+    {
+      "epoch": 0.178704,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.279354838709678e-05,
+      "loss": 0.1778,
+      "step": 11169
+    },
+    {
+      "epoch": 0.17872,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.279193548387098e-05,
+      "loss": 0.1695,
+      "step": 11170
+    },
+    {
+      "epoch": 0.178736,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.279032258064516e-05,
+      "loss": 0.1829,
+      "step": 11171
+    },
+    {
+      "epoch": 0.178752,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.278870967741936e-05,
+      "loss": 0.1862,
+      "step": 11172
+    },
+    {
+      "epoch": 0.178768,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.278709677419355e-05,
+      "loss": 0.1677,
+      "step": 11173
+    },
+    {
+      "epoch": 0.178784,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.278548387096775e-05,
+      "loss": 0.1581,
+      "step": 11174
+    },
+    {
+      "epoch": 0.1788,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.278387096774194e-05,
+      "loss": 0.1949,
+      "step": 11175
+    },
+    {
+      "epoch": 0.178816,
+      "grad_norm": 1.4140625,
+      "learning_rate": 8.278225806451614e-05,
+      "loss": 0.1846,
+      "step": 11176
+    },
+    {
+      "epoch": 0.178832,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.278064516129032e-05,
+      "loss": 0.1605,
+      "step": 11177
+    },
+    {
+      "epoch": 0.178848,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.277903225806452e-05,
+      "loss": 0.1279,
+      "step": 11178
+    },
+    {
+      "epoch": 0.178864,
+      "grad_norm": 1.40625,
+      "learning_rate": 8.277741935483872e-05,
+      "loss": 0.1997,
+      "step": 11179
+    },
+    {
+      "epoch": 0.17888,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.27758064516129e-05,
+      "loss": 0.1714,
+      "step": 11180
+    },
+    {
+      "epoch": 0.178896,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.27741935483871e-05,
+      "loss": 0.2142,
+      "step": 11181
+    },
+    {
+      "epoch": 0.178912,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.277258064516129e-05,
+      "loss": 0.18,
+      "step": 11182
+    },
+    {
+      "epoch": 0.178928,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.277096774193549e-05,
+      "loss": 0.2231,
+      "step": 11183
+    },
+    {
+      "epoch": 0.178944,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.276935483870968e-05,
+      "loss": 0.2158,
+      "step": 11184
+    },
+    {
+      "epoch": 0.17896,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.276774193548388e-05,
+      "loss": 0.1715,
+      "step": 11185
+    },
+    {
+      "epoch": 0.178976,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.276612903225806e-05,
+      "loss": 0.1511,
+      "step": 11186
+    },
+    {
+      "epoch": 0.178992,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.276451612903226e-05,
+      "loss": 0.1637,
+      "step": 11187
+    },
+    {
+      "epoch": 0.179008,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.276290322580645e-05,
+      "loss": 0.1556,
+      "step": 11188
+    },
+    {
+      "epoch": 0.179024,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.276129032258065e-05,
+      "loss": 0.1506,
+      "step": 11189
+    },
+    {
+      "epoch": 0.17904,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.275967741935485e-05,
+      "loss": 0.1475,
+      "step": 11190
+    },
+    {
+      "epoch": 0.179056,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.275806451612903e-05,
+      "loss": 0.1949,
+      "step": 11191
+    },
+    {
+      "epoch": 0.179072,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.275645161290323e-05,
+      "loss": 0.1528,
+      "step": 11192
+    },
+    {
+      "epoch": 0.179088,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.275483870967742e-05,
+      "loss": 0.1713,
+      "step": 11193
+    },
+    {
+      "epoch": 0.179104,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.275322580645162e-05,
+      "loss": 0.1399,
+      "step": 11194
+    },
+    {
+      "epoch": 0.17912,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.27516129032258e-05,
+      "loss": 0.1482,
+      "step": 11195
+    },
+    {
+      "epoch": 0.179136,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.275e-05,
+      "loss": 0.1749,
+      "step": 11196
+    },
+    {
+      "epoch": 0.179152,
+      "grad_norm": 1.125,
+      "learning_rate": 8.274838709677419e-05,
+      "loss": 0.1626,
+      "step": 11197
+    },
+    {
+      "epoch": 0.179168,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.274677419354839e-05,
+      "loss": 0.1741,
+      "step": 11198
+    },
+    {
+      "epoch": 0.179184,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.274516129032258e-05,
+      "loss": 0.1709,
+      "step": 11199
+    },
+    {
+      "epoch": 0.1792,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.274354838709678e-05,
+      "loss": 0.1531,
+      "step": 11200
+    },
+    {
+      "epoch": 0.179216,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.274193548387096e-05,
+      "loss": 0.1503,
+      "step": 11201
+    },
+    {
+      "epoch": 0.179232,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.274032258064516e-05,
+      "loss": 0.1765,
+      "step": 11202
+    },
+    {
+      "epoch": 0.179248,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.273870967741936e-05,
+      "loss": 0.1385,
+      "step": 11203
+    },
+    {
+      "epoch": 0.179264,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.273709677419356e-05,
+      "loss": 0.1593,
+      "step": 11204
+    },
+    {
+      "epoch": 0.17928,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.273548387096775e-05,
+      "loss": 0.1813,
+      "step": 11205
+    },
+    {
+      "epoch": 0.179296,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.273387096774195e-05,
+      "loss": 0.2135,
+      "step": 11206
+    },
+    {
+      "epoch": 0.179312,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.273225806451613e-05,
+      "loss": 0.1842,
+      "step": 11207
+    },
+    {
+      "epoch": 0.179328,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.273064516129032e-05,
+      "loss": 0.1357,
+      "step": 11208
+    },
+    {
+      "epoch": 0.179344,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.272903225806452e-05,
+      "loss": 0.1457,
+      "step": 11209
+    },
+    {
+      "epoch": 0.17936,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.27274193548387e-05,
+      "loss": 0.172,
+      "step": 11210
+    },
+    {
+      "epoch": 0.179376,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.27258064516129e-05,
+      "loss": 0.1401,
+      "step": 11211
+    },
+    {
+      "epoch": 0.179392,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.272419354838709e-05,
+      "loss": 0.1809,
+      "step": 11212
+    },
+    {
+      "epoch": 0.179408,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.272258064516129e-05,
+      "loss": 0.1736,
+      "step": 11213
+    },
+    {
+      "epoch": 0.179424,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.272096774193549e-05,
+      "loss": 0.1919,
+      "step": 11214
+    },
+    {
+      "epoch": 0.17944,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.271935483870969e-05,
+      "loss": 0.1768,
+      "step": 11215
+    },
+    {
+      "epoch": 0.179456,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.271774193548388e-05,
+      "loss": 0.2014,
+      "step": 11216
+    },
+    {
+      "epoch": 0.179472,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.271612903225808e-05,
+      "loss": 0.1585,
+      "step": 11217
+    },
+    {
+      "epoch": 0.179488,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.271451612903226e-05,
+      "loss": 0.1675,
+      "step": 11218
+    },
+    {
+      "epoch": 0.179504,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.271290322580646e-05,
+      "loss": 0.1853,
+      "step": 11219
+    },
+    {
+      "epoch": 0.17952,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.271129032258065e-05,
+      "loss": 0.1846,
+      "step": 11220
+    },
+    {
+      "epoch": 0.179536,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.270967741935485e-05,
+      "loss": 0.1726,
+      "step": 11221
+    },
+    {
+      "epoch": 0.179552,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.270806451612903e-05,
+      "loss": 0.192,
+      "step": 11222
+    },
+    {
+      "epoch": 0.179568,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.270645161290323e-05,
+      "loss": 0.1721,
+      "step": 11223
+    },
+    {
+      "epoch": 0.179584,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.270483870967742e-05,
+      "loss": 0.1692,
+      "step": 11224
+    },
+    {
+      "epoch": 0.1796,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.27032258064516e-05,
+      "loss": 0.1524,
+      "step": 11225
+    },
+    {
+      "epoch": 0.179616,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.27016129032258e-05,
+      "loss": 0.1716,
+      "step": 11226
+    },
+    {
+      "epoch": 0.179632,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.27e-05,
+      "loss": 0.1486,
+      "step": 11227
+    },
+    {
+      "epoch": 0.179648,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.26983870967742e-05,
+      "loss": 0.1816,
+      "step": 11228
+    },
+    {
+      "epoch": 0.179664,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.269677419354839e-05,
+      "loss": 0.1824,
+      "step": 11229
+    },
+    {
+      "epoch": 0.17968,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.269516129032259e-05,
+      "loss": 0.2031,
+      "step": 11230
+    },
+    {
+      "epoch": 0.179696,
+      "grad_norm": 1.8125,
+      "learning_rate": 8.269354838709678e-05,
+      "loss": 0.1566,
+      "step": 11231
+    },
+    {
+      "epoch": 0.179712,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.269193548387098e-05,
+      "loss": 0.1797,
+      "step": 11232
+    },
+    {
+      "epoch": 0.179728,
+      "grad_norm": 1.125,
+      "learning_rate": 8.269032258064516e-05,
+      "loss": 0.1737,
+      "step": 11233
+    },
+    {
+      "epoch": 0.179744,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.268870967741936e-05,
+      "loss": 0.1695,
+      "step": 11234
+    },
+    {
+      "epoch": 0.17976,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.268709677419355e-05,
+      "loss": 0.1755,
+      "step": 11235
+    },
+    {
+      "epoch": 0.179776,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.268548387096775e-05,
+      "loss": 0.1534,
+      "step": 11236
+    },
+    {
+      "epoch": 0.179792,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.268387096774193e-05,
+      "loss": 0.1491,
+      "step": 11237
+    },
+    {
+      "epoch": 0.179808,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.268225806451613e-05,
+      "loss": 0.1764,
+      "step": 11238
+    },
+    {
+      "epoch": 0.179824,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.268064516129033e-05,
+      "loss": 0.17,
+      "step": 11239
+    },
+    {
+      "epoch": 0.17984,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.267903225806452e-05,
+      "loss": 0.2086,
+      "step": 11240
+    },
+    {
+      "epoch": 0.179856,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.267741935483872e-05,
+      "loss": 0.1751,
+      "step": 11241
+    },
+    {
+      "epoch": 0.179872,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.26758064516129e-05,
+      "loss": 0.1844,
+      "step": 11242
+    },
+    {
+      "epoch": 0.179888,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.26741935483871e-05,
+      "loss": 0.1966,
+      "step": 11243
+    },
+    {
+      "epoch": 0.179904,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.267258064516129e-05,
+      "loss": 0.2022,
+      "step": 11244
+    },
+    {
+      "epoch": 0.17992,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.267096774193549e-05,
+      "loss": 0.2037,
+      "step": 11245
+    },
+    {
+      "epoch": 0.179936,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.266935483870968e-05,
+      "loss": 0.1745,
+      "step": 11246
+    },
+    {
+      "epoch": 0.179952,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.266774193548388e-05,
+      "loss": 0.1558,
+      "step": 11247
+    },
+    {
+      "epoch": 0.179968,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.266612903225806e-05,
+      "loss": 0.186,
+      "step": 11248
+    },
+    {
+      "epoch": 0.179984,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.266451612903226e-05,
+      "loss": 0.2319,
+      "step": 11249
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3671875,
+      "learning_rate": 8.266290322580646e-05,
+      "loss": 0.1753,
+      "step": 11250
+    },
+    {
+      "epoch": 0.180016,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.266129032258066e-05,
+      "loss": 0.246,
+      "step": 11251
+    },
+    {
+      "epoch": 0.180032,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.265967741935485e-05,
+      "loss": 0.1833,
+      "step": 11252
+    },
+    {
+      "epoch": 0.180048,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.265806451612905e-05,
+      "loss": 0.2222,
+      "step": 11253
+    },
+    {
+      "epoch": 0.180064,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.265645161290323e-05,
+      "loss": 0.1822,
+      "step": 11254
+    },
+    {
+      "epoch": 0.18008,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.265483870967742e-05,
+      "loss": 0.1517,
+      "step": 11255
+    },
+    {
+      "epoch": 0.180096,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.265322580645162e-05,
+      "loss": 0.1539,
+      "step": 11256
+    },
+    {
+      "epoch": 0.180112,
+      "grad_norm": 1.4609375,
+      "learning_rate": 8.26516129032258e-05,
+      "loss": 0.1883,
+      "step": 11257
+    },
+    {
+      "epoch": 0.180128,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.265e-05,
+      "loss": 0.1573,
+      "step": 11258
+    },
+    {
+      "epoch": 0.180144,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.264838709677419e-05,
+      "loss": 0.1335,
+      "step": 11259
+    },
+    {
+      "epoch": 0.18016,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.264677419354839e-05,
+      "loss": 0.1822,
+      "step": 11260
+    },
+    {
+      "epoch": 0.180176,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.264516129032257e-05,
+      "loss": 0.1995,
+      "step": 11261
+    },
+    {
+      "epoch": 0.180192,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.264354838709677e-05,
+      "loss": 0.1483,
+      "step": 11262
+    },
+    {
+      "epoch": 0.180208,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.264193548387097e-05,
+      "loss": 0.1974,
+      "step": 11263
+    },
+    {
+      "epoch": 0.180224,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.264032258064517e-05,
+      "loss": 0.1474,
+      "step": 11264
+    },
+    {
+      "epoch": 0.18024,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.263870967741936e-05,
+      "loss": 0.1887,
+      "step": 11265
+    },
+    {
+      "epoch": 0.180256,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.263709677419356e-05,
+      "loss": 0.1644,
+      "step": 11266
+    },
+    {
+      "epoch": 0.180272,
+      "grad_norm": 1.265625,
+      "learning_rate": 8.263548387096775e-05,
+      "loss": 0.1589,
+      "step": 11267
+    },
+    {
+      "epoch": 0.180288,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.263387096774195e-05,
+      "loss": 0.1827,
+      "step": 11268
+    },
+    {
+      "epoch": 0.180304,
+      "grad_norm": 1.3984375,
+      "learning_rate": 8.263225806451613e-05,
+      "loss": 0.2368,
+      "step": 11269
+    },
+    {
+      "epoch": 0.18032,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.263064516129033e-05,
+      "loss": 0.1665,
+      "step": 11270
+    },
+    {
+      "epoch": 0.180336,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.262903225806452e-05,
+      "loss": 0.1581,
+      "step": 11271
+    },
+    {
+      "epoch": 0.180352,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.26274193548387e-05,
+      "loss": 0.1985,
+      "step": 11272
+    },
+    {
+      "epoch": 0.180368,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.26258064516129e-05,
+      "loss": 0.1557,
+      "step": 11273
+    },
+    {
+      "epoch": 0.180384,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.26241935483871e-05,
+      "loss": 0.1656,
+      "step": 11274
+    },
+    {
+      "epoch": 0.1804,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.26225806451613e-05,
+      "loss": 0.1624,
+      "step": 11275
+    },
+    {
+      "epoch": 0.180416,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.262096774193549e-05,
+      "loss": 0.1692,
+      "step": 11276
+    },
+    {
+      "epoch": 0.180432,
+      "grad_norm": 1.3671875,
+      "learning_rate": 8.261935483870969e-05,
+      "loss": 0.1547,
+      "step": 11277
+    },
+    {
+      "epoch": 0.180448,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.261774193548387e-05,
+      "loss": 0.139,
+      "step": 11278
+    },
+    {
+      "epoch": 0.180464,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.261612903225807e-05,
+      "loss": 0.1645,
+      "step": 11279
+    },
+    {
+      "epoch": 0.18048,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.261451612903226e-05,
+      "loss": 0.1484,
+      "step": 11280
+    },
+    {
+      "epoch": 0.180496,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.261290322580646e-05,
+      "loss": 0.1516,
+      "step": 11281
+    },
+    {
+      "epoch": 0.180512,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.261129032258065e-05,
+      "loss": 0.1326,
+      "step": 11282
+    },
+    {
+      "epoch": 0.180528,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.260967741935484e-05,
+      "loss": 0.1924,
+      "step": 11283
+    },
+    {
+      "epoch": 0.180544,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.260806451612903e-05,
+      "loss": 0.1628,
+      "step": 11284
+    },
+    {
+      "epoch": 0.18056,
+      "grad_norm": 0.625,
+      "learning_rate": 8.260645161290323e-05,
+      "loss": 0.1508,
+      "step": 11285
+    },
+    {
+      "epoch": 0.180576,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.260483870967743e-05,
+      "loss": 0.153,
+      "step": 11286
+    },
+    {
+      "epoch": 0.180592,
+      "grad_norm": 1.3515625,
+      "learning_rate": 8.260322580645162e-05,
+      "loss": 0.2099,
+      "step": 11287
+    },
+    {
+      "epoch": 0.180608,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.260161290322582e-05,
+      "loss": 0.1288,
+      "step": 11288
+    },
+    {
+      "epoch": 0.180624,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.26e-05,
+      "loss": 0.1583,
+      "step": 11289
+    },
+    {
+      "epoch": 0.18064,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.25983870967742e-05,
+      "loss": 0.1501,
+      "step": 11290
+    },
+    {
+      "epoch": 0.180656,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.259677419354839e-05,
+      "loss": 0.1543,
+      "step": 11291
+    },
+    {
+      "epoch": 0.180672,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.259516129032259e-05,
+      "loss": 0.1502,
+      "step": 11292
+    },
+    {
+      "epoch": 0.180688,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.259354838709677e-05,
+      "loss": 0.1377,
+      "step": 11293
+    },
+    {
+      "epoch": 0.180704,
+      "grad_norm": 1.0,
+      "learning_rate": 8.259193548387097e-05,
+      "loss": 0.1631,
+      "step": 11294
+    },
+    {
+      "epoch": 0.18072,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.259032258064516e-05,
+      "loss": 0.1536,
+      "step": 11295
+    },
+    {
+      "epoch": 0.180736,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.258870967741936e-05,
+      "loss": 0.1325,
+      "step": 11296
+    },
+    {
+      "epoch": 0.180752,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.258709677419354e-05,
+      "loss": 0.2071,
+      "step": 11297
+    },
+    {
+      "epoch": 0.180768,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.258548387096774e-05,
+      "loss": 0.1728,
+      "step": 11298
+    },
+    {
+      "epoch": 0.180784,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.258387096774194e-05,
+      "loss": 0.1498,
+      "step": 11299
+    },
+    {
+      "epoch": 0.1808,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.258225806451614e-05,
+      "loss": 0.1481,
+      "step": 11300
+    },
+    {
+      "epoch": 0.180816,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.258064516129033e-05,
+      "loss": 0.1495,
+      "step": 11301
+    },
+    {
+      "epoch": 0.180832,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.257903225806452e-05,
+      "loss": 0.1466,
+      "step": 11302
+    },
+    {
+      "epoch": 0.180848,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.257741935483872e-05,
+      "loss": 0.1617,
+      "step": 11303
+    },
+    {
+      "epoch": 0.180864,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.25758064516129e-05,
+      "loss": 0.15,
+      "step": 11304
+    },
+    {
+      "epoch": 0.18088,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.25741935483871e-05,
+      "loss": 0.1733,
+      "step": 11305
+    },
+    {
+      "epoch": 0.180896,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.257258064516129e-05,
+      "loss": 0.2078,
+      "step": 11306
+    },
+    {
+      "epoch": 0.180912,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.257096774193549e-05,
+      "loss": 0.1691,
+      "step": 11307
+    },
+    {
+      "epoch": 0.180928,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.256935483870967e-05,
+      "loss": 0.2108,
+      "step": 11308
+    },
+    {
+      "epoch": 0.180944,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.256774193548387e-05,
+      "loss": 0.1573,
+      "step": 11309
+    },
+    {
+      "epoch": 0.18096,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.256612903225807e-05,
+      "loss": 0.1921,
+      "step": 11310
+    },
+    {
+      "epoch": 0.180976,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.256451612903227e-05,
+      "loss": 0.2095,
+      "step": 11311
+    },
+    {
+      "epoch": 0.180992,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.256290322580646e-05,
+      "loss": 0.1787,
+      "step": 11312
+    },
+    {
+      "epoch": 0.181008,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.256129032258066e-05,
+      "loss": 0.1458,
+      "step": 11313
+    },
+    {
+      "epoch": 0.181024,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.255967741935484e-05,
+      "loss": 0.1601,
+      "step": 11314
+    },
+    {
+      "epoch": 0.18104,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.255806451612904e-05,
+      "loss": 0.1603,
+      "step": 11315
+    },
+    {
+      "epoch": 0.181056,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.255645161290323e-05,
+      "loss": 0.1758,
+      "step": 11316
+    },
+    {
+      "epoch": 0.181072,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.255483870967742e-05,
+      "loss": 0.1779,
+      "step": 11317
+    },
+    {
+      "epoch": 0.181088,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.255322580645162e-05,
+      "loss": 0.2164,
+      "step": 11318
+    },
+    {
+      "epoch": 0.181104,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.25516129032258e-05,
+      "loss": 0.1586,
+      "step": 11319
+    },
+    {
+      "epoch": 0.18112,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.255e-05,
+      "loss": 0.1347,
+      "step": 11320
+    },
+    {
+      "epoch": 0.181136,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.254838709677419e-05,
+      "loss": 0.1841,
+      "step": 11321
+    },
+    {
+      "epoch": 0.181152,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.254677419354839e-05,
+      "loss": 0.1501,
+      "step": 11322
+    },
+    {
+      "epoch": 0.181168,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.254516129032259e-05,
+      "loss": 0.1713,
+      "step": 11323
+    },
+    {
+      "epoch": 0.181184,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.254354838709679e-05,
+      "loss": 0.2151,
+      "step": 11324
+    },
+    {
+      "epoch": 0.1812,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.254193548387097e-05,
+      "loss": 0.1769,
+      "step": 11325
+    },
+    {
+      "epoch": 0.181216,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.254032258064517e-05,
+      "loss": 0.1683,
+      "step": 11326
+    },
+    {
+      "epoch": 0.181232,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.253870967741936e-05,
+      "loss": 0.1415,
+      "step": 11327
+    },
+    {
+      "epoch": 0.181248,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.253709677419356e-05,
+      "loss": 0.1516,
+      "step": 11328
+    },
+    {
+      "epoch": 0.181264,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.253548387096774e-05,
+      "loss": 0.178,
+      "step": 11329
+    },
+    {
+      "epoch": 0.18128,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.253387096774194e-05,
+      "loss": 0.1873,
+      "step": 11330
+    },
+    {
+      "epoch": 0.181296,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.253225806451613e-05,
+      "loss": 0.1849,
+      "step": 11331
+    },
+    {
+      "epoch": 0.181312,
+      "grad_norm": 1.8203125,
+      "learning_rate": 8.253064516129033e-05,
+      "loss": 0.1874,
+      "step": 11332
+    },
+    {
+      "epoch": 0.181328,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.252903225806451e-05,
+      "loss": 0.1907,
+      "step": 11333
+    },
+    {
+      "epoch": 0.181344,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.252741935483871e-05,
+      "loss": 0.2005,
+      "step": 11334
+    },
+    {
+      "epoch": 0.18136,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.252580645161291e-05,
+      "loss": 0.1926,
+      "step": 11335
+    },
+    {
+      "epoch": 0.181376,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.25241935483871e-05,
+      "loss": 0.1668,
+      "step": 11336
+    },
+    {
+      "epoch": 0.181392,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.25225806451613e-05,
+      "loss": 0.1845,
+      "step": 11337
+    },
+    {
+      "epoch": 0.181408,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.252096774193549e-05,
+      "loss": 0.1641,
+      "step": 11338
+    },
+    {
+      "epoch": 0.181424,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.251935483870969e-05,
+      "loss": 0.193,
+      "step": 11339
+    },
+    {
+      "epoch": 0.18144,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.251774193548387e-05,
+      "loss": 0.1351,
+      "step": 11340
+    },
+    {
+      "epoch": 0.181456,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.251612903225807e-05,
+      "loss": 0.1392,
+      "step": 11341
+    },
+    {
+      "epoch": 0.181472,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.251451612903226e-05,
+      "loss": 0.1739,
+      "step": 11342
+    },
+    {
+      "epoch": 0.181488,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.251290322580646e-05,
+      "loss": 0.206,
+      "step": 11343
+    },
+    {
+      "epoch": 0.181504,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.251129032258064e-05,
+      "loss": 0.1728,
+      "step": 11344
+    },
+    {
+      "epoch": 0.18152,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.250967741935484e-05,
+      "loss": 0.1917,
+      "step": 11345
+    },
+    {
+      "epoch": 0.181536,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.250806451612904e-05,
+      "loss": 0.1619,
+      "step": 11346
+    },
+    {
+      "epoch": 0.181552,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.250645161290324e-05,
+      "loss": 0.1392,
+      "step": 11347
+    },
+    {
+      "epoch": 0.181568,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.250483870967743e-05,
+      "loss": 0.1619,
+      "step": 11348
+    },
+    {
+      "epoch": 0.181584,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.250322580645161e-05,
+      "loss": 0.1786,
+      "step": 11349
+    },
+    {
+      "epoch": 0.1816,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.250161290322581e-05,
+      "loss": 0.1635,
+      "step": 11350
+    },
+    {
+      "epoch": 0.181616,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.25e-05,
+      "loss": 0.199,
+      "step": 11351
+    },
+    {
+      "epoch": 0.181632,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.24983870967742e-05,
+      "loss": 0.1499,
+      "step": 11352
+    },
+    {
+      "epoch": 0.181648,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.249677419354839e-05,
+      "loss": 0.1685,
+      "step": 11353
+    },
+    {
+      "epoch": 0.181664,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.249516129032258e-05,
+      "loss": 0.1486,
+      "step": 11354
+    },
+    {
+      "epoch": 0.18168,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.249354838709677e-05,
+      "loss": 0.1538,
+      "step": 11355
+    },
+    {
+      "epoch": 0.181696,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.249193548387097e-05,
+      "loss": 0.1703,
+      "step": 11356
+    },
+    {
+      "epoch": 0.181712,
+      "grad_norm": 1.0,
+      "learning_rate": 8.249032258064516e-05,
+      "loss": 0.1854,
+      "step": 11357
+    },
+    {
+      "epoch": 0.181728,
+      "grad_norm": 0.95703125,
+      "learning_rate": 8.248870967741936e-05,
+      "loss": 0.2263,
+      "step": 11358
+    },
+    {
+      "epoch": 0.181744,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.248709677419356e-05,
+      "loss": 0.1946,
+      "step": 11359
+    },
+    {
+      "epoch": 0.18176,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.248548387096776e-05,
+      "loss": 0.1776,
+      "step": 11360
+    },
+    {
+      "epoch": 0.181776,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.248387096774194e-05,
+      "loss": 0.1848,
+      "step": 11361
+    },
+    {
+      "epoch": 0.181792,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.248225806451614e-05,
+      "loss": 0.2029,
+      "step": 11362
+    },
+    {
+      "epoch": 0.181808,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.248064516129033e-05,
+      "loss": 0.1246,
+      "step": 11363
+    },
+    {
+      "epoch": 0.181824,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.247903225806451e-05,
+      "loss": 0.1753,
+      "step": 11364
+    },
+    {
+      "epoch": 0.18184,
+      "grad_norm": 1.25,
+      "learning_rate": 8.247741935483871e-05,
+      "loss": 0.1777,
+      "step": 11365
+    },
+    {
+      "epoch": 0.181856,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.24758064516129e-05,
+      "loss": 0.2501,
+      "step": 11366
+    },
+    {
+      "epoch": 0.181872,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.24741935483871e-05,
+      "loss": 0.1518,
+      "step": 11367
+    },
+    {
+      "epoch": 0.181888,
+      "grad_norm": 0.875,
+      "learning_rate": 8.247258064516128e-05,
+      "loss": 0.1714,
+      "step": 11368
+    },
+    {
+      "epoch": 0.181904,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.247096774193548e-05,
+      "loss": 0.1499,
+      "step": 11369
+    },
+    {
+      "epoch": 0.18192,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.246935483870968e-05,
+      "loss": 0.1511,
+      "step": 11370
+    },
+    {
+      "epoch": 0.181936,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.246774193548388e-05,
+      "loss": 0.1466,
+      "step": 11371
+    },
+    {
+      "epoch": 0.181952,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.246612903225807e-05,
+      "loss": 0.1603,
+      "step": 11372
+    },
+    {
+      "epoch": 0.181968,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.246451612903227e-05,
+      "loss": 0.1873,
+      "step": 11373
+    },
+    {
+      "epoch": 0.181984,
+      "grad_norm": 1.3828125,
+      "learning_rate": 8.246290322580646e-05,
+      "loss": 0.1543,
+      "step": 11374
+    },
+    {
+      "epoch": 0.182,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.246129032258066e-05,
+      "loss": 0.1487,
+      "step": 11375
+    },
+    {
+      "epoch": 0.182016,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.245967741935484e-05,
+      "loss": 0.1698,
+      "step": 11376
+    },
+    {
+      "epoch": 0.182032,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.245806451612904e-05,
+      "loss": 0.152,
+      "step": 11377
+    },
+    {
+      "epoch": 0.182048,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.245645161290323e-05,
+      "loss": 0.1555,
+      "step": 11378
+    },
+    {
+      "epoch": 0.182064,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.245483870967743e-05,
+      "loss": 0.1295,
+      "step": 11379
+    },
+    {
+      "epoch": 0.18208,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.245322580645161e-05,
+      "loss": 0.2041,
+      "step": 11380
+    },
+    {
+      "epoch": 0.182096,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.245161290322581e-05,
+      "loss": 0.1636,
+      "step": 11381
+    },
+    {
+      "epoch": 0.182112,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.245e-05,
+      "loss": 0.1728,
+      "step": 11382
+    },
+    {
+      "epoch": 0.182128,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.24483870967742e-05,
+      "loss": 0.157,
+      "step": 11383
+    },
+    {
+      "epoch": 0.182144,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.24467741935484e-05,
+      "loss": 0.1629,
+      "step": 11384
+    },
+    {
+      "epoch": 0.18216,
+      "grad_norm": 1.3359375,
+      "learning_rate": 8.244516129032258e-05,
+      "loss": 0.1785,
+      "step": 11385
+    },
+    {
+      "epoch": 0.182176,
+      "grad_norm": 1.4375,
+      "learning_rate": 8.244354838709678e-05,
+      "loss": 0.1992,
+      "step": 11386
+    },
+    {
+      "epoch": 0.182192,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.244193548387097e-05,
+      "loss": 0.1542,
+      "step": 11387
+    },
+    {
+      "epoch": 0.182208,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.244032258064517e-05,
+      "loss": 0.1629,
+      "step": 11388
+    },
+    {
+      "epoch": 0.182224,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.243870967741936e-05,
+      "loss": 0.1643,
+      "step": 11389
+    },
+    {
+      "epoch": 0.18224,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.243709677419355e-05,
+      "loss": 0.1749,
+      "step": 11390
+    },
+    {
+      "epoch": 0.182256,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.243548387096774e-05,
+      "loss": 0.2004,
+      "step": 11391
+    },
+    {
+      "epoch": 0.182272,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.243387096774194e-05,
+      "loss": 0.1808,
+      "step": 11392
+    },
+    {
+      "epoch": 0.182288,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.243225806451613e-05,
+      "loss": 0.1697,
+      "step": 11393
+    },
+    {
+      "epoch": 0.182304,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.243064516129033e-05,
+      "loss": 0.1845,
+      "step": 11394
+    },
+    {
+      "epoch": 0.18232,
+      "grad_norm": 1.390625,
+      "learning_rate": 8.242903225806453e-05,
+      "loss": 0.1725,
+      "step": 11395
+    },
+    {
+      "epoch": 0.182336,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.242741935483871e-05,
+      "loss": 0.193,
+      "step": 11396
+    },
+    {
+      "epoch": 0.182352,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.242580645161291e-05,
+      "loss": 0.1471,
+      "step": 11397
+    },
+    {
+      "epoch": 0.182368,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.24241935483871e-05,
+      "loss": 0.1845,
+      "step": 11398
+    },
+    {
+      "epoch": 0.182384,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.24225806451613e-05,
+      "loss": 0.1479,
+      "step": 11399
+    },
+    {
+      "epoch": 0.1824,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.242096774193548e-05,
+      "loss": 0.1244,
+      "step": 11400
+    },
+    {
+      "epoch": 0.182416,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.241935483870968e-05,
+      "loss": 0.1454,
+      "step": 11401
+    },
+    {
+      "epoch": 0.182432,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.241774193548387e-05,
+      "loss": 0.1831,
+      "step": 11402
+    },
+    {
+      "epoch": 0.182448,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.241612903225807e-05,
+      "loss": 0.1494,
+      "step": 11403
+    },
+    {
+      "epoch": 0.182464,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.241451612903225e-05,
+      "loss": 0.1608,
+      "step": 11404
+    },
+    {
+      "epoch": 0.18248,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.241290322580645e-05,
+      "loss": 0.1715,
+      "step": 11405
+    },
+    {
+      "epoch": 0.182496,
+      "grad_norm": 1.46875,
+      "learning_rate": 8.241129032258065e-05,
+      "loss": 0.1739,
+      "step": 11406
+    },
+    {
+      "epoch": 0.182512,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.240967741935485e-05,
+      "loss": 0.1785,
+      "step": 11407
+    },
+    {
+      "epoch": 0.182528,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.240806451612904e-05,
+      "loss": 0.1973,
+      "step": 11408
+    },
+    {
+      "epoch": 0.182544,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.240645161290324e-05,
+      "loss": 0.157,
+      "step": 11409
+    },
+    {
+      "epoch": 0.18256,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.240483870967743e-05,
+      "loss": 0.1898,
+      "step": 11410
+    },
+    {
+      "epoch": 0.182576,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.240322580645161e-05,
+      "loss": 0.1539,
+      "step": 11411
+    },
+    {
+      "epoch": 0.182592,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.240161290322581e-05,
+      "loss": 0.1618,
+      "step": 11412
+    },
+    {
+      "epoch": 0.182608,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.24e-05,
+      "loss": 0.1687,
+      "step": 11413
+    },
+    {
+      "epoch": 0.182624,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.23983870967742e-05,
+      "loss": 0.184,
+      "step": 11414
+    },
+    {
+      "epoch": 0.18264,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.239677419354838e-05,
+      "loss": 0.1592,
+      "step": 11415
+    },
+    {
+      "epoch": 0.182656,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.239516129032258e-05,
+      "loss": 0.1784,
+      "step": 11416
+    },
+    {
+      "epoch": 0.182672,
+      "grad_norm": 0.625,
+      "learning_rate": 8.239354838709677e-05,
+      "loss": 0.1511,
+      "step": 11417
+    },
+    {
+      "epoch": 0.182688,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.239193548387097e-05,
+      "loss": 0.1645,
+      "step": 11418
+    },
+    {
+      "epoch": 0.182704,
+      "grad_norm": 0.75,
+      "learning_rate": 8.239032258064517e-05,
+      "loss": 0.2067,
+      "step": 11419
+    },
+    {
+      "epoch": 0.18272,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.238870967741937e-05,
+      "loss": 0.1807,
+      "step": 11420
+    },
+    {
+      "epoch": 0.182736,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.238709677419355e-05,
+      "loss": 0.1804,
+      "step": 11421
+    },
+    {
+      "epoch": 0.182752,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.238548387096775e-05,
+      "loss": 0.2026,
+      "step": 11422
+    },
+    {
+      "epoch": 0.182768,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.238387096774194e-05,
+      "loss": 0.1541,
+      "step": 11423
+    },
+    {
+      "epoch": 0.182784,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.238225806451614e-05,
+      "loss": 0.1664,
+      "step": 11424
+    },
+    {
+      "epoch": 0.1828,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.238064516129032e-05,
+      "loss": 0.1748,
+      "step": 11425
+    },
+    {
+      "epoch": 0.182816,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.237903225806451e-05,
+      "loss": 0.1922,
+      "step": 11426
+    },
+    {
+      "epoch": 0.182832,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.237741935483871e-05,
+      "loss": 0.1919,
+      "step": 11427
+    },
+    {
+      "epoch": 0.182848,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.23758064516129e-05,
+      "loss": 0.1406,
+      "step": 11428
+    },
+    {
+      "epoch": 0.182864,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.23741935483871e-05,
+      "loss": 0.1604,
+      "step": 11429
+    },
+    {
+      "epoch": 0.18288,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.23725806451613e-05,
+      "loss": 0.1558,
+      "step": 11430
+    },
+    {
+      "epoch": 0.182896,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.23709677419355e-05,
+      "loss": 0.1492,
+      "step": 11431
+    },
+    {
+      "epoch": 0.182912,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.236935483870968e-05,
+      "loss": 0.1866,
+      "step": 11432
+    },
+    {
+      "epoch": 0.182928,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.236774193548388e-05,
+      "loss": 0.1744,
+      "step": 11433
+    },
+    {
+      "epoch": 0.182944,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.236612903225807e-05,
+      "loss": 0.1519,
+      "step": 11434
+    },
+    {
+      "epoch": 0.18296,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.236451612903227e-05,
+      "loss": 0.1988,
+      "step": 11435
+    },
+    {
+      "epoch": 0.182976,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.236290322580645e-05,
+      "loss": 0.1568,
+      "step": 11436
+    },
+    {
+      "epoch": 0.182992,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.236129032258065e-05,
+      "loss": 0.1644,
+      "step": 11437
+    },
+    {
+      "epoch": 0.183008,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.235967741935484e-05,
+      "loss": 0.1505,
+      "step": 11438
+    },
+    {
+      "epoch": 0.183024,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.235806451612904e-05,
+      "loss": 0.1535,
+      "step": 11439
+    },
+    {
+      "epoch": 0.18304,
+      "grad_norm": 0.75,
+      "learning_rate": 8.235645161290322e-05,
+      "loss": 0.1733,
+      "step": 11440
+    },
+    {
+      "epoch": 0.183056,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.235483870967742e-05,
+      "loss": 0.1715,
+      "step": 11441
+    },
+    {
+      "epoch": 0.183072,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.235322580645162e-05,
+      "loss": 0.1889,
+      "step": 11442
+    },
+    {
+      "epoch": 0.183088,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.235161290322581e-05,
+      "loss": 0.1548,
+      "step": 11443
+    },
+    {
+      "epoch": 0.183104,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.235000000000001e-05,
+      "loss": 0.1932,
+      "step": 11444
+    },
+    {
+      "epoch": 0.18312,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.23483870967742e-05,
+      "loss": 0.1822,
+      "step": 11445
+    },
+    {
+      "epoch": 0.183136,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.23467741935484e-05,
+      "loss": 0.1725,
+      "step": 11446
+    },
+    {
+      "epoch": 0.183152,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.234516129032258e-05,
+      "loss": 0.1849,
+      "step": 11447
+    },
+    {
+      "epoch": 0.183168,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.234354838709678e-05,
+      "loss": 0.2292,
+      "step": 11448
+    },
+    {
+      "epoch": 0.183184,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.234193548387097e-05,
+      "loss": 0.193,
+      "step": 11449
+    },
+    {
+      "epoch": 0.1832,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.234032258064517e-05,
+      "loss": 0.1588,
+      "step": 11450
+    },
+    {
+      "epoch": 0.183216,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.233870967741935e-05,
+      "loss": 0.1612,
+      "step": 11451
+    },
+    {
+      "epoch": 0.183232,
+      "grad_norm": 1.484375,
+      "learning_rate": 8.233709677419355e-05,
+      "loss": 0.1529,
+      "step": 11452
+    },
+    {
+      "epoch": 0.183248,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.233548387096774e-05,
+      "loss": 0.1669,
+      "step": 11453
+    },
+    {
+      "epoch": 0.183264,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.233387096774194e-05,
+      "loss": 0.1786,
+      "step": 11454
+    },
+    {
+      "epoch": 0.18328,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.233225806451614e-05,
+      "loss": 0.1636,
+      "step": 11455
+    },
+    {
+      "epoch": 0.183296,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.233064516129034e-05,
+      "loss": 0.206,
+      "step": 11456
+    },
+    {
+      "epoch": 0.183312,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.232903225806452e-05,
+      "loss": 0.1906,
+      "step": 11457
+    },
+    {
+      "epoch": 0.183328,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.232741935483871e-05,
+      "loss": 0.2209,
+      "step": 11458
+    },
+    {
+      "epoch": 0.183344,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.232580645161291e-05,
+      "loss": 0.1387,
+      "step": 11459
+    },
+    {
+      "epoch": 0.18336,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.23241935483871e-05,
+      "loss": 0.1758,
+      "step": 11460
+    },
+    {
+      "epoch": 0.183376,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.23225806451613e-05,
+      "loss": 0.1624,
+      "step": 11461
+    },
+    {
+      "epoch": 0.183392,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.232096774193548e-05,
+      "loss": 0.1645,
+      "step": 11462
+    },
+    {
+      "epoch": 0.183408,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.231935483870968e-05,
+      "loss": 0.1608,
+      "step": 11463
+    },
+    {
+      "epoch": 0.183424,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.231774193548387e-05,
+      "loss": 0.1612,
+      "step": 11464
+    },
+    {
+      "epoch": 0.18344,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.231612903225807e-05,
+      "loss": 0.1635,
+      "step": 11465
+    },
+    {
+      "epoch": 0.183456,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.231451612903227e-05,
+      "loss": 0.2237,
+      "step": 11466
+    },
+    {
+      "epoch": 0.183472,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.231290322580647e-05,
+      "loss": 0.1794,
+      "step": 11467
+    },
+    {
+      "epoch": 0.183488,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.231129032258065e-05,
+      "loss": 0.1436,
+      "step": 11468
+    },
+    {
+      "epoch": 0.183504,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.230967741935485e-05,
+      "loss": 0.1773,
+      "step": 11469
+    },
+    {
+      "epoch": 0.18352,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.230806451612904e-05,
+      "loss": 0.1721,
+      "step": 11470
+    },
+    {
+      "epoch": 0.183536,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.230645161290324e-05,
+      "loss": 0.1644,
+      "step": 11471
+    },
+    {
+      "epoch": 0.183552,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.230483870967742e-05,
+      "loss": 0.1324,
+      "step": 11472
+    },
+    {
+      "epoch": 0.183568,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.230322580645161e-05,
+      "loss": 0.1671,
+      "step": 11473
+    },
+    {
+      "epoch": 0.183584,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.230161290322581e-05,
+      "loss": 0.1807,
+      "step": 11474
+    },
+    {
+      "epoch": 0.1836,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.23e-05,
+      "loss": 0.1679,
+      "step": 11475
+    },
+    {
+      "epoch": 0.183616,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.22983870967742e-05,
+      "loss": 0.2303,
+      "step": 11476
+    },
+    {
+      "epoch": 0.183632,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.229677419354838e-05,
+      "loss": 0.1878,
+      "step": 11477
+    },
+    {
+      "epoch": 0.183648,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.229516129032258e-05,
+      "loss": 0.1563,
+      "step": 11478
+    },
+    {
+      "epoch": 0.183664,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.229354838709678e-05,
+      "loss": 0.1425,
+      "step": 11479
+    },
+    {
+      "epoch": 0.18368,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.229193548387098e-05,
+      "loss": 0.2111,
+      "step": 11480
+    },
+    {
+      "epoch": 0.183696,
+      "grad_norm": 0.875,
+      "learning_rate": 8.229032258064517e-05,
+      "loss": 0.1678,
+      "step": 11481
+    },
+    {
+      "epoch": 0.183712,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.228870967741936e-05,
+      "loss": 0.1748,
+      "step": 11482
+    },
+    {
+      "epoch": 0.183728,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.228709677419355e-05,
+      "loss": 0.1539,
+      "step": 11483
+    },
+    {
+      "epoch": 0.183744,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.228548387096775e-05,
+      "loss": 0.1817,
+      "step": 11484
+    },
+    {
+      "epoch": 0.18376,
+      "grad_norm": 1.4296875,
+      "learning_rate": 8.228387096774194e-05,
+      "loss": 0.1625,
+      "step": 11485
+    },
+    {
+      "epoch": 0.183776,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.228225806451614e-05,
+      "loss": 0.1967,
+      "step": 11486
+    },
+    {
+      "epoch": 0.183792,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.228064516129032e-05,
+      "loss": 0.1486,
+      "step": 11487
+    },
+    {
+      "epoch": 0.183808,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.227903225806451e-05,
+      "loss": 0.1673,
+      "step": 11488
+    },
+    {
+      "epoch": 0.183824,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.227741935483871e-05,
+      "loss": 0.1775,
+      "step": 11489
+    },
+    {
+      "epoch": 0.18384,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.227580645161291e-05,
+      "loss": 0.1777,
+      "step": 11490
+    },
+    {
+      "epoch": 0.183856,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.227419354838711e-05,
+      "loss": 0.1831,
+      "step": 11491
+    },
+    {
+      "epoch": 0.183872,
+      "grad_norm": 1.7734375,
+      "learning_rate": 8.22725806451613e-05,
+      "loss": 0.1857,
+      "step": 11492
+    },
+    {
+      "epoch": 0.183888,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.227096774193549e-05,
+      "loss": 0.1977,
+      "step": 11493
+    },
+    {
+      "epoch": 0.183904,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.226935483870968e-05,
+      "loss": 0.225,
+      "step": 11494
+    },
+    {
+      "epoch": 0.18392,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.226774193548388e-05,
+      "loss": 0.1718,
+      "step": 11495
+    },
+    {
+      "epoch": 0.183936,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.226612903225806e-05,
+      "loss": 0.1453,
+      "step": 11496
+    },
+    {
+      "epoch": 0.183952,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.226451612903226e-05,
+      "loss": 0.19,
+      "step": 11497
+    },
+    {
+      "epoch": 0.183968,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.226290322580645e-05,
+      "loss": 0.1487,
+      "step": 11498
+    },
+    {
+      "epoch": 0.183984,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.226129032258065e-05,
+      "loss": 0.1743,
+      "step": 11499
+    },
+    {
+      "epoch": 0.184,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.225967741935484e-05,
+      "loss": 0.1531,
+      "step": 11500
+    },
+    {
+      "epoch": 0.184016,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.225806451612904e-05,
+      "loss": 0.1703,
+      "step": 11501
+    },
+    {
+      "epoch": 0.184032,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.225645161290324e-05,
+      "loss": 0.193,
+      "step": 11502
+    },
+    {
+      "epoch": 0.184048,
+      "grad_norm": 1.3984375,
+      "learning_rate": 8.225483870967744e-05,
+      "loss": 0.1647,
+      "step": 11503
+    },
+    {
+      "epoch": 0.184064,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.225322580645162e-05,
+      "loss": 0.1585,
+      "step": 11504
+    },
+    {
+      "epoch": 0.18408,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.225161290322581e-05,
+      "loss": 0.2334,
+      "step": 11505
+    },
+    {
+      "epoch": 0.184096,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.225000000000001e-05,
+      "loss": 0.173,
+      "step": 11506
+    },
+    {
+      "epoch": 0.184112,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.224838709677419e-05,
+      "loss": 0.1702,
+      "step": 11507
+    },
+    {
+      "epoch": 0.184128,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.224677419354839e-05,
+      "loss": 0.1641,
+      "step": 11508
+    },
+    {
+      "epoch": 0.184144,
+      "grad_norm": 1.28125,
+      "learning_rate": 8.224516129032258e-05,
+      "loss": 0.186,
+      "step": 11509
+    },
+    {
+      "epoch": 0.18416,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.224354838709678e-05,
+      "loss": 0.1466,
+      "step": 11510
+    },
+    {
+      "epoch": 0.184176,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.224193548387096e-05,
+      "loss": 0.1655,
+      "step": 11511
+    },
+    {
+      "epoch": 0.184192,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.224032258064516e-05,
+      "loss": 0.2207,
+      "step": 11512
+    },
+    {
+      "epoch": 0.184208,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.223870967741935e-05,
+      "loss": 0.1761,
+      "step": 11513
+    },
+    {
+      "epoch": 0.184224,
+      "grad_norm": 1.265625,
+      "learning_rate": 8.223709677419355e-05,
+      "loss": 0.1406,
+      "step": 11514
+    },
+    {
+      "epoch": 0.18424,
+      "grad_norm": 1.734375,
+      "learning_rate": 8.223548387096775e-05,
+      "loss": 0.2458,
+      "step": 11515
+    },
+    {
+      "epoch": 0.184256,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.223387096774195e-05,
+      "loss": 0.1619,
+      "step": 11516
+    },
+    {
+      "epoch": 0.184272,
+      "grad_norm": 1.40625,
+      "learning_rate": 8.223225806451614e-05,
+      "loss": 0.1495,
+      "step": 11517
+    },
+    {
+      "epoch": 0.184288,
+      "grad_norm": 1.3125,
+      "learning_rate": 8.223064516129033e-05,
+      "loss": 0.1833,
+      "step": 11518
+    },
+    {
+      "epoch": 0.184304,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.222903225806452e-05,
+      "loss": 0.142,
+      "step": 11519
+    },
+    {
+      "epoch": 0.18432,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.222741935483871e-05,
+      "loss": 0.1955,
+      "step": 11520
+    },
+    {
+      "epoch": 0.184336,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.22258064516129e-05,
+      "loss": 0.1663,
+      "step": 11521
+    },
+    {
+      "epoch": 0.184352,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.222419354838709e-05,
+      "loss": 0.1693,
+      "step": 11522
+    },
+    {
+      "epoch": 0.184368,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.222258064516129e-05,
+      "loss": 0.1358,
+      "step": 11523
+    },
+    {
+      "epoch": 0.184384,
+      "grad_norm": 0.423828125,
+      "learning_rate": 8.222096774193548e-05,
+      "loss": 0.1178,
+      "step": 11524
+    },
+    {
+      "epoch": 0.1844,
+      "grad_norm": 1.4140625,
+      "learning_rate": 8.221935483870968e-05,
+      "loss": 0.1666,
+      "step": 11525
+    },
+    {
+      "epoch": 0.184416,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.221774193548388e-05,
+      "loss": 0.1578,
+      "step": 11526
+    },
+    {
+      "epoch": 0.184432,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.221612903225808e-05,
+      "loss": 0.2036,
+      "step": 11527
+    },
+    {
+      "epoch": 0.184448,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.221451612903226e-05,
+      "loss": 0.1787,
+      "step": 11528
+    },
+    {
+      "epoch": 0.184464,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.221290322580646e-05,
+      "loss": 0.1571,
+      "step": 11529
+    },
+    {
+      "epoch": 0.18448,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.221129032258065e-05,
+      "loss": 0.1732,
+      "step": 11530
+    },
+    {
+      "epoch": 0.184496,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.220967741935485e-05,
+      "loss": 0.1427,
+      "step": 11531
+    },
+    {
+      "epoch": 0.184512,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.220806451612903e-05,
+      "loss": 0.2222,
+      "step": 11532
+    },
+    {
+      "epoch": 0.184528,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.220645161290323e-05,
+      "loss": 0.169,
+      "step": 11533
+    },
+    {
+      "epoch": 0.184544,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.220483870967742e-05,
+      "loss": 0.1361,
+      "step": 11534
+    },
+    {
+      "epoch": 0.18456,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.22032258064516e-05,
+      "loss": 0.177,
+      "step": 11535
+    },
+    {
+      "epoch": 0.184576,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.22016129032258e-05,
+      "loss": 0.1212,
+      "step": 11536
+    },
+    {
+      "epoch": 0.184592,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.22e-05,
+      "loss": 0.1983,
+      "step": 11537
+    },
+    {
+      "epoch": 0.184608,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.21983870967742e-05,
+      "loss": 0.1647,
+      "step": 11538
+    },
+    {
+      "epoch": 0.184624,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.219677419354839e-05,
+      "loss": 0.1433,
+      "step": 11539
+    },
+    {
+      "epoch": 0.18464,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.219516129032259e-05,
+      "loss": 0.1839,
+      "step": 11540
+    },
+    {
+      "epoch": 0.184656,
+      "grad_norm": 0.49609375,
+      "learning_rate": 8.219354838709678e-05,
+      "loss": 0.1259,
+      "step": 11541
+    },
+    {
+      "epoch": 0.184672,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.219193548387098e-05,
+      "loss": 0.154,
+      "step": 11542
+    },
+    {
+      "epoch": 0.184688,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.219032258064516e-05,
+      "loss": 0.1535,
+      "step": 11543
+    },
+    {
+      "epoch": 0.184704,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.218870967741936e-05,
+      "loss": 0.1512,
+      "step": 11544
+    },
+    {
+      "epoch": 0.18472,
+      "grad_norm": 1.3359375,
+      "learning_rate": 8.218709677419355e-05,
+      "loss": 0.1806,
+      "step": 11545
+    },
+    {
+      "epoch": 0.184736,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.218548387096775e-05,
+      "loss": 0.1629,
+      "step": 11546
+    },
+    {
+      "epoch": 0.184752,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.218387096774193e-05,
+      "loss": 0.1717,
+      "step": 11547
+    },
+    {
+      "epoch": 0.184768,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.218225806451613e-05,
+      "loss": 0.1668,
+      "step": 11548
+    },
+    {
+      "epoch": 0.184784,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.218064516129032e-05,
+      "loss": 0.1703,
+      "step": 11549
+    },
+    {
+      "epoch": 0.1848,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.217903225806452e-05,
+      "loss": 0.133,
+      "step": 11550
+    },
+    {
+      "epoch": 0.184816,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.217741935483872e-05,
+      "loss": 0.1882,
+      "step": 11551
+    },
+    {
+      "epoch": 0.184832,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.21758064516129e-05,
+      "loss": 0.1578,
+      "step": 11552
+    },
+    {
+      "epoch": 0.184848,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.21741935483871e-05,
+      "loss": 0.1638,
+      "step": 11553
+    },
+    {
+      "epoch": 0.184864,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.217258064516129e-05,
+      "loss": 0.2546,
+      "step": 11554
+    },
+    {
+      "epoch": 0.18488,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.217096774193549e-05,
+      "loss": 0.1669,
+      "step": 11555
+    },
+    {
+      "epoch": 0.184896,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.216935483870968e-05,
+      "loss": 0.1682,
+      "step": 11556
+    },
+    {
+      "epoch": 0.184912,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.216774193548388e-05,
+      "loss": 0.1688,
+      "step": 11557
+    },
+    {
+      "epoch": 0.184928,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.216612903225806e-05,
+      "loss": 0.198,
+      "step": 11558
+    },
+    {
+      "epoch": 0.184944,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.216451612903226e-05,
+      "loss": 0.1627,
+      "step": 11559
+    },
+    {
+      "epoch": 0.18496,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.216290322580645e-05,
+      "loss": 0.1515,
+      "step": 11560
+    },
+    {
+      "epoch": 0.184976,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.216129032258065e-05,
+      "loss": 0.2138,
+      "step": 11561
+    },
+    {
+      "epoch": 0.184992,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.215967741935485e-05,
+      "loss": 0.1733,
+      "step": 11562
+    },
+    {
+      "epoch": 0.185008,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.215806451612905e-05,
+      "loss": 0.1973,
+      "step": 11563
+    },
+    {
+      "epoch": 0.185024,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.215645161290323e-05,
+      "loss": 0.1609,
+      "step": 11564
+    },
+    {
+      "epoch": 0.18504,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.215483870967743e-05,
+      "loss": 0.155,
+      "step": 11565
+    },
+    {
+      "epoch": 0.185056,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.215322580645162e-05,
+      "loss": 0.1634,
+      "step": 11566
+    },
+    {
+      "epoch": 0.185072,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.21516129032258e-05,
+      "loss": 0.1744,
+      "step": 11567
+    },
+    {
+      "epoch": 0.185088,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.215e-05,
+      "loss": 0.1423,
+      "step": 11568
+    },
+    {
+      "epoch": 0.185104,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.214838709677419e-05,
+      "loss": 0.1547,
+      "step": 11569
+    },
+    {
+      "epoch": 0.18512,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.214677419354839e-05,
+      "loss": 0.1613,
+      "step": 11570
+    },
+    {
+      "epoch": 0.185136,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.214516129032258e-05,
+      "loss": 0.1832,
+      "step": 11571
+    },
+    {
+      "epoch": 0.185152,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.214354838709678e-05,
+      "loss": 0.1695,
+      "step": 11572
+    },
+    {
+      "epoch": 0.185168,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.214193548387096e-05,
+      "loss": 0.1904,
+      "step": 11573
+    },
+    {
+      "epoch": 0.185184,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.214032258064516e-05,
+      "loss": 0.1854,
+      "step": 11574
+    },
+    {
+      "epoch": 0.1852,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.213870967741936e-05,
+      "loss": 0.1521,
+      "step": 11575
+    },
+    {
+      "epoch": 0.185216,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.213709677419356e-05,
+      "loss": 0.1686,
+      "step": 11576
+    },
+    {
+      "epoch": 0.185232,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.213548387096775e-05,
+      "loss": 0.1794,
+      "step": 11577
+    },
+    {
+      "epoch": 0.185248,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.213387096774195e-05,
+      "loss": 0.1626,
+      "step": 11578
+    },
+    {
+      "epoch": 0.185264,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.213225806451613e-05,
+      "loss": 0.2098,
+      "step": 11579
+    },
+    {
+      "epoch": 0.18528,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.213064516129033e-05,
+      "loss": 0.1822,
+      "step": 11580
+    },
+    {
+      "epoch": 0.185296,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.212903225806452e-05,
+      "loss": 0.1676,
+      "step": 11581
+    },
+    {
+      "epoch": 0.185312,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.21274193548387e-05,
+      "loss": 0.1804,
+      "step": 11582
+    },
+    {
+      "epoch": 0.185328,
+      "grad_norm": 0.75,
+      "learning_rate": 8.21258064516129e-05,
+      "loss": 0.1554,
+      "step": 11583
+    },
+    {
+      "epoch": 0.185344,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.212419354838709e-05,
+      "loss": 0.1702,
+      "step": 11584
+    },
+    {
+      "epoch": 0.18536,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.212258064516129e-05,
+      "loss": 0.1372,
+      "step": 11585
+    },
+    {
+      "epoch": 0.185376,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.212096774193549e-05,
+      "loss": 0.1577,
+      "step": 11586
+    },
+    {
+      "epoch": 0.185392,
+      "grad_norm": 1.375,
+      "learning_rate": 8.211935483870969e-05,
+      "loss": 0.1729,
+      "step": 11587
+    },
+    {
+      "epoch": 0.185408,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.211774193548388e-05,
+      "loss": 0.1455,
+      "step": 11588
+    },
+    {
+      "epoch": 0.185424,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.211612903225807e-05,
+      "loss": 0.1764,
+      "step": 11589
+    },
+    {
+      "epoch": 0.18544,
+      "grad_norm": 0.625,
+      "learning_rate": 8.211451612903226e-05,
+      "loss": 0.1985,
+      "step": 11590
+    },
+    {
+      "epoch": 0.185456,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.211290322580646e-05,
+      "loss": 0.1916,
+      "step": 11591
+    },
+    {
+      "epoch": 0.185472,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.211129032258065e-05,
+      "loss": 0.1883,
+      "step": 11592
+    },
+    {
+      "epoch": 0.185488,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.210967741935485e-05,
+      "loss": 0.1741,
+      "step": 11593
+    },
+    {
+      "epoch": 0.185504,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.210806451612903e-05,
+      "loss": 0.1868,
+      "step": 11594
+    },
+    {
+      "epoch": 0.18552,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.210645161290323e-05,
+      "loss": 0.2097,
+      "step": 11595
+    },
+    {
+      "epoch": 0.185536,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.210483870967742e-05,
+      "loss": 0.2115,
+      "step": 11596
+    },
+    {
+      "epoch": 0.185552,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.210322580645162e-05,
+      "loss": 0.1712,
+      "step": 11597
+    },
+    {
+      "epoch": 0.185568,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.210161290322582e-05,
+      "loss": 0.1616,
+      "step": 11598
+    },
+    {
+      "epoch": 0.185584,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.21e-05,
+      "loss": 0.1548,
+      "step": 11599
+    },
+    {
+      "epoch": 0.1856,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.20983870967742e-05,
+      "loss": 0.1743,
+      "step": 11600
+    },
+    {
+      "epoch": 0.185616,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.209677419354839e-05,
+      "loss": 0.1529,
+      "step": 11601
+    },
+    {
+      "epoch": 0.185632,
+      "grad_norm": 1.3203125,
+      "learning_rate": 8.209516129032259e-05,
+      "loss": 0.2104,
+      "step": 11602
+    },
+    {
+      "epoch": 0.185648,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.209354838709677e-05,
+      "loss": 0.1844,
+      "step": 11603
+    },
+    {
+      "epoch": 0.185664,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.209193548387097e-05,
+      "loss": 0.1625,
+      "step": 11604
+    },
+    {
+      "epoch": 0.18568,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.209032258064516e-05,
+      "loss": 0.1856,
+      "step": 11605
+    },
+    {
+      "epoch": 0.185696,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.208870967741936e-05,
+      "loss": 0.1421,
+      "step": 11606
+    },
+    {
+      "epoch": 0.185712,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.208709677419355e-05,
+      "loss": 0.1959,
+      "step": 11607
+    },
+    {
+      "epoch": 0.185728,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.208548387096775e-05,
+      "loss": 0.2039,
+      "step": 11608
+    },
+    {
+      "epoch": 0.185744,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.208387096774193e-05,
+      "loss": 0.1697,
+      "step": 11609
+    },
+    {
+      "epoch": 0.18576,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.208225806451613e-05,
+      "loss": 0.1682,
+      "step": 11610
+    },
+    {
+      "epoch": 0.185776,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.208064516129033e-05,
+      "loss": 0.2089,
+      "step": 11611
+    },
+    {
+      "epoch": 0.185792,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.207903225806453e-05,
+      "loss": 0.2176,
+      "step": 11612
+    },
+    {
+      "epoch": 0.185808,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.207741935483872e-05,
+      "loss": 0.1667,
+      "step": 11613
+    },
+    {
+      "epoch": 0.185824,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.20758064516129e-05,
+      "loss": 0.1936,
+      "step": 11614
+    },
+    {
+      "epoch": 0.18584,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.20741935483871e-05,
+      "loss": 0.143,
+      "step": 11615
+    },
+    {
+      "epoch": 0.185856,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.207258064516129e-05,
+      "loss": 0.1762,
+      "step": 11616
+    },
+    {
+      "epoch": 0.185872,
+      "grad_norm": 1.234375,
+      "learning_rate": 8.207096774193549e-05,
+      "loss": 0.2172,
+      "step": 11617
+    },
+    {
+      "epoch": 0.185888,
+      "grad_norm": 1.4296875,
+      "learning_rate": 8.206935483870967e-05,
+      "loss": 0.1581,
+      "step": 11618
+    },
+    {
+      "epoch": 0.185904,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.206774193548387e-05,
+      "loss": 0.1741,
+      "step": 11619
+    },
+    {
+      "epoch": 0.18592,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.206612903225806e-05,
+      "loss": 0.1567,
+      "step": 11620
+    },
+    {
+      "epoch": 0.185936,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.206451612903226e-05,
+      "loss": 0.1237,
+      "step": 11621
+    },
+    {
+      "epoch": 0.185952,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.206290322580646e-05,
+      "loss": 0.165,
+      "step": 11622
+    },
+    {
+      "epoch": 0.185968,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.206129032258066e-05,
+      "loss": 0.1642,
+      "step": 11623
+    },
+    {
+      "epoch": 0.185984,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.205967741935485e-05,
+      "loss": 0.1941,
+      "step": 11624
+    },
+    {
+      "epoch": 0.186,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.205806451612904e-05,
+      "loss": 0.1424,
+      "step": 11625
+    },
+    {
+      "epoch": 0.186016,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.205645161290323e-05,
+      "loss": 0.1857,
+      "step": 11626
+    },
+    {
+      "epoch": 0.186032,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.205483870967743e-05,
+      "loss": 0.2087,
+      "step": 11627
+    },
+    {
+      "epoch": 0.186048,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.205322580645162e-05,
+      "loss": 0.1649,
+      "step": 11628
+    },
+    {
+      "epoch": 0.186064,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.20516129032258e-05,
+      "loss": 0.1612,
+      "step": 11629
+    },
+    {
+      "epoch": 0.18608,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.205e-05,
+      "loss": 0.21,
+      "step": 11630
+    },
+    {
+      "epoch": 0.186096,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.204838709677419e-05,
+      "loss": 0.1546,
+      "step": 11631
+    },
+    {
+      "epoch": 0.186112,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.204677419354839e-05,
+      "loss": 0.1803,
+      "step": 11632
+    },
+    {
+      "epoch": 0.186128,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.204516129032259e-05,
+      "loss": 0.1736,
+      "step": 11633
+    },
+    {
+      "epoch": 0.186144,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.204354838709677e-05,
+      "loss": 0.1804,
+      "step": 11634
+    },
+    {
+      "epoch": 0.18616,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.204193548387097e-05,
+      "loss": 0.1376,
+      "step": 11635
+    },
+    {
+      "epoch": 0.186176,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.204032258064517e-05,
+      "loss": 0.1532,
+      "step": 11636
+    },
+    {
+      "epoch": 0.186192,
+      "grad_norm": 1.40625,
+      "learning_rate": 8.203870967741936e-05,
+      "loss": 0.2074,
+      "step": 11637
+    },
+    {
+      "epoch": 0.186208,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.203709677419356e-05,
+      "loss": 0.1985,
+      "step": 11638
+    },
+    {
+      "epoch": 0.186224,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.203548387096774e-05,
+      "loss": 0.1481,
+      "step": 11639
+    },
+    {
+      "epoch": 0.18624,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.203387096774194e-05,
+      "loss": 0.1739,
+      "step": 11640
+    },
+    {
+      "epoch": 0.186256,
+      "grad_norm": 1.0,
+      "learning_rate": 8.203225806451613e-05,
+      "loss": 0.1629,
+      "step": 11641
+    },
+    {
+      "epoch": 0.186272,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.203064516129033e-05,
+      "loss": 0.1993,
+      "step": 11642
+    },
+    {
+      "epoch": 0.186288,
+      "grad_norm": 1.34375,
+      "learning_rate": 8.202903225806452e-05,
+      "loss": 0.2134,
+      "step": 11643
+    },
+    {
+      "epoch": 0.186304,
+      "grad_norm": 0.75,
+      "learning_rate": 8.20274193548387e-05,
+      "loss": 0.1293,
+      "step": 11644
+    },
+    {
+      "epoch": 0.18632,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.20258064516129e-05,
+      "loss": 0.1387,
+      "step": 11645
+    },
+    {
+      "epoch": 0.186336,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.20241935483871e-05,
+      "loss": 0.2035,
+      "step": 11646
+    },
+    {
+      "epoch": 0.186352,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.20225806451613e-05,
+      "loss": 0.1545,
+      "step": 11647
+    },
+    {
+      "epoch": 0.186368,
+      "grad_norm": 0.625,
+      "learning_rate": 8.202096774193549e-05,
+      "loss": 0.1483,
+      "step": 11648
+    },
+    {
+      "epoch": 0.186384,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.201935483870969e-05,
+      "loss": 0.18,
+      "step": 11649
+    },
+    {
+      "epoch": 0.1864,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.201774193548387e-05,
+      "loss": 0.157,
+      "step": 11650
+    },
+    {
+      "epoch": 0.186416,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.201612903225807e-05,
+      "loss": 0.1573,
+      "step": 11651
+    },
+    {
+      "epoch": 0.186432,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.201451612903226e-05,
+      "loss": 0.1672,
+      "step": 11652
+    },
+    {
+      "epoch": 0.186448,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.201290322580646e-05,
+      "loss": 0.1457,
+      "step": 11653
+    },
+    {
+      "epoch": 0.186464,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.201129032258064e-05,
+      "loss": 0.1229,
+      "step": 11654
+    },
+    {
+      "epoch": 0.18648,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.200967741935484e-05,
+      "loss": 0.1674,
+      "step": 11655
+    },
+    {
+      "epoch": 0.186496,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.200806451612903e-05,
+      "loss": 0.1675,
+      "step": 11656
+    },
+    {
+      "epoch": 0.186512,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.200645161290323e-05,
+      "loss": 0.1831,
+      "step": 11657
+    },
+    {
+      "epoch": 0.186528,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.200483870967743e-05,
+      "loss": 0.1677,
+      "step": 11658
+    },
+    {
+      "epoch": 0.186544,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.200322580645163e-05,
+      "loss": 0.1918,
+      "step": 11659
+    },
+    {
+      "epoch": 0.18656,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.200161290322581e-05,
+      "loss": 0.1648,
+      "step": 11660
+    },
+    {
+      "epoch": 0.186576,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.2e-05,
+      "loss": 0.1634,
+      "step": 11661
+    },
+    {
+      "epoch": 0.186592,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.19983870967742e-05,
+      "loss": 0.1748,
+      "step": 11662
+    },
+    {
+      "epoch": 0.186608,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.199677419354839e-05,
+      "loss": 0.1899,
+      "step": 11663
+    },
+    {
+      "epoch": 0.186624,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.199516129032259e-05,
+      "loss": 0.136,
+      "step": 11664
+    },
+    {
+      "epoch": 0.18664,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.199354838709677e-05,
+      "loss": 0.1825,
+      "step": 11665
+    },
+    {
+      "epoch": 0.186656,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.199193548387097e-05,
+      "loss": 0.1611,
+      "step": 11666
+    },
+    {
+      "epoch": 0.186672,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.199032258064516e-05,
+      "loss": 0.1688,
+      "step": 11667
+    },
+    {
+      "epoch": 0.186688,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.198870967741936e-05,
+      "loss": 0.183,
+      "step": 11668
+    },
+    {
+      "epoch": 0.186704,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.198709677419354e-05,
+      "loss": 0.1768,
+      "step": 11669
+    },
+    {
+      "epoch": 0.18672,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.198548387096774e-05,
+      "loss": 0.2067,
+      "step": 11670
+    },
+    {
+      "epoch": 0.186736,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.198387096774194e-05,
+      "loss": 0.1728,
+      "step": 11671
+    },
+    {
+      "epoch": 0.186752,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.198225806451614e-05,
+      "loss": 0.1603,
+      "step": 11672
+    },
+    {
+      "epoch": 0.186768,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.198064516129033e-05,
+      "loss": 0.1483,
+      "step": 11673
+    },
+    {
+      "epoch": 0.186784,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.197903225806453e-05,
+      "loss": 0.2117,
+      "step": 11674
+    },
+    {
+      "epoch": 0.1868,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.197741935483871e-05,
+      "loss": 0.1893,
+      "step": 11675
+    },
+    {
+      "epoch": 0.186816,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.19758064516129e-05,
+      "loss": 0.2035,
+      "step": 11676
+    },
+    {
+      "epoch": 0.186832,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.19741935483871e-05,
+      "loss": 0.1636,
+      "step": 11677
+    },
+    {
+      "epoch": 0.186848,
+      "grad_norm": 1.2890625,
+      "learning_rate": 8.197258064516129e-05,
+      "loss": 0.1791,
+      "step": 11678
+    },
+    {
+      "epoch": 0.186864,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.197096774193549e-05,
+      "loss": 0.1845,
+      "step": 11679
+    },
+    {
+      "epoch": 0.18688,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.196935483870967e-05,
+      "loss": 0.1476,
+      "step": 11680
+    },
+    {
+      "epoch": 0.186896,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.196774193548387e-05,
+      "loss": 0.1708,
+      "step": 11681
+    },
+    {
+      "epoch": 0.186912,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.196612903225807e-05,
+      "loss": 0.1857,
+      "step": 11682
+    },
+    {
+      "epoch": 0.186928,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.196451612903227e-05,
+      "loss": 0.1419,
+      "step": 11683
+    },
+    {
+      "epoch": 0.186944,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.196290322580646e-05,
+      "loss": 0.1359,
+      "step": 11684
+    },
+    {
+      "epoch": 0.18696,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.196129032258066e-05,
+      "loss": 0.1957,
+      "step": 11685
+    },
+    {
+      "epoch": 0.186976,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.195967741935484e-05,
+      "loss": 0.1399,
+      "step": 11686
+    },
+    {
+      "epoch": 0.186992,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.195806451612904e-05,
+      "loss": 0.1828,
+      "step": 11687
+    },
+    {
+      "epoch": 0.187008,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.195645161290323e-05,
+      "loss": 0.1813,
+      "step": 11688
+    },
+    {
+      "epoch": 0.187024,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.195483870967743e-05,
+      "loss": 0.1547,
+      "step": 11689
+    },
+    {
+      "epoch": 0.18704,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.195322580645161e-05,
+      "loss": 0.1792,
+      "step": 11690
+    },
+    {
+      "epoch": 0.187056,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.19516129032258e-05,
+      "loss": 0.2117,
+      "step": 11691
+    },
+    {
+      "epoch": 0.187072,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.195e-05,
+      "loss": 0.1753,
+      "step": 11692
+    },
+    {
+      "epoch": 0.187088,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.19483870967742e-05,
+      "loss": 0.1603,
+      "step": 11693
+    },
+    {
+      "epoch": 0.187104,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.19467741935484e-05,
+      "loss": 0.1603,
+      "step": 11694
+    },
+    {
+      "epoch": 0.18712,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.194516129032259e-05,
+      "loss": 0.1848,
+      "step": 11695
+    },
+    {
+      "epoch": 0.187136,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.194354838709678e-05,
+      "loss": 0.1654,
+      "step": 11696
+    },
+    {
+      "epoch": 0.187152,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.194193548387097e-05,
+      "loss": 0.2095,
+      "step": 11697
+    },
+    {
+      "epoch": 0.187168,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.194032258064517e-05,
+      "loss": 0.1998,
+      "step": 11698
+    },
+    {
+      "epoch": 0.187184,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.193870967741936e-05,
+      "loss": 0.231,
+      "step": 11699
+    },
+    {
+      "epoch": 0.1872,
+      "grad_norm": 0.75,
+      "learning_rate": 8.193709677419356e-05,
+      "loss": 0.1752,
+      "step": 11700
+    },
+    {
+      "epoch": 0.187216,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.193548387096774e-05,
+      "loss": 0.1803,
+      "step": 11701
+    },
+    {
+      "epoch": 0.187232,
+      "grad_norm": 1.359375,
+      "learning_rate": 8.193387096774194e-05,
+      "loss": 0.2039,
+      "step": 11702
+    },
+    {
+      "epoch": 0.187248,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.193225806451613e-05,
+      "loss": 0.1623,
+      "step": 11703
+    },
+    {
+      "epoch": 0.187264,
+      "grad_norm": 1.453125,
+      "learning_rate": 8.193064516129033e-05,
+      "loss": 0.1541,
+      "step": 11704
+    },
+    {
+      "epoch": 0.18728,
+      "grad_norm": 1.359375,
+      "learning_rate": 8.192903225806451e-05,
+      "loss": 0.1754,
+      "step": 11705
+    },
+    {
+      "epoch": 0.187296,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.192741935483871e-05,
+      "loss": 0.1374,
+      "step": 11706
+    },
+    {
+      "epoch": 0.187312,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.192580645161291e-05,
+      "loss": 0.1472,
+      "step": 11707
+    },
+    {
+      "epoch": 0.187328,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.19241935483871e-05,
+      "loss": 0.1781,
+      "step": 11708
+    },
+    {
+      "epoch": 0.187344,
+      "grad_norm": 1.3046875,
+      "learning_rate": 8.19225806451613e-05,
+      "loss": 0.195,
+      "step": 11709
+    },
+    {
+      "epoch": 0.18736,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.192096774193548e-05,
+      "loss": 0.179,
+      "step": 11710
+    },
+    {
+      "epoch": 0.187376,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.191935483870968e-05,
+      "loss": 0.1549,
+      "step": 11711
+    },
+    {
+      "epoch": 0.187392,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.191774193548387e-05,
+      "loss": 0.194,
+      "step": 11712
+    },
+    {
+      "epoch": 0.187408,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.191612903225807e-05,
+      "loss": 0.1853,
+      "step": 11713
+    },
+    {
+      "epoch": 0.187424,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.191451612903226e-05,
+      "loss": 0.2266,
+      "step": 11714
+    },
+    {
+      "epoch": 0.18744,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.191290322580646e-05,
+      "loss": 0.1536,
+      "step": 11715
+    },
+    {
+      "epoch": 0.187456,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.191129032258064e-05,
+      "loss": 0.1878,
+      "step": 11716
+    },
+    {
+      "epoch": 0.187472,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.190967741935484e-05,
+      "loss": 0.1377,
+      "step": 11717
+    },
+    {
+      "epoch": 0.187488,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.190806451612904e-05,
+      "loss": 0.2064,
+      "step": 11718
+    },
+    {
+      "epoch": 0.187504,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.190645161290324e-05,
+      "loss": 0.1775,
+      "step": 11719
+    },
+    {
+      "epoch": 0.18752,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.190483870967743e-05,
+      "loss": 0.1258,
+      "step": 11720
+    },
+    {
+      "epoch": 0.187536,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.190322580645163e-05,
+      "loss": 0.1395,
+      "step": 11721
+    },
+    {
+      "epoch": 0.187552,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.190161290322581e-05,
+      "loss": 0.186,
+      "step": 11722
+    },
+    {
+      "epoch": 0.187568,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.19e-05,
+      "loss": 0.1475,
+      "step": 11723
+    },
+    {
+      "epoch": 0.187584,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.18983870967742e-05,
+      "loss": 0.1525,
+      "step": 11724
+    },
+    {
+      "epoch": 0.1876,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.189677419354838e-05,
+      "loss": 0.1505,
+      "step": 11725
+    },
+    {
+      "epoch": 0.187616,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.189516129032258e-05,
+      "loss": 0.1794,
+      "step": 11726
+    },
+    {
+      "epoch": 0.187632,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.189354838709677e-05,
+      "loss": 0.1213,
+      "step": 11727
+    },
+    {
+      "epoch": 0.187648,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.189193548387097e-05,
+      "loss": 0.1905,
+      "step": 11728
+    },
+    {
+      "epoch": 0.187664,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.189032258064516e-05,
+      "loss": 0.181,
+      "step": 11729
+    },
+    {
+      "epoch": 0.18768,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.188870967741936e-05,
+      "loss": 0.1645,
+      "step": 11730
+    },
+    {
+      "epoch": 0.187696,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.188709677419355e-05,
+      "loss": 0.1501,
+      "step": 11731
+    },
+    {
+      "epoch": 0.187712,
+      "grad_norm": 1.25,
+      "learning_rate": 8.188548387096775e-05,
+      "loss": 0.187,
+      "step": 11732
+    },
+    {
+      "epoch": 0.187728,
+      "grad_norm": 1.2109375,
+      "learning_rate": 8.188387096774194e-05,
+      "loss": 0.2004,
+      "step": 11733
+    },
+    {
+      "epoch": 0.187744,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.188225806451614e-05,
+      "loss": 0.1543,
+      "step": 11734
+    },
+    {
+      "epoch": 0.18776,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.188064516129033e-05,
+      "loss": 0.1682,
+      "step": 11735
+    },
+    {
+      "epoch": 0.187776,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.187903225806453e-05,
+      "loss": 0.1689,
+      "step": 11736
+    },
+    {
+      "epoch": 0.187792,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.187741935483871e-05,
+      "loss": 0.1619,
+      "step": 11737
+    },
+    {
+      "epoch": 0.187808,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.18758064516129e-05,
+      "loss": 0.178,
+      "step": 11738
+    },
+    {
+      "epoch": 0.187824,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.18741935483871e-05,
+      "loss": 0.1647,
+      "step": 11739
+    },
+    {
+      "epoch": 0.18784,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.187258064516128e-05,
+      "loss": 0.193,
+      "step": 11740
+    },
+    {
+      "epoch": 0.187856,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.187096774193548e-05,
+      "loss": 0.1792,
+      "step": 11741
+    },
+    {
+      "epoch": 0.187872,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.186935483870968e-05,
+      "loss": 0.1995,
+      "step": 11742
+    },
+    {
+      "epoch": 0.187888,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.186774193548388e-05,
+      "loss": 0.1553,
+      "step": 11743
+    },
+    {
+      "epoch": 0.187904,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.186612903225807e-05,
+      "loss": 0.1789,
+      "step": 11744
+    },
+    {
+      "epoch": 0.18792,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.186451612903227e-05,
+      "loss": 0.1995,
+      "step": 11745
+    },
+    {
+      "epoch": 0.187936,
+      "grad_norm": 1.375,
+      "learning_rate": 8.186290322580645e-05,
+      "loss": 0.2142,
+      "step": 11746
+    },
+    {
+      "epoch": 0.187952,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.186129032258065e-05,
+      "loss": 0.1963,
+      "step": 11747
+    },
+    {
+      "epoch": 0.187968,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.185967741935484e-05,
+      "loss": 0.1773,
+      "step": 11748
+    },
+    {
+      "epoch": 0.187984,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.185806451612904e-05,
+      "loss": 0.1496,
+      "step": 11749
+    },
+    {
+      "epoch": 0.188,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.185645161290323e-05,
+      "loss": 0.1558,
+      "step": 11750
+    },
+    {
+      "epoch": 0.188016,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.185483870967743e-05,
+      "loss": 0.2001,
+      "step": 11751
+    },
+    {
+      "epoch": 0.188032,
+      "grad_norm": 1.25,
+      "learning_rate": 8.185322580645161e-05,
+      "loss": 0.2103,
+      "step": 11752
+    },
+    {
+      "epoch": 0.188048,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.185161290322581e-05,
+      "loss": 0.2269,
+      "step": 11753
+    },
+    {
+      "epoch": 0.188064,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.185000000000001e-05,
+      "loss": 0.1663,
+      "step": 11754
+    },
+    {
+      "epoch": 0.18808,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.18483870967742e-05,
+      "loss": 0.1757,
+      "step": 11755
+    },
+    {
+      "epoch": 0.188096,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.18467741935484e-05,
+      "loss": 0.1994,
+      "step": 11756
+    },
+    {
+      "epoch": 0.188112,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.184516129032258e-05,
+      "loss": 0.2263,
+      "step": 11757
+    },
+    {
+      "epoch": 0.188128,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.184354838709678e-05,
+      "loss": 0.1757,
+      "step": 11758
+    },
+    {
+      "epoch": 0.188144,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.184193548387097e-05,
+      "loss": 0.1693,
+      "step": 11759
+    },
+    {
+      "epoch": 0.18816,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.184032258064517e-05,
+      "loss": 0.2191,
+      "step": 11760
+    },
+    {
+      "epoch": 0.188176,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.183870967741935e-05,
+      "loss": 0.1643,
+      "step": 11761
+    },
+    {
+      "epoch": 0.188192,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.183709677419355e-05,
+      "loss": 0.1999,
+      "step": 11762
+    },
+    {
+      "epoch": 0.188208,
+      "grad_norm": 1.328125,
+      "learning_rate": 8.183548387096774e-05,
+      "loss": 0.1445,
+      "step": 11763
+    },
+    {
+      "epoch": 0.188224,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.183387096774194e-05,
+      "loss": 0.1916,
+      "step": 11764
+    },
+    {
+      "epoch": 0.18824,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.183225806451613e-05,
+      "loss": 0.138,
+      "step": 11765
+    },
+    {
+      "epoch": 0.188256,
+      "grad_norm": 1.3828125,
+      "learning_rate": 8.183064516129033e-05,
+      "loss": 0.1942,
+      "step": 11766
+    },
+    {
+      "epoch": 0.188272,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.182903225806452e-05,
+      "loss": 0.1834,
+      "step": 11767
+    },
+    {
+      "epoch": 0.188288,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.182741935483872e-05,
+      "loss": 0.1724,
+      "step": 11768
+    },
+    {
+      "epoch": 0.188304,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.182580645161291e-05,
+      "loss": 0.1566,
+      "step": 11769
+    },
+    {
+      "epoch": 0.18832,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.18241935483871e-05,
+      "loss": 0.1905,
+      "step": 11770
+    },
+    {
+      "epoch": 0.188336,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.18225806451613e-05,
+      "loss": 0.1889,
+      "step": 11771
+    },
+    {
+      "epoch": 0.188352,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.182096774193548e-05,
+      "loss": 0.1641,
+      "step": 11772
+    },
+    {
+      "epoch": 0.188368,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.181935483870968e-05,
+      "loss": 0.1671,
+      "step": 11773
+    },
+    {
+      "epoch": 0.188384,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.181774193548387e-05,
+      "loss": 0.1761,
+      "step": 11774
+    },
+    {
+      "epoch": 0.1884,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.181612903225807e-05,
+      "loss": 0.1948,
+      "step": 11775
+    },
+    {
+      "epoch": 0.188416,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.181451612903225e-05,
+      "loss": 0.1615,
+      "step": 11776
+    },
+    {
+      "epoch": 0.188432,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.181290322580645e-05,
+      "loss": 0.1519,
+      "step": 11777
+    },
+    {
+      "epoch": 0.188448,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.181129032258065e-05,
+      "loss": 0.166,
+      "step": 11778
+    },
+    {
+      "epoch": 0.188464,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.180967741935485e-05,
+      "loss": 0.1458,
+      "step": 11779
+    },
+    {
+      "epoch": 0.18848,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.180806451612904e-05,
+      "loss": 0.1725,
+      "step": 11780
+    },
+    {
+      "epoch": 0.188496,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.180645161290324e-05,
+      "loss": 0.1849,
+      "step": 11781
+    },
+    {
+      "epoch": 0.188512,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.180483870967742e-05,
+      "loss": 0.1583,
+      "step": 11782
+    },
+    {
+      "epoch": 0.188528,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.180322580645162e-05,
+      "loss": 0.1685,
+      "step": 11783
+    },
+    {
+      "epoch": 0.188544,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.180161290322581e-05,
+      "loss": 0.1662,
+      "step": 11784
+    },
+    {
+      "epoch": 0.18856,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.18e-05,
+      "loss": 0.1331,
+      "step": 11785
+    },
+    {
+      "epoch": 0.188576,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.17983870967742e-05,
+      "loss": 0.1492,
+      "step": 11786
+    },
+    {
+      "epoch": 0.188592,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.179677419354838e-05,
+      "loss": 0.1427,
+      "step": 11787
+    },
+    {
+      "epoch": 0.188608,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.179516129032258e-05,
+      "loss": 0.1588,
+      "step": 11788
+    },
+    {
+      "epoch": 0.188624,
+      "grad_norm": 1.5078125,
+      "learning_rate": 8.179354838709678e-05,
+      "loss": 0.1787,
+      "step": 11789
+    },
+    {
+      "epoch": 0.18864,
+      "grad_norm": 0.75,
+      "learning_rate": 8.179193548387098e-05,
+      "loss": 0.1879,
+      "step": 11790
+    },
+    {
+      "epoch": 0.188656,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.179032258064517e-05,
+      "loss": 0.1767,
+      "step": 11791
+    },
+    {
+      "epoch": 0.188672,
+      "grad_norm": 1.328125,
+      "learning_rate": 8.178870967741937e-05,
+      "loss": 0.1727,
+      "step": 11792
+    },
+    {
+      "epoch": 0.188688,
+      "grad_norm": 1.2109375,
+      "learning_rate": 8.178709677419355e-05,
+      "loss": 0.1867,
+      "step": 11793
+    },
+    {
+      "epoch": 0.188704,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.178548387096775e-05,
+      "loss": 0.1711,
+      "step": 11794
+    },
+    {
+      "epoch": 0.18872,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.178387096774194e-05,
+      "loss": 0.1545,
+      "step": 11795
+    },
+    {
+      "epoch": 0.188736,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.178225806451614e-05,
+      "loss": 0.1861,
+      "step": 11796
+    },
+    {
+      "epoch": 0.188752,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.178064516129032e-05,
+      "loss": 0.1624,
+      "step": 11797
+    },
+    {
+      "epoch": 0.188768,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.177903225806452e-05,
+      "loss": 0.143,
+      "step": 11798
+    },
+    {
+      "epoch": 0.188784,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.177741935483871e-05,
+      "loss": 0.185,
+      "step": 11799
+    },
+    {
+      "epoch": 0.1888,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.17758064516129e-05,
+      "loss": 0.1613,
+      "step": 11800
+    },
+    {
+      "epoch": 0.188816,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.17741935483871e-05,
+      "loss": 0.2045,
+      "step": 11801
+    },
+    {
+      "epoch": 0.188832,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.17725806451613e-05,
+      "loss": 0.1763,
+      "step": 11802
+    },
+    {
+      "epoch": 0.188848,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.17709677419355e-05,
+      "loss": 0.1717,
+      "step": 11803
+    },
+    {
+      "epoch": 0.188864,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.176935483870968e-05,
+      "loss": 0.1842,
+      "step": 11804
+    },
+    {
+      "epoch": 0.18888,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.176774193548388e-05,
+      "loss": 0.1532,
+      "step": 11805
+    },
+    {
+      "epoch": 0.188896,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.176612903225807e-05,
+      "loss": 0.1883,
+      "step": 11806
+    },
+    {
+      "epoch": 0.188912,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.176451612903227e-05,
+      "loss": 0.1786,
+      "step": 11807
+    },
+    {
+      "epoch": 0.188928,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.176290322580645e-05,
+      "loss": 0.1761,
+      "step": 11808
+    },
+    {
+      "epoch": 0.188944,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.176129032258065e-05,
+      "loss": 0.212,
+      "step": 11809
+    },
+    {
+      "epoch": 0.18896,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.175967741935484e-05,
+      "loss": 0.1688,
+      "step": 11810
+    },
+    {
+      "epoch": 0.188976,
+      "grad_norm": 1.0,
+      "learning_rate": 8.175806451612904e-05,
+      "loss": 0.1846,
+      "step": 11811
+    },
+    {
+      "epoch": 0.188992,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.175645161290322e-05,
+      "loss": 0.1667,
+      "step": 11812
+    },
+    {
+      "epoch": 0.189008,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.175483870967742e-05,
+      "loss": 0.1782,
+      "step": 11813
+    },
+    {
+      "epoch": 0.189024,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.175322580645162e-05,
+      "loss": 0.1796,
+      "step": 11814
+    },
+    {
+      "epoch": 0.18904,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.175161290322581e-05,
+      "loss": 0.1626,
+      "step": 11815
+    },
+    {
+      "epoch": 0.189056,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.175000000000001e-05,
+      "loss": 0.1419,
+      "step": 11816
+    },
+    {
+      "epoch": 0.189072,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.17483870967742e-05,
+      "loss": 0.2111,
+      "step": 11817
+    },
+    {
+      "epoch": 0.189088,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.17467741935484e-05,
+      "loss": 0.1743,
+      "step": 11818
+    },
+    {
+      "epoch": 0.189104,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.174516129032258e-05,
+      "loss": 0.1832,
+      "step": 11819
+    },
+    {
+      "epoch": 0.18912,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.174354838709678e-05,
+      "loss": 0.148,
+      "step": 11820
+    },
+    {
+      "epoch": 0.189136,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.174193548387097e-05,
+      "loss": 0.1555,
+      "step": 11821
+    },
+    {
+      "epoch": 0.189152,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.174032258064517e-05,
+      "loss": 0.1527,
+      "step": 11822
+    },
+    {
+      "epoch": 0.189168,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.173870967741935e-05,
+      "loss": 0.1813,
+      "step": 11823
+    },
+    {
+      "epoch": 0.189184,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.173709677419355e-05,
+      "loss": 0.1921,
+      "step": 11824
+    },
+    {
+      "epoch": 0.1892,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.173548387096774e-05,
+      "loss": 0.1319,
+      "step": 11825
+    },
+    {
+      "epoch": 0.189216,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.173387096774194e-05,
+      "loss": 0.1814,
+      "step": 11826
+    },
+    {
+      "epoch": 0.189232,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.173225806451614e-05,
+      "loss": 0.1808,
+      "step": 11827
+    },
+    {
+      "epoch": 0.189248,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.173064516129034e-05,
+      "loss": 0.1779,
+      "step": 11828
+    },
+    {
+      "epoch": 0.189264,
+      "grad_norm": 1.3203125,
+      "learning_rate": 8.172903225806452e-05,
+      "loss": 0.1779,
+      "step": 11829
+    },
+    {
+      "epoch": 0.18928,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.172741935483872e-05,
+      "loss": 0.1516,
+      "step": 11830
+    },
+    {
+      "epoch": 0.189296,
+      "grad_norm": 1.0,
+      "learning_rate": 8.172580645161291e-05,
+      "loss": 0.1968,
+      "step": 11831
+    },
+    {
+      "epoch": 0.189312,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.17241935483871e-05,
+      "loss": 0.1938,
+      "step": 11832
+    },
+    {
+      "epoch": 0.189328,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.17225806451613e-05,
+      "loss": 0.174,
+      "step": 11833
+    },
+    {
+      "epoch": 0.189344,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.172096774193548e-05,
+      "loss": 0.1553,
+      "step": 11834
+    },
+    {
+      "epoch": 0.18936,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.171935483870968e-05,
+      "loss": 0.1979,
+      "step": 11835
+    },
+    {
+      "epoch": 0.189376,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.171774193548387e-05,
+      "loss": 0.1997,
+      "step": 11836
+    },
+    {
+      "epoch": 0.189392,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.171612903225807e-05,
+      "loss": 0.1593,
+      "step": 11837
+    },
+    {
+      "epoch": 0.189408,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.171451612903226e-05,
+      "loss": 0.1684,
+      "step": 11838
+    },
+    {
+      "epoch": 0.189424,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.171290322580646e-05,
+      "loss": 0.2207,
+      "step": 11839
+    },
+    {
+      "epoch": 0.18944,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.171129032258065e-05,
+      "loss": 0.143,
+      "step": 11840
+    },
+    {
+      "epoch": 0.189456,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.170967741935485e-05,
+      "loss": 0.2107,
+      "step": 11841
+    },
+    {
+      "epoch": 0.189472,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.170806451612904e-05,
+      "loss": 0.1406,
+      "step": 11842
+    },
+    {
+      "epoch": 0.189488,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.170645161290324e-05,
+      "loss": 0.1632,
+      "step": 11843
+    },
+    {
+      "epoch": 0.189504,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.170483870967742e-05,
+      "loss": 0.194,
+      "step": 11844
+    },
+    {
+      "epoch": 0.18952,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.170322580645162e-05,
+      "loss": 0.1515,
+      "step": 11845
+    },
+    {
+      "epoch": 0.189536,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.170161290322581e-05,
+      "loss": 0.1596,
+      "step": 11846
+    },
+    {
+      "epoch": 0.189552,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.17e-05,
+      "loss": 0.1616,
+      "step": 11847
+    },
+    {
+      "epoch": 0.189568,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.16983870967742e-05,
+      "loss": 0.1903,
+      "step": 11848
+    },
+    {
+      "epoch": 0.189584,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.169677419354839e-05,
+      "loss": 0.1855,
+      "step": 11849
+    },
+    {
+      "epoch": 0.1896,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.169516129032259e-05,
+      "loss": 0.1769,
+      "step": 11850
+    },
+    {
+      "epoch": 0.189616,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.169354838709678e-05,
+      "loss": 0.1889,
+      "step": 11851
+    },
+    {
+      "epoch": 0.189632,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.169193548387098e-05,
+      "loss": 0.2284,
+      "step": 11852
+    },
+    {
+      "epoch": 0.189648,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.169032258064516e-05,
+      "loss": 0.1745,
+      "step": 11853
+    },
+    {
+      "epoch": 0.189664,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.168870967741936e-05,
+      "loss": 0.1863,
+      "step": 11854
+    },
+    {
+      "epoch": 0.18968,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.168709677419355e-05,
+      "loss": 0.1375,
+      "step": 11855
+    },
+    {
+      "epoch": 0.189696,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.168548387096775e-05,
+      "loss": 0.1741,
+      "step": 11856
+    },
+    {
+      "epoch": 0.189712,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.168387096774194e-05,
+      "loss": 0.1357,
+      "step": 11857
+    },
+    {
+      "epoch": 0.189728,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.168225806451614e-05,
+      "loss": 0.1918,
+      "step": 11858
+    },
+    {
+      "epoch": 0.189744,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.168064516129032e-05,
+      "loss": 0.1886,
+      "step": 11859
+    },
+    {
+      "epoch": 0.18976,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.167903225806452e-05,
+      "loss": 0.1799,
+      "step": 11860
+    },
+    {
+      "epoch": 0.189776,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.167741935483871e-05,
+      "loss": 0.1738,
+      "step": 11861
+    },
+    {
+      "epoch": 0.189792,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.167580645161291e-05,
+      "loss": 0.1568,
+      "step": 11862
+    },
+    {
+      "epoch": 0.189808,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.16741935483871e-05,
+      "loss": 0.1549,
+      "step": 11863
+    },
+    {
+      "epoch": 0.189824,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.167258064516129e-05,
+      "loss": 0.1727,
+      "step": 11864
+    },
+    {
+      "epoch": 0.18984,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.167096774193549e-05,
+      "loss": 0.1692,
+      "step": 11865
+    },
+    {
+      "epoch": 0.189856,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.166935483870968e-05,
+      "loss": 0.1662,
+      "step": 11866
+    },
+    {
+      "epoch": 0.189872,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.166774193548388e-05,
+      "loss": 0.156,
+      "step": 11867
+    },
+    {
+      "epoch": 0.189888,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.166612903225806e-05,
+      "loss": 0.2154,
+      "step": 11868
+    },
+    {
+      "epoch": 0.189904,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.166451612903226e-05,
+      "loss": 0.1458,
+      "step": 11869
+    },
+    {
+      "epoch": 0.18992,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.166290322580645e-05,
+      "loss": 0.1763,
+      "step": 11870
+    },
+    {
+      "epoch": 0.189936,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.166129032258065e-05,
+      "loss": 0.1738,
+      "step": 11871
+    },
+    {
+      "epoch": 0.189952,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.165967741935484e-05,
+      "loss": 0.1282,
+      "step": 11872
+    },
+    {
+      "epoch": 0.189968,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.165806451612903e-05,
+      "loss": 0.1729,
+      "step": 11873
+    },
+    {
+      "epoch": 0.189984,
+      "grad_norm": 1.125,
+      "learning_rate": 8.165645161290323e-05,
+      "loss": 0.1918,
+      "step": 11874
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.165483870967743e-05,
+      "loss": 0.1431,
+      "step": 11875
+    },
+    {
+      "epoch": 0.190016,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.165322580645162e-05,
+      "loss": 0.1337,
+      "step": 11876
+    },
+    {
+      "epoch": 0.190032,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.165161290322582e-05,
+      "loss": 0.1308,
+      "step": 11877
+    },
+    {
+      "epoch": 0.190048,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.165e-05,
+      "loss": 0.1978,
+      "step": 11878
+    },
+    {
+      "epoch": 0.190064,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.164838709677419e-05,
+      "loss": 0.1601,
+      "step": 11879
+    },
+    {
+      "epoch": 0.19008,
+      "grad_norm": 1.0,
+      "learning_rate": 8.164677419354839e-05,
+      "loss": 0.1987,
+      "step": 11880
+    },
+    {
+      "epoch": 0.190096,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.164516129032258e-05,
+      "loss": 0.1673,
+      "step": 11881
+    },
+    {
+      "epoch": 0.190112,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.164354838709678e-05,
+      "loss": 0.1659,
+      "step": 11882
+    },
+    {
+      "epoch": 0.190128,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.164193548387096e-05,
+      "loss": 0.1569,
+      "step": 11883
+    },
+    {
+      "epoch": 0.190144,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.164032258064516e-05,
+      "loss": 0.1894,
+      "step": 11884
+    },
+    {
+      "epoch": 0.19016,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.163870967741936e-05,
+      "loss": 0.2247,
+      "step": 11885
+    },
+    {
+      "epoch": 0.190176,
+      "grad_norm": 1.2578125,
+      "learning_rate": 8.163709677419355e-05,
+      "loss": 0.2484,
+      "step": 11886
+    },
+    {
+      "epoch": 0.190192,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.163548387096775e-05,
+      "loss": 0.1584,
+      "step": 11887
+    },
+    {
+      "epoch": 0.190208,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.163387096774195e-05,
+      "loss": 0.1516,
+      "step": 11888
+    },
+    {
+      "epoch": 0.190224,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.163225806451613e-05,
+      "loss": 0.1572,
+      "step": 11889
+    },
+    {
+      "epoch": 0.19024,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.163064516129033e-05,
+      "loss": 0.1681,
+      "step": 11890
+    },
+    {
+      "epoch": 0.190256,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.162903225806452e-05,
+      "loss": 0.1858,
+      "step": 11891
+    },
+    {
+      "epoch": 0.190272,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.162741935483872e-05,
+      "loss": 0.1856,
+      "step": 11892
+    },
+    {
+      "epoch": 0.190288,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.16258064516129e-05,
+      "loss": 0.2081,
+      "step": 11893
+    },
+    {
+      "epoch": 0.190304,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.162419354838709e-05,
+      "loss": 0.1729,
+      "step": 11894
+    },
+    {
+      "epoch": 0.19032,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.162258064516129e-05,
+      "loss": 0.14,
+      "step": 11895
+    },
+    {
+      "epoch": 0.190336,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.162096774193548e-05,
+      "loss": 0.209,
+      "step": 11896
+    },
+    {
+      "epoch": 0.190352,
+      "grad_norm": 1.125,
+      "learning_rate": 8.161935483870968e-05,
+      "loss": 0.1572,
+      "step": 11897
+    },
+    {
+      "epoch": 0.190368,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.161774193548388e-05,
+      "loss": 0.1597,
+      "step": 11898
+    },
+    {
+      "epoch": 0.190384,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.161612903225808e-05,
+      "loss": 0.186,
+      "step": 11899
+    },
+    {
+      "epoch": 0.1904,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.161451612903226e-05,
+      "loss": 0.1794,
+      "step": 11900
+    },
+    {
+      "epoch": 0.190416,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.161290322580646e-05,
+      "loss": 0.1776,
+      "step": 11901
+    },
+    {
+      "epoch": 0.190432,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.161129032258065e-05,
+      "loss": 0.1443,
+      "step": 11902
+    },
+    {
+      "epoch": 0.190448,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.160967741935485e-05,
+      "loss": 0.1626,
+      "step": 11903
+    },
+    {
+      "epoch": 0.190464,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.160806451612903e-05,
+      "loss": 0.2084,
+      "step": 11904
+    },
+    {
+      "epoch": 0.19048,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.160645161290323e-05,
+      "loss": 0.1847,
+      "step": 11905
+    },
+    {
+      "epoch": 0.190496,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.160483870967742e-05,
+      "loss": 0.1838,
+      "step": 11906
+    },
+    {
+      "epoch": 0.190512,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.160322580645162e-05,
+      "loss": 0.1841,
+      "step": 11907
+    },
+    {
+      "epoch": 0.190528,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.16016129032258e-05,
+      "loss": 0.143,
+      "step": 11908
+    },
+    {
+      "epoch": 0.190544,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.16e-05,
+      "loss": 0.2129,
+      "step": 11909
+    },
+    {
+      "epoch": 0.19056,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.15983870967742e-05,
+      "loss": 0.1723,
+      "step": 11910
+    },
+    {
+      "epoch": 0.190576,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.159677419354839e-05,
+      "loss": 0.1284,
+      "step": 11911
+    },
+    {
+      "epoch": 0.190592,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.159516129032259e-05,
+      "loss": 0.2074,
+      "step": 11912
+    },
+    {
+      "epoch": 0.190608,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.159354838709678e-05,
+      "loss": 0.1905,
+      "step": 11913
+    },
+    {
+      "epoch": 0.190624,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.159193548387098e-05,
+      "loss": 0.1929,
+      "step": 11914
+    },
+    {
+      "epoch": 0.19064,
+      "grad_norm": 1.359375,
+      "learning_rate": 8.159032258064516e-05,
+      "loss": 0.1772,
+      "step": 11915
+    },
+    {
+      "epoch": 0.190656,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.158870967741936e-05,
+      "loss": 0.1642,
+      "step": 11916
+    },
+    {
+      "epoch": 0.190672,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.158709677419355e-05,
+      "loss": 0.1563,
+      "step": 11917
+    },
+    {
+      "epoch": 0.190688,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.158548387096775e-05,
+      "loss": 0.1964,
+      "step": 11918
+    },
+    {
+      "epoch": 0.190704,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.158387096774193e-05,
+      "loss": 0.1582,
+      "step": 11919
+    },
+    {
+      "epoch": 0.19072,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.158225806451613e-05,
+      "loss": 0.2306,
+      "step": 11920
+    },
+    {
+      "epoch": 0.190736,
+      "grad_norm": 0.4921875,
+      "learning_rate": 8.158064516129032e-05,
+      "loss": 0.1567,
+      "step": 11921
+    },
+    {
+      "epoch": 0.190752,
+      "grad_norm": 0.625,
+      "learning_rate": 8.157903225806452e-05,
+      "loss": 0.2066,
+      "step": 11922
+    },
+    {
+      "epoch": 0.190768,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.157741935483872e-05,
+      "loss": 0.1827,
+      "step": 11923
+    },
+    {
+      "epoch": 0.190784,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.15758064516129e-05,
+      "loss": 0.1495,
+      "step": 11924
+    },
+    {
+      "epoch": 0.1908,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.15741935483871e-05,
+      "loss": 0.1605,
+      "step": 11925
+    },
+    {
+      "epoch": 0.190816,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.157258064516129e-05,
+      "loss": 0.1725,
+      "step": 11926
+    },
+    {
+      "epoch": 0.190832,
+      "grad_norm": 0.875,
+      "learning_rate": 8.157096774193549e-05,
+      "loss": 0.1624,
+      "step": 11927
+    },
+    {
+      "epoch": 0.190848,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.156935483870968e-05,
+      "loss": 0.1711,
+      "step": 11928
+    },
+    {
+      "epoch": 0.190864,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.156774193548388e-05,
+      "loss": 0.1741,
+      "step": 11929
+    },
+    {
+      "epoch": 0.19088,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.156612903225806e-05,
+      "loss": 0.133,
+      "step": 11930
+    },
+    {
+      "epoch": 0.190896,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.156451612903226e-05,
+      "loss": 0.1753,
+      "step": 11931
+    },
+    {
+      "epoch": 0.190912,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.156290322580645e-05,
+      "loss": 0.1737,
+      "step": 11932
+    },
+    {
+      "epoch": 0.190928,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.156129032258065e-05,
+      "loss": 0.1813,
+      "step": 11933
+    },
+    {
+      "epoch": 0.190944,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.155967741935485e-05,
+      "loss": 0.1477,
+      "step": 11934
+    },
+    {
+      "epoch": 0.19096,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.155806451612905e-05,
+      "loss": 0.1872,
+      "step": 11935
+    },
+    {
+      "epoch": 0.190976,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.155645161290323e-05,
+      "loss": 0.1918,
+      "step": 11936
+    },
+    {
+      "epoch": 0.190992,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.155483870967743e-05,
+      "loss": 0.1449,
+      "step": 11937
+    },
+    {
+      "epoch": 0.191008,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.155322580645162e-05,
+      "loss": 0.1408,
+      "step": 11938
+    },
+    {
+      "epoch": 0.191024,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.155161290322582e-05,
+      "loss": 0.1845,
+      "step": 11939
+    },
+    {
+      "epoch": 0.19104,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.155e-05,
+      "loss": 0.1856,
+      "step": 11940
+    },
+    {
+      "epoch": 0.191056,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.154838709677419e-05,
+      "loss": 0.1769,
+      "step": 11941
+    },
+    {
+      "epoch": 0.191072,
+      "grad_norm": 1.125,
+      "learning_rate": 8.154677419354839e-05,
+      "loss": 0.1756,
+      "step": 11942
+    },
+    {
+      "epoch": 0.191088,
+      "grad_norm": 1.5703125,
+      "learning_rate": 8.154516129032258e-05,
+      "loss": 0.1936,
+      "step": 11943
+    },
+    {
+      "epoch": 0.191104,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.154354838709677e-05,
+      "loss": 0.1781,
+      "step": 11944
+    },
+    {
+      "epoch": 0.19112,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.154193548387097e-05,
+      "loss": 0.1908,
+      "step": 11945
+    },
+    {
+      "epoch": 0.191136,
+      "grad_norm": 1.2890625,
+      "learning_rate": 8.154032258064517e-05,
+      "loss": 0.2165,
+      "step": 11946
+    },
+    {
+      "epoch": 0.191152,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.153870967741936e-05,
+      "loss": 0.149,
+      "step": 11947
+    },
+    {
+      "epoch": 0.191168,
+      "grad_norm": 1.484375,
+      "learning_rate": 8.153709677419356e-05,
+      "loss": 0.1877,
+      "step": 11948
+    },
+    {
+      "epoch": 0.191184,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.153548387096775e-05,
+      "loss": 0.1539,
+      "step": 11949
+    },
+    {
+      "epoch": 0.1912,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.153387096774195e-05,
+      "loss": 0.1881,
+      "step": 11950
+    },
+    {
+      "epoch": 0.191216,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.153225806451613e-05,
+      "loss": 0.1992,
+      "step": 11951
+    },
+    {
+      "epoch": 0.191232,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.153064516129033e-05,
+      "loss": 0.1678,
+      "step": 11952
+    },
+    {
+      "epoch": 0.191248,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.152903225806452e-05,
+      "loss": 0.1645,
+      "step": 11953
+    },
+    {
+      "epoch": 0.191264,
+      "grad_norm": 0.75,
+      "learning_rate": 8.152741935483872e-05,
+      "loss": 0.1634,
+      "step": 11954
+    },
+    {
+      "epoch": 0.19128,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.15258064516129e-05,
+      "loss": 0.2071,
+      "step": 11955
+    },
+    {
+      "epoch": 0.191296,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.152419354838709e-05,
+      "loss": 0.1893,
+      "step": 11956
+    },
+    {
+      "epoch": 0.191312,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.152258064516129e-05,
+      "loss": 0.1728,
+      "step": 11957
+    },
+    {
+      "epoch": 0.191328,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.152096774193549e-05,
+      "loss": 0.2387,
+      "step": 11958
+    },
+    {
+      "epoch": 0.191344,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.151935483870969e-05,
+      "loss": 0.1949,
+      "step": 11959
+    },
+    {
+      "epoch": 0.19136,
+      "grad_norm": 0.494140625,
+      "learning_rate": 8.151774193548387e-05,
+      "loss": 0.1426,
+      "step": 11960
+    },
+    {
+      "epoch": 0.191376,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.151612903225807e-05,
+      "loss": 0.1599,
+      "step": 11961
+    },
+    {
+      "epoch": 0.191392,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.151451612903226e-05,
+      "loss": 0.1756,
+      "step": 11962
+    },
+    {
+      "epoch": 0.191408,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.151290322580646e-05,
+      "loss": 0.1509,
+      "step": 11963
+    },
+    {
+      "epoch": 0.191424,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.151129032258065e-05,
+      "loss": 0.1677,
+      "step": 11964
+    },
+    {
+      "epoch": 0.19144,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.150967741935485e-05,
+      "loss": 0.1858,
+      "step": 11965
+    },
+    {
+      "epoch": 0.191456,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.150806451612903e-05,
+      "loss": 0.1741,
+      "step": 11966
+    },
+    {
+      "epoch": 0.191472,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.150645161290323e-05,
+      "loss": 0.1919,
+      "step": 11967
+    },
+    {
+      "epoch": 0.191488,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.150483870967742e-05,
+      "loss": 0.1871,
+      "step": 11968
+    },
+    {
+      "epoch": 0.191504,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.150322580645162e-05,
+      "loss": 0.1406,
+      "step": 11969
+    },
+    {
+      "epoch": 0.19152,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.150161290322582e-05,
+      "loss": 0.1842,
+      "step": 11970
+    },
+    {
+      "epoch": 0.191536,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.15e-05,
+      "loss": 0.167,
+      "step": 11971
+    },
+    {
+      "epoch": 0.191552,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.14983870967742e-05,
+      "loss": 0.1892,
+      "step": 11972
+    },
+    {
+      "epoch": 0.191568,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.149677419354839e-05,
+      "loss": 0.206,
+      "step": 11973
+    },
+    {
+      "epoch": 0.191584,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.149516129032259e-05,
+      "loss": 0.1523,
+      "step": 11974
+    },
+    {
+      "epoch": 0.1916,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.149354838709677e-05,
+      "loss": 0.1579,
+      "step": 11975
+    },
+    {
+      "epoch": 0.191616,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.149193548387097e-05,
+      "loss": 0.177,
+      "step": 11976
+    },
+    {
+      "epoch": 0.191632,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.149032258064516e-05,
+      "loss": 0.1818,
+      "step": 11977
+    },
+    {
+      "epoch": 0.191648,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.148870967741936e-05,
+      "loss": 0.1722,
+      "step": 11978
+    },
+    {
+      "epoch": 0.191664,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.148709677419355e-05,
+      "loss": 0.186,
+      "step": 11979
+    },
+    {
+      "epoch": 0.19168,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.148548387096774e-05,
+      "loss": 0.1677,
+      "step": 11980
+    },
+    {
+      "epoch": 0.191696,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.148387096774193e-05,
+      "loss": 0.1743,
+      "step": 11981
+    },
+    {
+      "epoch": 0.191712,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.148225806451613e-05,
+      "loss": 0.1682,
+      "step": 11982
+    },
+    {
+      "epoch": 0.191728,
+      "grad_norm": 1.7890625,
+      "learning_rate": 8.148064516129033e-05,
+      "loss": 0.1973,
+      "step": 11983
+    },
+    {
+      "epoch": 0.191744,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.147903225806453e-05,
+      "loss": 0.1673,
+      "step": 11984
+    },
+    {
+      "epoch": 0.19176,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.147741935483872e-05,
+      "loss": 0.1794,
+      "step": 11985
+    },
+    {
+      "epoch": 0.191776,
+      "grad_norm": 1.7421875,
+      "learning_rate": 8.14758064516129e-05,
+      "loss": 0.1803,
+      "step": 11986
+    },
+    {
+      "epoch": 0.191792,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.14741935483871e-05,
+      "loss": 0.1747,
+      "step": 11987
+    },
+    {
+      "epoch": 0.191808,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.147258064516129e-05,
+      "loss": 0.1757,
+      "step": 11988
+    },
+    {
+      "epoch": 0.191824,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.147096774193549e-05,
+      "loss": 0.1569,
+      "step": 11989
+    },
+    {
+      "epoch": 0.19184,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.146935483870967e-05,
+      "loss": 0.1541,
+      "step": 11990
+    },
+    {
+      "epoch": 0.191856,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.146774193548387e-05,
+      "loss": 0.16,
+      "step": 11991
+    },
+    {
+      "epoch": 0.191872,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.146612903225806e-05,
+      "loss": 0.176,
+      "step": 11992
+    },
+    {
+      "epoch": 0.191888,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.146451612903226e-05,
+      "loss": 0.1452,
+      "step": 11993
+    },
+    {
+      "epoch": 0.191904,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.146290322580646e-05,
+      "loss": 0.1893,
+      "step": 11994
+    },
+    {
+      "epoch": 0.19192,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.146129032258066e-05,
+      "loss": 0.1444,
+      "step": 11995
+    },
+    {
+      "epoch": 0.191936,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.145967741935484e-05,
+      "loss": 0.1599,
+      "step": 11996
+    },
+    {
+      "epoch": 0.191952,
+      "grad_norm": 0.75,
+      "learning_rate": 8.145806451612904e-05,
+      "loss": 0.1713,
+      "step": 11997
+    },
+    {
+      "epoch": 0.191968,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.145645161290323e-05,
+      "loss": 0.1782,
+      "step": 11998
+    },
+    {
+      "epoch": 0.191984,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.145483870967743e-05,
+      "loss": 0.18,
+      "step": 11999
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.145322580645162e-05,
+      "loss": 0.2018,
+      "step": 12000
+    },
+    {
+      "epoch": 0.192016,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.145161290322582e-05,
+      "loss": 0.157,
+      "step": 12001
+    },
+    {
+      "epoch": 0.192032,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.145e-05,
+      "loss": 0.2087,
+      "step": 12002
+    },
+    {
+      "epoch": 0.192048,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.144838709677419e-05,
+      "loss": 0.228,
+      "step": 12003
+    },
+    {
+      "epoch": 0.192064,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.144677419354839e-05,
+      "loss": 0.1512,
+      "step": 12004
+    },
+    {
+      "epoch": 0.19208,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.144516129032259e-05,
+      "loss": 0.1682,
+      "step": 12005
+    },
+    {
+      "epoch": 0.192096,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.144354838709679e-05,
+      "loss": 0.1486,
+      "step": 12006
+    },
+    {
+      "epoch": 0.192112,
+      "grad_norm": 1.28125,
+      "learning_rate": 8.144193548387097e-05,
+      "loss": 0.2095,
+      "step": 12007
+    },
+    {
+      "epoch": 0.192128,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.144032258064517e-05,
+      "loss": 0.1398,
+      "step": 12008
+    },
+    {
+      "epoch": 0.192144,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.143870967741936e-05,
+      "loss": 0.1058,
+      "step": 12009
+    },
+    {
+      "epoch": 0.19216,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.143709677419356e-05,
+      "loss": 0.1678,
+      "step": 12010
+    },
+    {
+      "epoch": 0.192176,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.143548387096774e-05,
+      "loss": 0.1366,
+      "step": 12011
+    },
+    {
+      "epoch": 0.192192,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.143387096774194e-05,
+      "loss": 0.178,
+      "step": 12012
+    },
+    {
+      "epoch": 0.192208,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.143225806451613e-05,
+      "loss": 0.1335,
+      "step": 12013
+    },
+    {
+      "epoch": 0.192224,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.143064516129033e-05,
+      "loss": 0.1176,
+      "step": 12014
+    },
+    {
+      "epoch": 0.19224,
+      "grad_norm": 1.3203125,
+      "learning_rate": 8.142903225806452e-05,
+      "loss": 0.1799,
+      "step": 12015
+    },
+    {
+      "epoch": 0.192256,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.142741935483871e-05,
+      "loss": 0.1879,
+      "step": 12016
+    },
+    {
+      "epoch": 0.192272,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.14258064516129e-05,
+      "loss": 0.1544,
+      "step": 12017
+    },
+    {
+      "epoch": 0.192288,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.14241935483871e-05,
+      "loss": 0.1833,
+      "step": 12018
+    },
+    {
+      "epoch": 0.192304,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.14225806451613e-05,
+      "loss": 0.1928,
+      "step": 12019
+    },
+    {
+      "epoch": 0.19232,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.142096774193549e-05,
+      "loss": 0.1658,
+      "step": 12020
+    },
+    {
+      "epoch": 0.192336,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.141935483870969e-05,
+      "loss": 0.189,
+      "step": 12021
+    },
+    {
+      "epoch": 0.192352,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.141774193548387e-05,
+      "loss": 0.1168,
+      "step": 12022
+    },
+    {
+      "epoch": 0.192368,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.141612903225807e-05,
+      "loss": 0.2125,
+      "step": 12023
+    },
+    {
+      "epoch": 0.192384,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.141451612903226e-05,
+      "loss": 0.187,
+      "step": 12024
+    },
+    {
+      "epoch": 0.1924,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.141290322580646e-05,
+      "loss": 0.1665,
+      "step": 12025
+    },
+    {
+      "epoch": 0.192416,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.141129032258064e-05,
+      "loss": 0.1607,
+      "step": 12026
+    },
+    {
+      "epoch": 0.192432,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.140967741935484e-05,
+      "loss": 0.1551,
+      "step": 12027
+    },
+    {
+      "epoch": 0.192448,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.140806451612903e-05,
+      "loss": 0.1531,
+      "step": 12028
+    },
+    {
+      "epoch": 0.192464,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.140645161290323e-05,
+      "loss": 0.1802,
+      "step": 12029
+    },
+    {
+      "epoch": 0.19248,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.140483870967743e-05,
+      "loss": 0.1597,
+      "step": 12030
+    },
+    {
+      "epoch": 0.192496,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.140322580645163e-05,
+      "loss": 0.1184,
+      "step": 12031
+    },
+    {
+      "epoch": 0.192512,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.140161290322581e-05,
+      "loss": 0.1725,
+      "step": 12032
+    },
+    {
+      "epoch": 0.192528,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.14e-05,
+      "loss": 0.174,
+      "step": 12033
+    },
+    {
+      "epoch": 0.192544,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.13983870967742e-05,
+      "loss": 0.1714,
+      "step": 12034
+    },
+    {
+      "epoch": 0.19256,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.139677419354839e-05,
+      "loss": 0.1922,
+      "step": 12035
+    },
+    {
+      "epoch": 0.192576,
+      "grad_norm": 0.486328125,
+      "learning_rate": 8.139516129032259e-05,
+      "loss": 0.159,
+      "step": 12036
+    },
+    {
+      "epoch": 0.192592,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.139354838709677e-05,
+      "loss": 0.1534,
+      "step": 12037
+    },
+    {
+      "epoch": 0.192608,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.139193548387097e-05,
+      "loss": 0.1872,
+      "step": 12038
+    },
+    {
+      "epoch": 0.192624,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.139032258064516e-05,
+      "loss": 0.1959,
+      "step": 12039
+    },
+    {
+      "epoch": 0.19264,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.138870967741936e-05,
+      "loss": 0.1618,
+      "step": 12040
+    },
+    {
+      "epoch": 0.192656,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.138709677419356e-05,
+      "loss": 0.1648,
+      "step": 12041
+    },
+    {
+      "epoch": 0.192672,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.138548387096776e-05,
+      "loss": 0.2303,
+      "step": 12042
+    },
+    {
+      "epoch": 0.192688,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.138387096774194e-05,
+      "loss": 0.1595,
+      "step": 12043
+    },
+    {
+      "epoch": 0.192704,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.138225806451614e-05,
+      "loss": 0.1584,
+      "step": 12044
+    },
+    {
+      "epoch": 0.19272,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.138064516129033e-05,
+      "loss": 0.1857,
+      "step": 12045
+    },
+    {
+      "epoch": 0.192736,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.137903225806453e-05,
+      "loss": 0.202,
+      "step": 12046
+    },
+    {
+      "epoch": 0.192752,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.137741935483871e-05,
+      "loss": 0.1729,
+      "step": 12047
+    },
+    {
+      "epoch": 0.192768,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.137580645161291e-05,
+      "loss": 0.1547,
+      "step": 12048
+    },
+    {
+      "epoch": 0.192784,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.13741935483871e-05,
+      "loss": 0.2029,
+      "step": 12049
+    },
+    {
+      "epoch": 0.1928,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.137258064516129e-05,
+      "loss": 0.1664,
+      "step": 12050
+    },
+    {
+      "epoch": 0.192816,
+      "grad_norm": 0.49609375,
+      "learning_rate": 8.137096774193548e-05,
+      "loss": 0.1722,
+      "step": 12051
+    },
+    {
+      "epoch": 0.192832,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.136935483870967e-05,
+      "loss": 0.1888,
+      "step": 12052
+    },
+    {
+      "epoch": 0.192848,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.136774193548387e-05,
+      "loss": 0.1363,
+      "step": 12053
+    },
+    {
+      "epoch": 0.192864,
+      "grad_norm": 1.40625,
+      "learning_rate": 8.136612903225807e-05,
+      "loss": 0.1707,
+      "step": 12054
+    },
+    {
+      "epoch": 0.19288,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.136451612903227e-05,
+      "loss": 0.1562,
+      "step": 12055
+    },
+    {
+      "epoch": 0.192896,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.136290322580646e-05,
+      "loss": 0.1891,
+      "step": 12056
+    },
+    {
+      "epoch": 0.192912,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.136129032258066e-05,
+      "loss": 0.1921,
+      "step": 12057
+    },
+    {
+      "epoch": 0.192928,
+      "grad_norm": 1.3984375,
+      "learning_rate": 8.135967741935484e-05,
+      "loss": 0.1801,
+      "step": 12058
+    },
+    {
+      "epoch": 0.192944,
+      "grad_norm": 1.390625,
+      "learning_rate": 8.135806451612904e-05,
+      "loss": 0.1697,
+      "step": 12059
+    },
+    {
+      "epoch": 0.19296,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.135645161290323e-05,
+      "loss": 0.1813,
+      "step": 12060
+    },
+    {
+      "epoch": 0.192976,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.135483870967743e-05,
+      "loss": 0.1752,
+      "step": 12061
+    },
+    {
+      "epoch": 0.192992,
+      "grad_norm": 1.3203125,
+      "learning_rate": 8.135322580645161e-05,
+      "loss": 0.2173,
+      "step": 12062
+    },
+    {
+      "epoch": 0.193008,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.135161290322581e-05,
+      "loss": 0.153,
+      "step": 12063
+    },
+    {
+      "epoch": 0.193024,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.135e-05,
+      "loss": 0.1572,
+      "step": 12064
+    },
+    {
+      "epoch": 0.19304,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.13483870967742e-05,
+      "loss": 0.1846,
+      "step": 12065
+    },
+    {
+      "epoch": 0.193056,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.13467741935484e-05,
+      "loss": 0.1786,
+      "step": 12066
+    },
+    {
+      "epoch": 0.193072,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.134516129032258e-05,
+      "loss": 0.1661,
+      "step": 12067
+    },
+    {
+      "epoch": 0.193088,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.134354838709678e-05,
+      "loss": 0.171,
+      "step": 12068
+    },
+    {
+      "epoch": 0.193104,
+      "grad_norm": 0.875,
+      "learning_rate": 8.134193548387097e-05,
+      "loss": 0.1798,
+      "step": 12069
+    },
+    {
+      "epoch": 0.19312,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.134032258064517e-05,
+      "loss": 0.1812,
+      "step": 12070
+    },
+    {
+      "epoch": 0.193136,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.133870967741936e-05,
+      "loss": 0.1941,
+      "step": 12071
+    },
+    {
+      "epoch": 0.193152,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.133709677419356e-05,
+      "loss": 0.1695,
+      "step": 12072
+    },
+    {
+      "epoch": 0.193168,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.133548387096774e-05,
+      "loss": 0.1642,
+      "step": 12073
+    },
+    {
+      "epoch": 0.193184,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.133387096774194e-05,
+      "loss": 0.1804,
+      "step": 12074
+    },
+    {
+      "epoch": 0.1932,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.133225806451613e-05,
+      "loss": 0.1946,
+      "step": 12075
+    },
+    {
+      "epoch": 0.193216,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.133064516129033e-05,
+      "loss": 0.2143,
+      "step": 12076
+    },
+    {
+      "epoch": 0.193232,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.132903225806451e-05,
+      "loss": 0.1608,
+      "step": 12077
+    },
+    {
+      "epoch": 0.193248,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.132741935483871e-05,
+      "loss": 0.164,
+      "step": 12078
+    },
+    {
+      "epoch": 0.193264,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.132580645161291e-05,
+      "loss": 0.2016,
+      "step": 12079
+    },
+    {
+      "epoch": 0.19328,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.13241935483871e-05,
+      "loss": 0.192,
+      "step": 12080
+    },
+    {
+      "epoch": 0.193296,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.13225806451613e-05,
+      "loss": 0.111,
+      "step": 12081
+    },
+    {
+      "epoch": 0.193312,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.132096774193548e-05,
+      "loss": 0.1485,
+      "step": 12082
+    },
+    {
+      "epoch": 0.193328,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.131935483870968e-05,
+      "loss": 0.1667,
+      "step": 12083
+    },
+    {
+      "epoch": 0.193344,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.131774193548387e-05,
+      "loss": 0.1931,
+      "step": 12084
+    },
+    {
+      "epoch": 0.19336,
+      "grad_norm": 1.265625,
+      "learning_rate": 8.131612903225807e-05,
+      "loss": 0.1854,
+      "step": 12085
+    },
+    {
+      "epoch": 0.193376,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.131451612903226e-05,
+      "loss": 0.1901,
+      "step": 12086
+    },
+    {
+      "epoch": 0.193392,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.131290322580645e-05,
+      "loss": 0.1508,
+      "step": 12087
+    },
+    {
+      "epoch": 0.193408,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.131129032258064e-05,
+      "loss": 0.1578,
+      "step": 12088
+    },
+    {
+      "epoch": 0.193424,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.130967741935484e-05,
+      "loss": 0.1929,
+      "step": 12089
+    },
+    {
+      "epoch": 0.19344,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.130806451612904e-05,
+      "loss": 0.1711,
+      "step": 12090
+    },
+    {
+      "epoch": 0.193456,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.130645161290324e-05,
+      "loss": 0.1416,
+      "step": 12091
+    },
+    {
+      "epoch": 0.193472,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.130483870967743e-05,
+      "loss": 0.1592,
+      "step": 12092
+    },
+    {
+      "epoch": 0.193488,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.130322580645163e-05,
+      "loss": 0.1681,
+      "step": 12093
+    },
+    {
+      "epoch": 0.193504,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.130161290322581e-05,
+      "loss": 0.1595,
+      "step": 12094
+    },
+    {
+      "epoch": 0.19352,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.13e-05,
+      "loss": 0.186,
+      "step": 12095
+    },
+    {
+      "epoch": 0.193536,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.12983870967742e-05,
+      "loss": 0.1877,
+      "step": 12096
+    },
+    {
+      "epoch": 0.193552,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.129677419354838e-05,
+      "loss": 0.1667,
+      "step": 12097
+    },
+    {
+      "epoch": 0.193568,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.129516129032258e-05,
+      "loss": 0.1691,
+      "step": 12098
+    },
+    {
+      "epoch": 0.193584,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.129354838709677e-05,
+      "loss": 0.1647,
+      "step": 12099
+    },
+    {
+      "epoch": 0.1936,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.129193548387097e-05,
+      "loss": 0.1933,
+      "step": 12100
+    },
+    {
+      "epoch": 0.193616,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.129032258064517e-05,
+      "loss": 0.1766,
+      "step": 12101
+    },
+    {
+      "epoch": 0.193632,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.128870967741937e-05,
+      "loss": 0.2322,
+      "step": 12102
+    },
+    {
+      "epoch": 0.193648,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.128709677419355e-05,
+      "loss": 0.1867,
+      "step": 12103
+    },
+    {
+      "epoch": 0.193664,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.128548387096775e-05,
+      "loss": 0.138,
+      "step": 12104
+    },
+    {
+      "epoch": 0.19368,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.128387096774194e-05,
+      "loss": 0.1641,
+      "step": 12105
+    },
+    {
+      "epoch": 0.193696,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.128225806451614e-05,
+      "loss": 0.15,
+      "step": 12106
+    },
+    {
+      "epoch": 0.193712,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.128064516129033e-05,
+      "loss": 0.1457,
+      "step": 12107
+    },
+    {
+      "epoch": 0.193728,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.127903225806452e-05,
+      "loss": 0.1346,
+      "step": 12108
+    },
+    {
+      "epoch": 0.193744,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.127741935483871e-05,
+      "loss": 0.1804,
+      "step": 12109
+    },
+    {
+      "epoch": 0.19376,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.127580645161291e-05,
+      "loss": 0.1332,
+      "step": 12110
+    },
+    {
+      "epoch": 0.193776,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.12741935483871e-05,
+      "loss": 0.1553,
+      "step": 12111
+    },
+    {
+      "epoch": 0.193792,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.127258064516128e-05,
+      "loss": 0.1946,
+      "step": 12112
+    },
+    {
+      "epoch": 0.193808,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.127096774193548e-05,
+      "loss": 0.174,
+      "step": 12113
+    },
+    {
+      "epoch": 0.193824,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.126935483870968e-05,
+      "loss": 0.1923,
+      "step": 12114
+    },
+    {
+      "epoch": 0.19384,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.126774193548388e-05,
+      "loss": 0.1452,
+      "step": 12115
+    },
+    {
+      "epoch": 0.193856,
+      "grad_norm": 1.2734375,
+      "learning_rate": 8.126612903225807e-05,
+      "loss": 0.1872,
+      "step": 12116
+    },
+    {
+      "epoch": 0.193872,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.126451612903227e-05,
+      "loss": 0.1879,
+      "step": 12117
+    },
+    {
+      "epoch": 0.193888,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.126290322580645e-05,
+      "loss": 0.167,
+      "step": 12118
+    },
+    {
+      "epoch": 0.193904,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.126129032258065e-05,
+      "loss": 0.1867,
+      "step": 12119
+    },
+    {
+      "epoch": 0.19392,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.125967741935484e-05,
+      "loss": 0.1867,
+      "step": 12120
+    },
+    {
+      "epoch": 0.193936,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.125806451612904e-05,
+      "loss": 0.1586,
+      "step": 12121
+    },
+    {
+      "epoch": 0.193952,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.125645161290322e-05,
+      "loss": 0.1476,
+      "step": 12122
+    },
+    {
+      "epoch": 0.193968,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.125483870967742e-05,
+      "loss": 0.1974,
+      "step": 12123
+    },
+    {
+      "epoch": 0.193984,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.125322580645161e-05,
+      "loss": 0.1662,
+      "step": 12124
+    },
+    {
+      "epoch": 0.194,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.125161290322581e-05,
+      "loss": 0.1659,
+      "step": 12125
+    },
+    {
+      "epoch": 0.194016,
+      "grad_norm": 0.478515625,
+      "learning_rate": 8.125000000000001e-05,
+      "loss": 0.148,
+      "step": 12126
+    },
+    {
+      "epoch": 0.194032,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.12483870967742e-05,
+      "loss": 0.1741,
+      "step": 12127
+    },
+    {
+      "epoch": 0.194048,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.12467741935484e-05,
+      "loss": 0.1897,
+      "step": 12128
+    },
+    {
+      "epoch": 0.194064,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.124516129032258e-05,
+      "loss": 0.2078,
+      "step": 12129
+    },
+    {
+      "epoch": 0.19408,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.124354838709678e-05,
+      "loss": 0.1602,
+      "step": 12130
+    },
+    {
+      "epoch": 0.194096,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.124193548387097e-05,
+      "loss": 0.1594,
+      "step": 12131
+    },
+    {
+      "epoch": 0.194112,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.124032258064517e-05,
+      "loss": 0.1788,
+      "step": 12132
+    },
+    {
+      "epoch": 0.194128,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.123870967741935e-05,
+      "loss": 0.2022,
+      "step": 12133
+    },
+    {
+      "epoch": 0.194144,
+      "grad_norm": 1.234375,
+      "learning_rate": 8.123709677419355e-05,
+      "loss": 0.1839,
+      "step": 12134
+    },
+    {
+      "epoch": 0.19416,
+      "grad_norm": 1.0,
+      "learning_rate": 8.123548387096774e-05,
+      "loss": 0.1549,
+      "step": 12135
+    },
+    {
+      "epoch": 0.194176,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.123387096774194e-05,
+      "loss": 0.1506,
+      "step": 12136
+    },
+    {
+      "epoch": 0.194192,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.123225806451614e-05,
+      "loss": 0.1937,
+      "step": 12137
+    },
+    {
+      "epoch": 0.194208,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.123064516129032e-05,
+      "loss": 0.1785,
+      "step": 12138
+    },
+    {
+      "epoch": 0.194224,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.122903225806452e-05,
+      "loss": 0.1681,
+      "step": 12139
+    },
+    {
+      "epoch": 0.19424,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.122741935483872e-05,
+      "loss": 0.1451,
+      "step": 12140
+    },
+    {
+      "epoch": 0.194256,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.122580645161291e-05,
+      "loss": 0.1667,
+      "step": 12141
+    },
+    {
+      "epoch": 0.194272,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.12241935483871e-05,
+      "loss": 0.1608,
+      "step": 12142
+    },
+    {
+      "epoch": 0.194288,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.12225806451613e-05,
+      "loss": 0.1764,
+      "step": 12143
+    },
+    {
+      "epoch": 0.194304,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.122096774193548e-05,
+      "loss": 0.177,
+      "step": 12144
+    },
+    {
+      "epoch": 0.19432,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.121935483870968e-05,
+      "loss": 0.1126,
+      "step": 12145
+    },
+    {
+      "epoch": 0.194336,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.121774193548387e-05,
+      "loss": 0.2262,
+      "step": 12146
+    },
+    {
+      "epoch": 0.194352,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.121612903225807e-05,
+      "loss": 0.1712,
+      "step": 12147
+    },
+    {
+      "epoch": 0.194368,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.121451612903225e-05,
+      "loss": 0.2126,
+      "step": 12148
+    },
+    {
+      "epoch": 0.194384,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.121290322580645e-05,
+      "loss": 0.1356,
+      "step": 12149
+    },
+    {
+      "epoch": 0.1944,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.121129032258065e-05,
+      "loss": 0.2239,
+      "step": 12150
+    },
+    {
+      "epoch": 0.194416,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.120967741935485e-05,
+      "loss": 0.139,
+      "step": 12151
+    },
+    {
+      "epoch": 0.194432,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.120806451612904e-05,
+      "loss": 0.1564,
+      "step": 12152
+    },
+    {
+      "epoch": 0.194448,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.120645161290324e-05,
+      "loss": 0.169,
+      "step": 12153
+    },
+    {
+      "epoch": 0.194464,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.120483870967742e-05,
+      "loss": 0.1695,
+      "step": 12154
+    },
+    {
+      "epoch": 0.19448,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.120322580645162e-05,
+      "loss": 0.1858,
+      "step": 12155
+    },
+    {
+      "epoch": 0.194496,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.120161290322581e-05,
+      "loss": 0.1584,
+      "step": 12156
+    },
+    {
+      "epoch": 0.194512,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.120000000000001e-05,
+      "loss": 0.1821,
+      "step": 12157
+    },
+    {
+      "epoch": 0.194528,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.11983870967742e-05,
+      "loss": 0.1811,
+      "step": 12158
+    },
+    {
+      "epoch": 0.194544,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.119677419354838e-05,
+      "loss": 0.207,
+      "step": 12159
+    },
+    {
+      "epoch": 0.19456,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.119516129032258e-05,
+      "loss": 0.1486,
+      "step": 12160
+    },
+    {
+      "epoch": 0.194576,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.119354838709678e-05,
+      "loss": 0.1467,
+      "step": 12161
+    },
+    {
+      "epoch": 0.194592,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.119193548387098e-05,
+      "loss": 0.1534,
+      "step": 12162
+    },
+    {
+      "epoch": 0.194608,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.119032258064517e-05,
+      "loss": 0.1906,
+      "step": 12163
+    },
+    {
+      "epoch": 0.194624,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.118870967741937e-05,
+      "loss": 0.2123,
+      "step": 12164
+    },
+    {
+      "epoch": 0.19464,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.118709677419355e-05,
+      "loss": 0.1516,
+      "step": 12165
+    },
+    {
+      "epoch": 0.194656,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.118548387096775e-05,
+      "loss": 0.1565,
+      "step": 12166
+    },
+    {
+      "epoch": 0.194672,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.118387096774194e-05,
+      "loss": 0.1937,
+      "step": 12167
+    },
+    {
+      "epoch": 0.194688,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.118225806451614e-05,
+      "loss": 0.1642,
+      "step": 12168
+    },
+    {
+      "epoch": 0.194704,
+      "grad_norm": 1.28125,
+      "learning_rate": 8.118064516129032e-05,
+      "loss": 0.1577,
+      "step": 12169
+    },
+    {
+      "epoch": 0.19472,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.117903225806452e-05,
+      "loss": 0.1961,
+      "step": 12170
+    },
+    {
+      "epoch": 0.194736,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.117741935483871e-05,
+      "loss": 0.2025,
+      "step": 12171
+    },
+    {
+      "epoch": 0.194752,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.117580645161291e-05,
+      "loss": 0.164,
+      "step": 12172
+    },
+    {
+      "epoch": 0.194768,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.11741935483871e-05,
+      "loss": 0.1416,
+      "step": 12173
+    },
+    {
+      "epoch": 0.194784,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.11725806451613e-05,
+      "loss": 0.1581,
+      "step": 12174
+    },
+    {
+      "epoch": 0.1948,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.11709677419355e-05,
+      "loss": 0.1599,
+      "step": 12175
+    },
+    {
+      "epoch": 0.194816,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.116935483870968e-05,
+      "loss": 0.1409,
+      "step": 12176
+    },
+    {
+      "epoch": 0.194832,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.116774193548388e-05,
+      "loss": 0.1616,
+      "step": 12177
+    },
+    {
+      "epoch": 0.194848,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.116612903225807e-05,
+      "loss": 0.1784,
+      "step": 12178
+    },
+    {
+      "epoch": 0.194864,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.116451612903226e-05,
+      "loss": 0.1564,
+      "step": 12179
+    },
+    {
+      "epoch": 0.19488,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.116290322580645e-05,
+      "loss": 0.1625,
+      "step": 12180
+    },
+    {
+      "epoch": 0.194896,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.116129032258065e-05,
+      "loss": 0.1424,
+      "step": 12181
+    },
+    {
+      "epoch": 0.194912,
+      "grad_norm": 1.28125,
+      "learning_rate": 8.115967741935484e-05,
+      "loss": 0.1892,
+      "step": 12182
+    },
+    {
+      "epoch": 0.194928,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.115806451612904e-05,
+      "loss": 0.1679,
+      "step": 12183
+    },
+    {
+      "epoch": 0.194944,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.115645161290322e-05,
+      "loss": 0.187,
+      "step": 12184
+    },
+    {
+      "epoch": 0.19496,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.115483870967742e-05,
+      "loss": 0.1676,
+      "step": 12185
+    },
+    {
+      "epoch": 0.194976,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.115322580645162e-05,
+      "loss": 0.17,
+      "step": 12186
+    },
+    {
+      "epoch": 0.194992,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.115161290322582e-05,
+      "loss": 0.1781,
+      "step": 12187
+    },
+    {
+      "epoch": 0.195008,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.115000000000001e-05,
+      "loss": 0.1647,
+      "step": 12188
+    },
+    {
+      "epoch": 0.195024,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.11483870967742e-05,
+      "loss": 0.1568,
+      "step": 12189
+    },
+    {
+      "epoch": 0.19504,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.114677419354839e-05,
+      "loss": 0.142,
+      "step": 12190
+    },
+    {
+      "epoch": 0.195056,
+      "grad_norm": 1.3359375,
+      "learning_rate": 8.114516129032258e-05,
+      "loss": 0.1246,
+      "step": 12191
+    },
+    {
+      "epoch": 0.195072,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.114354838709678e-05,
+      "loss": 0.1874,
+      "step": 12192
+    },
+    {
+      "epoch": 0.195088,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.114193548387096e-05,
+      "loss": 0.1348,
+      "step": 12193
+    },
+    {
+      "epoch": 0.195104,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.114032258064516e-05,
+      "loss": 0.1631,
+      "step": 12194
+    },
+    {
+      "epoch": 0.19512,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.113870967741935e-05,
+      "loss": 0.2125,
+      "step": 12195
+    },
+    {
+      "epoch": 0.195136,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.113709677419355e-05,
+      "loss": 0.1679,
+      "step": 12196
+    },
+    {
+      "epoch": 0.195152,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.113548387096775e-05,
+      "loss": 0.1551,
+      "step": 12197
+    },
+    {
+      "epoch": 0.195168,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.113387096774195e-05,
+      "loss": 0.1657,
+      "step": 12198
+    },
+    {
+      "epoch": 0.195184,
+      "grad_norm": 0.5,
+      "learning_rate": 8.113225806451614e-05,
+      "loss": 0.1662,
+      "step": 12199
+    },
+    {
+      "epoch": 0.1952,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.113064516129034e-05,
+      "loss": 0.1484,
+      "step": 12200
+    },
+    {
+      "epoch": 0.195216,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.112903225806452e-05,
+      "loss": 0.1459,
+      "step": 12201
+    },
+    {
+      "epoch": 0.195232,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.112741935483872e-05,
+      "loss": 0.216,
+      "step": 12202
+    },
+    {
+      "epoch": 0.195248,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.112580645161291e-05,
+      "loss": 0.2083,
+      "step": 12203
+    },
+    {
+      "epoch": 0.195264,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.112419354838709e-05,
+      "loss": 0.15,
+      "step": 12204
+    },
+    {
+      "epoch": 0.19528,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.112258064516129e-05,
+      "loss": 0.1738,
+      "step": 12205
+    },
+    {
+      "epoch": 0.195296,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.112096774193548e-05,
+      "loss": 0.2009,
+      "step": 12206
+    },
+    {
+      "epoch": 0.195312,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.111935483870968e-05,
+      "loss": 0.1495,
+      "step": 12207
+    },
+    {
+      "epoch": 0.195328,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.111774193548386e-05,
+      "loss": 0.1983,
+      "step": 12208
+    },
+    {
+      "epoch": 0.195344,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.111612903225806e-05,
+      "loss": 0.1825,
+      "step": 12209
+    },
+    {
+      "epoch": 0.19536,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.111451612903226e-05,
+      "loss": 0.1727,
+      "step": 12210
+    },
+    {
+      "epoch": 0.195376,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.111290322580646e-05,
+      "loss": 0.1722,
+      "step": 12211
+    },
+    {
+      "epoch": 0.195392,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.111129032258065e-05,
+      "loss": 0.1576,
+      "step": 12212
+    },
+    {
+      "epoch": 0.195408,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.110967741935485e-05,
+      "loss": 0.1755,
+      "step": 12213
+    },
+    {
+      "epoch": 0.195424,
+      "grad_norm": 0.875,
+      "learning_rate": 8.110806451612904e-05,
+      "loss": 0.1567,
+      "step": 12214
+    },
+    {
+      "epoch": 0.19544,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.110645161290323e-05,
+      "loss": 0.1743,
+      "step": 12215
+    },
+    {
+      "epoch": 0.195456,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.110483870967742e-05,
+      "loss": 0.1785,
+      "step": 12216
+    },
+    {
+      "epoch": 0.195472,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.110322580645162e-05,
+      "loss": 0.1545,
+      "step": 12217
+    },
+    {
+      "epoch": 0.195488,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.11016129032258e-05,
+      "loss": 0.1809,
+      "step": 12218
+    },
+    {
+      "epoch": 0.195504,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.11e-05,
+      "loss": 0.1792,
+      "step": 12219
+    },
+    {
+      "epoch": 0.19552,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.109838709677419e-05,
+      "loss": 0.1687,
+      "step": 12220
+    },
+    {
+      "epoch": 0.195536,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.109677419354839e-05,
+      "loss": 0.1857,
+      "step": 12221
+    },
+    {
+      "epoch": 0.195552,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.109516129032259e-05,
+      "loss": 0.1989,
+      "step": 12222
+    },
+    {
+      "epoch": 0.195568,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.109354838709678e-05,
+      "loss": 0.1726,
+      "step": 12223
+    },
+    {
+      "epoch": 0.195584,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.109193548387098e-05,
+      "loss": 0.154,
+      "step": 12224
+    },
+    {
+      "epoch": 0.1956,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.109032258064516e-05,
+      "loss": 0.1533,
+      "step": 12225
+    },
+    {
+      "epoch": 0.195616,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.108870967741936e-05,
+      "loss": 0.2213,
+      "step": 12226
+    },
+    {
+      "epoch": 0.195632,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.108709677419355e-05,
+      "loss": 0.1181,
+      "step": 12227
+    },
+    {
+      "epoch": 0.195648,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.108548387096775e-05,
+      "loss": 0.1945,
+      "step": 12228
+    },
+    {
+      "epoch": 0.195664,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.108387096774193e-05,
+      "loss": 0.1679,
+      "step": 12229
+    },
+    {
+      "epoch": 0.19568,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.108225806451613e-05,
+      "loss": 0.1737,
+      "step": 12230
+    },
+    {
+      "epoch": 0.195696,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.108064516129032e-05,
+      "loss": 0.2112,
+      "step": 12231
+    },
+    {
+      "epoch": 0.195712,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.107903225806452e-05,
+      "loss": 0.1768,
+      "step": 12232
+    },
+    {
+      "epoch": 0.195728,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.10774193548387e-05,
+      "loss": 0.1771,
+      "step": 12233
+    },
+    {
+      "epoch": 0.195744,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.10758064516129e-05,
+      "loss": 0.1596,
+      "step": 12234
+    },
+    {
+      "epoch": 0.19576,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.10741935483871e-05,
+      "loss": 0.1561,
+      "step": 12235
+    },
+    {
+      "epoch": 0.195776,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.107258064516129e-05,
+      "loss": 0.1539,
+      "step": 12236
+    },
+    {
+      "epoch": 0.195792,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.107096774193549e-05,
+      "loss": 0.1507,
+      "step": 12237
+    },
+    {
+      "epoch": 0.195808,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.106935483870968e-05,
+      "loss": 0.1699,
+      "step": 12238
+    },
+    {
+      "epoch": 0.195824,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.106774193548388e-05,
+      "loss": 0.1706,
+      "step": 12239
+    },
+    {
+      "epoch": 0.19584,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.106612903225806e-05,
+      "loss": 0.1689,
+      "step": 12240
+    },
+    {
+      "epoch": 0.195856,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.106451612903226e-05,
+      "loss": 0.1462,
+      "step": 12241
+    },
+    {
+      "epoch": 0.195872,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.106290322580645e-05,
+      "loss": 0.1845,
+      "step": 12242
+    },
+    {
+      "epoch": 0.195888,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.106129032258065e-05,
+      "loss": 0.1715,
+      "step": 12243
+    },
+    {
+      "epoch": 0.195904,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.105967741935483e-05,
+      "loss": 0.1401,
+      "step": 12244
+    },
+    {
+      "epoch": 0.19592,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.105806451612903e-05,
+      "loss": 0.1873,
+      "step": 12245
+    },
+    {
+      "epoch": 0.195936,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.105645161290323e-05,
+      "loss": 0.1485,
+      "step": 12246
+    },
+    {
+      "epoch": 0.195952,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.105483870967743e-05,
+      "loss": 0.2135,
+      "step": 12247
+    },
+    {
+      "epoch": 0.195968,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.105322580645162e-05,
+      "loss": 0.1548,
+      "step": 12248
+    },
+    {
+      "epoch": 0.195984,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.105161290322582e-05,
+      "loss": 0.2036,
+      "step": 12249
+    },
+    {
+      "epoch": 0.196,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.105e-05,
+      "loss": 0.1566,
+      "step": 12250
+    },
+    {
+      "epoch": 0.196016,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.104838709677419e-05,
+      "loss": 0.1512,
+      "step": 12251
+    },
+    {
+      "epoch": 0.196032,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.104677419354839e-05,
+      "loss": 0.1925,
+      "step": 12252
+    },
+    {
+      "epoch": 0.196048,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.104516129032258e-05,
+      "loss": 0.1524,
+      "step": 12253
+    },
+    {
+      "epoch": 0.196064,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.104354838709678e-05,
+      "loss": 0.1853,
+      "step": 12254
+    },
+    {
+      "epoch": 0.19608,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.104193548387096e-05,
+      "loss": 0.175,
+      "step": 12255
+    },
+    {
+      "epoch": 0.196096,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.104032258064516e-05,
+      "loss": 0.166,
+      "step": 12256
+    },
+    {
+      "epoch": 0.196112,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.103870967741936e-05,
+      "loss": 0.12,
+      "step": 12257
+    },
+    {
+      "epoch": 0.196128,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.103709677419356e-05,
+      "loss": 0.2063,
+      "step": 12258
+    },
+    {
+      "epoch": 0.196144,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.103548387096775e-05,
+      "loss": 0.1731,
+      "step": 12259
+    },
+    {
+      "epoch": 0.19616,
+      "grad_norm": 1.0,
+      "learning_rate": 8.103387096774195e-05,
+      "loss": 0.1774,
+      "step": 12260
+    },
+    {
+      "epoch": 0.196176,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.103225806451613e-05,
+      "loss": 0.1717,
+      "step": 12261
+    },
+    {
+      "epoch": 0.196192,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.103064516129033e-05,
+      "loss": 0.1558,
+      "step": 12262
+    },
+    {
+      "epoch": 0.196208,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.102903225806452e-05,
+      "loss": 0.2342,
+      "step": 12263
+    },
+    {
+      "epoch": 0.196224,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.102741935483872e-05,
+      "loss": 0.1625,
+      "step": 12264
+    },
+    {
+      "epoch": 0.19624,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.10258064516129e-05,
+      "loss": 0.1362,
+      "step": 12265
+    },
+    {
+      "epoch": 0.196256,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.10241935483871e-05,
+      "loss": 0.1672,
+      "step": 12266
+    },
+    {
+      "epoch": 0.196272,
+      "grad_norm": 1.421875,
+      "learning_rate": 8.102258064516129e-05,
+      "loss": 0.1936,
+      "step": 12267
+    },
+    {
+      "epoch": 0.196288,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.102096774193548e-05,
+      "loss": 0.223,
+      "step": 12268
+    },
+    {
+      "epoch": 0.196304,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.101935483870968e-05,
+      "loss": 0.1559,
+      "step": 12269
+    },
+    {
+      "epoch": 0.19632,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.101774193548388e-05,
+      "loss": 0.136,
+      "step": 12270
+    },
+    {
+      "epoch": 0.196336,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.101612903225808e-05,
+      "loss": 0.2124,
+      "step": 12271
+    },
+    {
+      "epoch": 0.196352,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.101451612903226e-05,
+      "loss": 0.173,
+      "step": 12272
+    },
+    {
+      "epoch": 0.196368,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.101290322580646e-05,
+      "loss": 0.1647,
+      "step": 12273
+    },
+    {
+      "epoch": 0.196384,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.101129032258065e-05,
+      "loss": 0.1915,
+      "step": 12274
+    },
+    {
+      "epoch": 0.1964,
+      "grad_norm": 1.2578125,
+      "learning_rate": 8.100967741935485e-05,
+      "loss": 0.173,
+      "step": 12275
+    },
+    {
+      "epoch": 0.196416,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.100806451612903e-05,
+      "loss": 0.1724,
+      "step": 12276
+    },
+    {
+      "epoch": 0.196432,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.100645161290323e-05,
+      "loss": 0.146,
+      "step": 12277
+    },
+    {
+      "epoch": 0.196448,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.100483870967742e-05,
+      "loss": 0.1528,
+      "step": 12278
+    },
+    {
+      "epoch": 0.196464,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.100322580645162e-05,
+      "loss": 0.1611,
+      "step": 12279
+    },
+    {
+      "epoch": 0.19648,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.10016129032258e-05,
+      "loss": 0.1702,
+      "step": 12280
+    },
+    {
+      "epoch": 0.196496,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.1e-05,
+      "loss": 0.1682,
+      "step": 12281
+    },
+    {
+      "epoch": 0.196512,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.09983870967742e-05,
+      "loss": 0.1471,
+      "step": 12282
+    },
+    {
+      "epoch": 0.196528,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.099677419354839e-05,
+      "loss": 0.1648,
+      "step": 12283
+    },
+    {
+      "epoch": 0.196544,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.099516129032259e-05,
+      "loss": 0.1633,
+      "step": 12284
+    },
+    {
+      "epoch": 0.19656,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.099354838709678e-05,
+      "loss": 0.1761,
+      "step": 12285
+    },
+    {
+      "epoch": 0.196576,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.099193548387097e-05,
+      "loss": 0.169,
+      "step": 12286
+    },
+    {
+      "epoch": 0.196592,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.099032258064516e-05,
+      "loss": 0.1368,
+      "step": 12287
+    },
+    {
+      "epoch": 0.196608,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.098870967741936e-05,
+      "loss": 0.129,
+      "step": 12288
+    },
+    {
+      "epoch": 0.196624,
+      "grad_norm": 1.0,
+      "learning_rate": 8.098709677419355e-05,
+      "loss": 0.1496,
+      "step": 12289
+    },
+    {
+      "epoch": 0.19664,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.098548387096775e-05,
+      "loss": 0.1909,
+      "step": 12290
+    },
+    {
+      "epoch": 0.196656,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.098387096774193e-05,
+      "loss": 0.1816,
+      "step": 12291
+    },
+    {
+      "epoch": 0.196672,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.098225806451613e-05,
+      "loss": 0.1408,
+      "step": 12292
+    },
+    {
+      "epoch": 0.196688,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.098064516129033e-05,
+      "loss": 0.1941,
+      "step": 12293
+    },
+    {
+      "epoch": 0.196704,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.097903225806453e-05,
+      "loss": 0.1764,
+      "step": 12294
+    },
+    {
+      "epoch": 0.19672,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.097741935483872e-05,
+      "loss": 0.2216,
+      "step": 12295
+    },
+    {
+      "epoch": 0.196736,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.097580645161292e-05,
+      "loss": 0.1797,
+      "step": 12296
+    },
+    {
+      "epoch": 0.196752,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.09741935483871e-05,
+      "loss": 0.1891,
+      "step": 12297
+    },
+    {
+      "epoch": 0.196768,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.097258064516129e-05,
+      "loss": 0.1641,
+      "step": 12298
+    },
+    {
+      "epoch": 0.196784,
+      "grad_norm": 0.75,
+      "learning_rate": 8.097096774193549e-05,
+      "loss": 0.1693,
+      "step": 12299
+    },
+    {
+      "epoch": 0.1968,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.096935483870967e-05,
+      "loss": 0.1522,
+      "step": 12300
+    },
+    {
+      "epoch": 0.196816,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.096774193548387e-05,
+      "loss": 0.1568,
+      "step": 12301
+    },
+    {
+      "epoch": 0.196832,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.096612903225806e-05,
+      "loss": 0.1791,
+      "step": 12302
+    },
+    {
+      "epoch": 0.196848,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.096451612903226e-05,
+      "loss": 0.165,
+      "step": 12303
+    },
+    {
+      "epoch": 0.196864,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.096290322580645e-05,
+      "loss": 0.1916,
+      "step": 12304
+    },
+    {
+      "epoch": 0.19688,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.096129032258065e-05,
+      "loss": 0.1238,
+      "step": 12305
+    },
+    {
+      "epoch": 0.196896,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.095967741935485e-05,
+      "loss": 0.1758,
+      "step": 12306
+    },
+    {
+      "epoch": 0.196912,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.095806451612905e-05,
+      "loss": 0.1174,
+      "step": 12307
+    },
+    {
+      "epoch": 0.196928,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.095645161290323e-05,
+      "loss": 0.1613,
+      "step": 12308
+    },
+    {
+      "epoch": 0.196944,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.095483870967743e-05,
+      "loss": 0.1662,
+      "step": 12309
+    },
+    {
+      "epoch": 0.19696,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.095322580645162e-05,
+      "loss": 0.2133,
+      "step": 12310
+    },
+    {
+      "epoch": 0.196976,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.095161290322582e-05,
+      "loss": 0.1541,
+      "step": 12311
+    },
+    {
+      "epoch": 0.196992,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.095e-05,
+      "loss": 0.187,
+      "step": 12312
+    },
+    {
+      "epoch": 0.197008,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.094838709677419e-05,
+      "loss": 0.2014,
+      "step": 12313
+    },
+    {
+      "epoch": 0.197024,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.094677419354839e-05,
+      "loss": 0.1628,
+      "step": 12314
+    },
+    {
+      "epoch": 0.19704,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.094516129032257e-05,
+      "loss": 0.15,
+      "step": 12315
+    },
+    {
+      "epoch": 0.197056,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.094354838709677e-05,
+      "loss": 0.1846,
+      "step": 12316
+    },
+    {
+      "epoch": 0.197072,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.094193548387097e-05,
+      "loss": 0.1995,
+      "step": 12317
+    },
+    {
+      "epoch": 0.197088,
+      "grad_norm": 1.484375,
+      "learning_rate": 8.094032258064517e-05,
+      "loss": 0.1915,
+      "step": 12318
+    },
+    {
+      "epoch": 0.197104,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.093870967741936e-05,
+      "loss": 0.1813,
+      "step": 12319
+    },
+    {
+      "epoch": 0.19712,
+      "grad_norm": 0.75,
+      "learning_rate": 8.093709677419356e-05,
+      "loss": 0.1849,
+      "step": 12320
+    },
+    {
+      "epoch": 0.197136,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.093548387096774e-05,
+      "loss": 0.2195,
+      "step": 12321
+    },
+    {
+      "epoch": 0.197152,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.093387096774194e-05,
+      "loss": 0.1771,
+      "step": 12322
+    },
+    {
+      "epoch": 0.197168,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.093225806451613e-05,
+      "loss": 0.1866,
+      "step": 12323
+    },
+    {
+      "epoch": 0.197184,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.093064516129033e-05,
+      "loss": 0.1454,
+      "step": 12324
+    },
+    {
+      "epoch": 0.1972,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.092903225806452e-05,
+      "loss": 0.1327,
+      "step": 12325
+    },
+    {
+      "epoch": 0.197216,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.092741935483872e-05,
+      "loss": 0.1905,
+      "step": 12326
+    },
+    {
+      "epoch": 0.197232,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.09258064516129e-05,
+      "loss": 0.2279,
+      "step": 12327
+    },
+    {
+      "epoch": 0.197248,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.09241935483871e-05,
+      "loss": 0.1585,
+      "step": 12328
+    },
+    {
+      "epoch": 0.197264,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.092258064516129e-05,
+      "loss": 0.1601,
+      "step": 12329
+    },
+    {
+      "epoch": 0.19728,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.092096774193549e-05,
+      "loss": 0.171,
+      "step": 12330
+    },
+    {
+      "epoch": 0.197296,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.091935483870969e-05,
+      "loss": 0.1619,
+      "step": 12331
+    },
+    {
+      "epoch": 0.197312,
+      "grad_norm": 0.625,
+      "learning_rate": 8.091774193548387e-05,
+      "loss": 0.1421,
+      "step": 12332
+    },
+    {
+      "epoch": 0.197328,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.091612903225807e-05,
+      "loss": 0.1795,
+      "step": 12333
+    },
+    {
+      "epoch": 0.197344,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.091451612903226e-05,
+      "loss": 0.1801,
+      "step": 12334
+    },
+    {
+      "epoch": 0.19736,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.091290322580646e-05,
+      "loss": 0.1949,
+      "step": 12335
+    },
+    {
+      "epoch": 0.197376,
+      "grad_norm": 0.95703125,
+      "learning_rate": 8.091129032258064e-05,
+      "loss": 0.2004,
+      "step": 12336
+    },
+    {
+      "epoch": 0.197392,
+      "grad_norm": 1.3203125,
+      "learning_rate": 8.090967741935484e-05,
+      "loss": 0.2072,
+      "step": 12337
+    },
+    {
+      "epoch": 0.197408,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.090806451612903e-05,
+      "loss": 0.1937,
+      "step": 12338
+    },
+    {
+      "epoch": 0.197424,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.090645161290323e-05,
+      "loss": 0.2079,
+      "step": 12339
+    },
+    {
+      "epoch": 0.19744,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.090483870967742e-05,
+      "loss": 0.1571,
+      "step": 12340
+    },
+    {
+      "epoch": 0.197456,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.090322580645162e-05,
+      "loss": 0.1603,
+      "step": 12341
+    },
+    {
+      "epoch": 0.197472,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.090161290322582e-05,
+      "loss": 0.1849,
+      "step": 12342
+    },
+    {
+      "epoch": 0.197488,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.090000000000001e-05,
+      "loss": 0.1649,
+      "step": 12343
+    },
+    {
+      "epoch": 0.197504,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.08983870967742e-05,
+      "loss": 0.1674,
+      "step": 12344
+    },
+    {
+      "epoch": 0.19752,
+      "grad_norm": 1.234375,
+      "learning_rate": 8.089677419354839e-05,
+      "loss": 0.1918,
+      "step": 12345
+    },
+    {
+      "epoch": 0.197536,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.089516129032259e-05,
+      "loss": 0.1966,
+      "step": 12346
+    },
+    {
+      "epoch": 0.197552,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.089354838709677e-05,
+      "loss": 0.1713,
+      "step": 12347
+    },
+    {
+      "epoch": 0.197568,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.089193548387097e-05,
+      "loss": 0.1699,
+      "step": 12348
+    },
+    {
+      "epoch": 0.197584,
+      "grad_norm": 0.474609375,
+      "learning_rate": 8.089032258064516e-05,
+      "loss": 0.1211,
+      "step": 12349
+    },
+    {
+      "epoch": 0.1976,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.088870967741936e-05,
+      "loss": 0.2076,
+      "step": 12350
+    },
+    {
+      "epoch": 0.197616,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.088709677419354e-05,
+      "loss": 0.1332,
+      "step": 12351
+    },
+    {
+      "epoch": 0.197632,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.088548387096774e-05,
+      "loss": 0.1346,
+      "step": 12352
+    },
+    {
+      "epoch": 0.197648,
+      "grad_norm": 1.125,
+      "learning_rate": 8.088387096774194e-05,
+      "loss": 0.2035,
+      "step": 12353
+    },
+    {
+      "epoch": 0.197664,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.088225806451614e-05,
+      "loss": 0.1503,
+      "step": 12354
+    },
+    {
+      "epoch": 0.19768,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.088064516129033e-05,
+      "loss": 0.1803,
+      "step": 12355
+    },
+    {
+      "epoch": 0.197696,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.087903225806453e-05,
+      "loss": 0.1876,
+      "step": 12356
+    },
+    {
+      "epoch": 0.197712,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.087741935483871e-05,
+      "loss": 0.1696,
+      "step": 12357
+    },
+    {
+      "epoch": 0.197728,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.087580645161291e-05,
+      "loss": 0.1784,
+      "step": 12358
+    },
+    {
+      "epoch": 0.197744,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.08741935483871e-05,
+      "loss": 0.1688,
+      "step": 12359
+    },
+    {
+      "epoch": 0.19776,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.087258064516129e-05,
+      "loss": 0.1506,
+      "step": 12360
+    },
+    {
+      "epoch": 0.197776,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.087096774193549e-05,
+      "loss": 0.1558,
+      "step": 12361
+    },
+    {
+      "epoch": 0.197792,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.086935483870967e-05,
+      "loss": 0.185,
+      "step": 12362
+    },
+    {
+      "epoch": 0.197808,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.086774193548387e-05,
+      "loss": 0.1382,
+      "step": 12363
+    },
+    {
+      "epoch": 0.197824,
+      "grad_norm": 0.95703125,
+      "learning_rate": 8.086612903225806e-05,
+      "loss": 0.2008,
+      "step": 12364
+    },
+    {
+      "epoch": 0.19784,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.086451612903226e-05,
+      "loss": 0.2191,
+      "step": 12365
+    },
+    {
+      "epoch": 0.197856,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.086290322580646e-05,
+      "loss": 0.1775,
+      "step": 12366
+    },
+    {
+      "epoch": 0.197872,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.086129032258066e-05,
+      "loss": 0.1953,
+      "step": 12367
+    },
+    {
+      "epoch": 0.197888,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.085967741935484e-05,
+      "loss": 0.1836,
+      "step": 12368
+    },
+    {
+      "epoch": 0.197904,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.085806451612904e-05,
+      "loss": 0.151,
+      "step": 12369
+    },
+    {
+      "epoch": 0.19792,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.085645161290323e-05,
+      "loss": 0.1506,
+      "step": 12370
+    },
+    {
+      "epoch": 0.197936,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.085483870967743e-05,
+      "loss": 0.1828,
+      "step": 12371
+    },
+    {
+      "epoch": 0.197952,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.085322580645161e-05,
+      "loss": 0.1523,
+      "step": 12372
+    },
+    {
+      "epoch": 0.197968,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.085161290322581e-05,
+      "loss": 0.1667,
+      "step": 12373
+    },
+    {
+      "epoch": 0.197984,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.085e-05,
+      "loss": 0.1852,
+      "step": 12374
+    },
+    {
+      "epoch": 0.198,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.084838709677419e-05,
+      "loss": 0.2022,
+      "step": 12375
+    },
+    {
+      "epoch": 0.198016,
+      "grad_norm": 0.75,
+      "learning_rate": 8.084677419354839e-05,
+      "loss": 0.1514,
+      "step": 12376
+    },
+    {
+      "epoch": 0.198032,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.084516129032259e-05,
+      "loss": 0.1671,
+      "step": 12377
+    },
+    {
+      "epoch": 0.198048,
+      "grad_norm": 1.4453125,
+      "learning_rate": 8.084354838709679e-05,
+      "loss": 0.1934,
+      "step": 12378
+    },
+    {
+      "epoch": 0.198064,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.084193548387097e-05,
+      "loss": 0.1271,
+      "step": 12379
+    },
+    {
+      "epoch": 0.19808,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.084032258064517e-05,
+      "loss": 0.1527,
+      "step": 12380
+    },
+    {
+      "epoch": 0.198096,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.083870967741936e-05,
+      "loss": 0.2002,
+      "step": 12381
+    },
+    {
+      "epoch": 0.198112,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.083709677419356e-05,
+      "loss": 0.1702,
+      "step": 12382
+    },
+    {
+      "epoch": 0.198128,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.083548387096774e-05,
+      "loss": 0.1585,
+      "step": 12383
+    },
+    {
+      "epoch": 0.198144,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.083387096774194e-05,
+      "loss": 0.178,
+      "step": 12384
+    },
+    {
+      "epoch": 0.19816,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.083225806451613e-05,
+      "loss": 0.215,
+      "step": 12385
+    },
+    {
+      "epoch": 0.198176,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.083064516129033e-05,
+      "loss": 0.1504,
+      "step": 12386
+    },
+    {
+      "epoch": 0.198192,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.082903225806451e-05,
+      "loss": 0.2135,
+      "step": 12387
+    },
+    {
+      "epoch": 0.198208,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.082741935483871e-05,
+      "loss": 0.1529,
+      "step": 12388
+    },
+    {
+      "epoch": 0.198224,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.082580645161291e-05,
+      "loss": 0.1814,
+      "step": 12389
+    },
+    {
+      "epoch": 0.19824,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.08241935483871e-05,
+      "loss": 0.2156,
+      "step": 12390
+    },
+    {
+      "epoch": 0.198256,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.08225806451613e-05,
+      "loss": 0.1943,
+      "step": 12391
+    },
+    {
+      "epoch": 0.198272,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.082096774193549e-05,
+      "loss": 0.163,
+      "step": 12392
+    },
+    {
+      "epoch": 0.198288,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.081935483870968e-05,
+      "loss": 0.1553,
+      "step": 12393
+    },
+    {
+      "epoch": 0.198304,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.081774193548387e-05,
+      "loss": 0.1401,
+      "step": 12394
+    },
+    {
+      "epoch": 0.19832,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.081612903225807e-05,
+      "loss": 0.1682,
+      "step": 12395
+    },
+    {
+      "epoch": 0.198336,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.081451612903226e-05,
+      "loss": 0.2463,
+      "step": 12396
+    },
+    {
+      "epoch": 0.198352,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.081290322580646e-05,
+      "loss": 0.1642,
+      "step": 12397
+    },
+    {
+      "epoch": 0.198368,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.081129032258064e-05,
+      "loss": 0.2097,
+      "step": 12398
+    },
+    {
+      "epoch": 0.198384,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.080967741935484e-05,
+      "loss": 0.1985,
+      "step": 12399
+    },
+    {
+      "epoch": 0.1984,
+      "grad_norm": 1.453125,
+      "learning_rate": 8.080806451612903e-05,
+      "loss": 0.16,
+      "step": 12400
+    },
+    {
+      "epoch": 0.198416,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.080645161290323e-05,
+      "loss": 0.1666,
+      "step": 12401
+    },
+    {
+      "epoch": 0.198432,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.080483870967743e-05,
+      "loss": 0.1558,
+      "step": 12402
+    },
+    {
+      "epoch": 0.198448,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.080322580645163e-05,
+      "loss": 0.1691,
+      "step": 12403
+    },
+    {
+      "epoch": 0.198464,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.080161290322581e-05,
+      "loss": 0.2335,
+      "step": 12404
+    },
+    {
+      "epoch": 0.19848,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.080000000000001e-05,
+      "loss": 0.1657,
+      "step": 12405
+    },
+    {
+      "epoch": 0.198496,
+      "grad_norm": 0.75,
+      "learning_rate": 8.07983870967742e-05,
+      "loss": 0.153,
+      "step": 12406
+    },
+    {
+      "epoch": 0.198512,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.079677419354838e-05,
+      "loss": 0.1551,
+      "step": 12407
+    },
+    {
+      "epoch": 0.198528,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.079516129032258e-05,
+      "loss": 0.1983,
+      "step": 12408
+    },
+    {
+      "epoch": 0.198544,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.079354838709677e-05,
+      "loss": 0.1726,
+      "step": 12409
+    },
+    {
+      "epoch": 0.19856,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.079193548387097e-05,
+      "loss": 0.1792,
+      "step": 12410
+    },
+    {
+      "epoch": 0.198576,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.079032258064516e-05,
+      "loss": 0.162,
+      "step": 12411
+    },
+    {
+      "epoch": 0.198592,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.078870967741936e-05,
+      "loss": 0.1665,
+      "step": 12412
+    },
+    {
+      "epoch": 0.198608,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.078709677419356e-05,
+      "loss": 0.174,
+      "step": 12413
+    },
+    {
+      "epoch": 0.198624,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.078548387096775e-05,
+      "loss": 0.1937,
+      "step": 12414
+    },
+    {
+      "epoch": 0.19864,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.078387096774194e-05,
+      "loss": 0.118,
+      "step": 12415
+    },
+    {
+      "epoch": 0.198656,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.078225806451614e-05,
+      "loss": 0.1543,
+      "step": 12416
+    },
+    {
+      "epoch": 0.198672,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.078064516129033e-05,
+      "loss": 0.1772,
+      "step": 12417
+    },
+    {
+      "epoch": 0.198688,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.077903225806453e-05,
+      "loss": 0.1878,
+      "step": 12418
+    },
+    {
+      "epoch": 0.198704,
+      "grad_norm": 1.3984375,
+      "learning_rate": 8.077741935483871e-05,
+      "loss": 0.1968,
+      "step": 12419
+    },
+    {
+      "epoch": 0.19872,
+      "grad_norm": 1.78125,
+      "learning_rate": 8.077580645161291e-05,
+      "loss": 0.2392,
+      "step": 12420
+    },
+    {
+      "epoch": 0.198736,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.07741935483871e-05,
+      "loss": 0.1731,
+      "step": 12421
+    },
+    {
+      "epoch": 0.198752,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.077258064516128e-05,
+      "loss": 0.1482,
+      "step": 12422
+    },
+    {
+      "epoch": 0.198768,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.077096774193548e-05,
+      "loss": 0.1842,
+      "step": 12423
+    },
+    {
+      "epoch": 0.198784,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.076935483870967e-05,
+      "loss": 0.1393,
+      "step": 12424
+    },
+    {
+      "epoch": 0.1988,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.076774193548387e-05,
+      "loss": 0.1427,
+      "step": 12425
+    },
+    {
+      "epoch": 0.198816,
+      "grad_norm": 1.5,
+      "learning_rate": 8.076612903225807e-05,
+      "loss": 0.2047,
+      "step": 12426
+    },
+    {
+      "epoch": 0.198832,
+      "grad_norm": 0.95703125,
+      "learning_rate": 8.076451612903227e-05,
+      "loss": 0.1759,
+      "step": 12427
+    },
+    {
+      "epoch": 0.198848,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.076290322580645e-05,
+      "loss": 0.176,
+      "step": 12428
+    },
+    {
+      "epoch": 0.198864,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.076129032258065e-05,
+      "loss": 0.1798,
+      "step": 12429
+    },
+    {
+      "epoch": 0.19888,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.075967741935484e-05,
+      "loss": 0.2376,
+      "step": 12430
+    },
+    {
+      "epoch": 0.198896,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.075806451612904e-05,
+      "loss": 0.1672,
+      "step": 12431
+    },
+    {
+      "epoch": 0.198912,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.075645161290323e-05,
+      "loss": 0.2054,
+      "step": 12432
+    },
+    {
+      "epoch": 0.198928,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.075483870967743e-05,
+      "loss": 0.1528,
+      "step": 12433
+    },
+    {
+      "epoch": 0.198944,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.075322580645161e-05,
+      "loss": 0.1727,
+      "step": 12434
+    },
+    {
+      "epoch": 0.19896,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.075161290322581e-05,
+      "loss": 0.1748,
+      "step": 12435
+    },
+    {
+      "epoch": 0.198976,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.075e-05,
+      "loss": 0.1929,
+      "step": 12436
+    },
+    {
+      "epoch": 0.198992,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.07483870967742e-05,
+      "loss": 0.1611,
+      "step": 12437
+    },
+    {
+      "epoch": 0.199008,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.07467741935484e-05,
+      "loss": 0.1481,
+      "step": 12438
+    },
+    {
+      "epoch": 0.199024,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.074516129032258e-05,
+      "loss": 0.1661,
+      "step": 12439
+    },
+    {
+      "epoch": 0.19904,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.074354838709678e-05,
+      "loss": 0.1793,
+      "step": 12440
+    },
+    {
+      "epoch": 0.199056,
+      "grad_norm": 1.3046875,
+      "learning_rate": 8.074193548387097e-05,
+      "loss": 0.2046,
+      "step": 12441
+    },
+    {
+      "epoch": 0.199072,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.074032258064517e-05,
+      "loss": 0.1611,
+      "step": 12442
+    },
+    {
+      "epoch": 0.199088,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.073870967741935e-05,
+      "loss": 0.2131,
+      "step": 12443
+    },
+    {
+      "epoch": 0.199104,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.073709677419355e-05,
+      "loss": 0.1778,
+      "step": 12444
+    },
+    {
+      "epoch": 0.19912,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.073548387096774e-05,
+      "loss": 0.164,
+      "step": 12445
+    },
+    {
+      "epoch": 0.199136,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.073387096774194e-05,
+      "loss": 0.1845,
+      "step": 12446
+    },
+    {
+      "epoch": 0.199152,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.073225806451613e-05,
+      "loss": 0.191,
+      "step": 12447
+    },
+    {
+      "epoch": 0.199168,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.073064516129033e-05,
+      "loss": 0.1649,
+      "step": 12448
+    },
+    {
+      "epoch": 0.199184,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.072903225806453e-05,
+      "loss": 0.1864,
+      "step": 12449
+    },
+    {
+      "epoch": 0.1992,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.072741935483872e-05,
+      "loss": 0.1497,
+      "step": 12450
+    },
+    {
+      "epoch": 0.199216,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.072580645161291e-05,
+      "loss": 0.1691,
+      "step": 12451
+    },
+    {
+      "epoch": 0.199232,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.072419354838711e-05,
+      "loss": 0.1921,
+      "step": 12452
+    },
+    {
+      "epoch": 0.199248,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.07225806451613e-05,
+      "loss": 0.2133,
+      "step": 12453
+    },
+    {
+      "epoch": 0.199264,
+      "grad_norm": 1.8046875,
+      "learning_rate": 8.072096774193548e-05,
+      "loss": 0.201,
+      "step": 12454
+    },
+    {
+      "epoch": 0.19928,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.071935483870968e-05,
+      "loss": 0.1639,
+      "step": 12455
+    },
+    {
+      "epoch": 0.199296,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.071774193548387e-05,
+      "loss": 0.1556,
+      "step": 12456
+    },
+    {
+      "epoch": 0.199312,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.071612903225807e-05,
+      "loss": 0.2058,
+      "step": 12457
+    },
+    {
+      "epoch": 0.199328,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.071451612903225e-05,
+      "loss": 0.1519,
+      "step": 12458
+    },
+    {
+      "epoch": 0.199344,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.071290322580645e-05,
+      "loss": 0.1634,
+      "step": 12459
+    },
+    {
+      "epoch": 0.19936,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.071129032258064e-05,
+      "loss": 0.1815,
+      "step": 12460
+    },
+    {
+      "epoch": 0.199376,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.070967741935484e-05,
+      "loss": 0.1712,
+      "step": 12461
+    },
+    {
+      "epoch": 0.199392,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.070806451612904e-05,
+      "loss": 0.1745,
+      "step": 12462
+    },
+    {
+      "epoch": 0.199408,
+      "grad_norm": 1.375,
+      "learning_rate": 8.070645161290324e-05,
+      "loss": 0.2,
+      "step": 12463
+    },
+    {
+      "epoch": 0.199424,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.070483870967742e-05,
+      "loss": 0.1581,
+      "step": 12464
+    },
+    {
+      "epoch": 0.19944,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.070322580645162e-05,
+      "loss": 0.1978,
+      "step": 12465
+    },
+    {
+      "epoch": 0.199456,
+      "grad_norm": 1.328125,
+      "learning_rate": 8.070161290322581e-05,
+      "loss": 0.1818,
+      "step": 12466
+    },
+    {
+      "epoch": 0.199472,
+      "grad_norm": 1.484375,
+      "learning_rate": 8.070000000000001e-05,
+      "loss": 0.2174,
+      "step": 12467
+    },
+    {
+      "epoch": 0.199488,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.06983870967742e-05,
+      "loss": 0.1432,
+      "step": 12468
+    },
+    {
+      "epoch": 0.199504,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.069677419354838e-05,
+      "loss": 0.1961,
+      "step": 12469
+    },
+    {
+      "epoch": 0.19952,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.069516129032258e-05,
+      "loss": 0.1874,
+      "step": 12470
+    },
+    {
+      "epoch": 0.199536,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.069354838709677e-05,
+      "loss": 0.1748,
+      "step": 12471
+    },
+    {
+      "epoch": 0.199552,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.069193548387097e-05,
+      "loss": 0.1597,
+      "step": 12472
+    },
+    {
+      "epoch": 0.199568,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.069032258064517e-05,
+      "loss": 0.1632,
+      "step": 12473
+    },
+    {
+      "epoch": 0.199584,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.068870967741937e-05,
+      "loss": 0.1974,
+      "step": 12474
+    },
+    {
+      "epoch": 0.1996,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.068709677419355e-05,
+      "loss": 0.1933,
+      "step": 12475
+    },
+    {
+      "epoch": 0.199616,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.068548387096775e-05,
+      "loss": 0.1658,
+      "step": 12476
+    },
+    {
+      "epoch": 0.199632,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.068387096774194e-05,
+      "loss": 0.1827,
+      "step": 12477
+    },
+    {
+      "epoch": 0.199648,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.068225806451614e-05,
+      "loss": 0.1685,
+      "step": 12478
+    },
+    {
+      "epoch": 0.199664,
+      "grad_norm": 1.2109375,
+      "learning_rate": 8.068064516129032e-05,
+      "loss": 0.1899,
+      "step": 12479
+    },
+    {
+      "epoch": 0.19968,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.067903225806452e-05,
+      "loss": 0.1497,
+      "step": 12480
+    },
+    {
+      "epoch": 0.199696,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.067741935483871e-05,
+      "loss": 0.1454,
+      "step": 12481
+    },
+    {
+      "epoch": 0.199712,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.067580645161291e-05,
+      "loss": 0.1892,
+      "step": 12482
+    },
+    {
+      "epoch": 0.199728,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.06741935483871e-05,
+      "loss": 0.1426,
+      "step": 12483
+    },
+    {
+      "epoch": 0.199744,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.06725806451613e-05,
+      "loss": 0.145,
+      "step": 12484
+    },
+    {
+      "epoch": 0.19976,
+      "grad_norm": 0.95703125,
+      "learning_rate": 8.067096774193548e-05,
+      "loss": 0.1616,
+      "step": 12485
+    },
+    {
+      "epoch": 0.199776,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.066935483870968e-05,
+      "loss": 0.2073,
+      "step": 12486
+    },
+    {
+      "epoch": 0.199792,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.066774193548388e-05,
+      "loss": 0.1622,
+      "step": 12487
+    },
+    {
+      "epoch": 0.199808,
+      "grad_norm": 1.25,
+      "learning_rate": 8.066612903225807e-05,
+      "loss": 0.2179,
+      "step": 12488
+    },
+    {
+      "epoch": 0.199824,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.066451612903227e-05,
+      "loss": 0.1761,
+      "step": 12489
+    },
+    {
+      "epoch": 0.19984,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.066290322580645e-05,
+      "loss": 0.1627,
+      "step": 12490
+    },
+    {
+      "epoch": 0.199856,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.066129032258065e-05,
+      "loss": 0.1985,
+      "step": 12491
+    },
+    {
+      "epoch": 0.199872,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.065967741935484e-05,
+      "loss": 0.1758,
+      "step": 12492
+    },
+    {
+      "epoch": 0.199888,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.065806451612904e-05,
+      "loss": 0.1423,
+      "step": 12493
+    },
+    {
+      "epoch": 0.199904,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.065645161290322e-05,
+      "loss": 0.1598,
+      "step": 12494
+    },
+    {
+      "epoch": 0.19992,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.065483870967742e-05,
+      "loss": 0.1931,
+      "step": 12495
+    },
+    {
+      "epoch": 0.199936,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.065322580645161e-05,
+      "loss": 0.1859,
+      "step": 12496
+    },
+    {
+      "epoch": 0.199952,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.065161290322581e-05,
+      "loss": 0.1598,
+      "step": 12497
+    },
+    {
+      "epoch": 0.199968,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.065000000000001e-05,
+      "loss": 0.1531,
+      "step": 12498
+    },
+    {
+      "epoch": 0.199984,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.064838709677421e-05,
+      "loss": 0.1232,
+      "step": 12499
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.06467741935484e-05,
+      "loss": 0.1401,
+      "step": 12500
+    },
+    {
+      "epoch": 0.200016,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.064516129032258e-05,
+      "loss": 0.1151,
+      "step": 12501
+    },
+    {
+      "epoch": 0.200032,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.064354838709678e-05,
+      "loss": 0.1475,
+      "step": 12502
+    },
+    {
+      "epoch": 0.200048,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.064193548387097e-05,
+      "loss": 0.1784,
+      "step": 12503
+    },
+    {
+      "epoch": 0.200064,
+      "grad_norm": 0.625,
+      "learning_rate": 8.064032258064517e-05,
+      "loss": 0.1464,
+      "step": 12504
+    },
+    {
+      "epoch": 0.20008,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.063870967741935e-05,
+      "loss": 0.1558,
+      "step": 12505
+    },
+    {
+      "epoch": 0.200096,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.063709677419355e-05,
+      "loss": 0.1411,
+      "step": 12506
+    },
+    {
+      "epoch": 0.200112,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.063548387096774e-05,
+      "loss": 0.1357,
+      "step": 12507
+    },
+    {
+      "epoch": 0.200128,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.063387096774194e-05,
+      "loss": 0.1656,
+      "step": 12508
+    },
+    {
+      "epoch": 0.200144,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.063225806451614e-05,
+      "loss": 0.163,
+      "step": 12509
+    },
+    {
+      "epoch": 0.20016,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.063064516129034e-05,
+      "loss": 0.159,
+      "step": 12510
+    },
+    {
+      "epoch": 0.200176,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.062903225806452e-05,
+      "loss": 0.2026,
+      "step": 12511
+    },
+    {
+      "epoch": 0.200192,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.062741935483872e-05,
+      "loss": 0.2079,
+      "step": 12512
+    },
+    {
+      "epoch": 0.200208,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.062580645161291e-05,
+      "loss": 0.1699,
+      "step": 12513
+    },
+    {
+      "epoch": 0.200224,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.062419354838711e-05,
+      "loss": 0.178,
+      "step": 12514
+    },
+    {
+      "epoch": 0.20024,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.06225806451613e-05,
+      "loss": 0.1965,
+      "step": 12515
+    },
+    {
+      "epoch": 0.200256,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.062096774193548e-05,
+      "loss": 0.1549,
+      "step": 12516
+    },
+    {
+      "epoch": 0.200272,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.061935483870968e-05,
+      "loss": 0.159,
+      "step": 12517
+    },
+    {
+      "epoch": 0.200288,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.061774193548387e-05,
+      "loss": 0.1235,
+      "step": 12518
+    },
+    {
+      "epoch": 0.200304,
+      "grad_norm": 0.875,
+      "learning_rate": 8.061612903225807e-05,
+      "loss": 0.1779,
+      "step": 12519
+    },
+    {
+      "epoch": 0.20032,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.061451612903225e-05,
+      "loss": 0.185,
+      "step": 12520
+    },
+    {
+      "epoch": 0.200336,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.061290322580645e-05,
+      "loss": 0.1697,
+      "step": 12521
+    },
+    {
+      "epoch": 0.200352,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.061129032258065e-05,
+      "loss": 0.143,
+      "step": 12522
+    },
+    {
+      "epoch": 0.200368,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.060967741935485e-05,
+      "loss": 0.1828,
+      "step": 12523
+    },
+    {
+      "epoch": 0.200384,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.060806451612904e-05,
+      "loss": 0.1846,
+      "step": 12524
+    },
+    {
+      "epoch": 0.2004,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.060645161290324e-05,
+      "loss": 0.1564,
+      "step": 12525
+    },
+    {
+      "epoch": 0.200416,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.060483870967742e-05,
+      "loss": 0.1601,
+      "step": 12526
+    },
+    {
+      "epoch": 0.200432,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.060322580645162e-05,
+      "loss": 0.2183,
+      "step": 12527
+    },
+    {
+      "epoch": 0.200448,
+      "grad_norm": 0.875,
+      "learning_rate": 8.060161290322581e-05,
+      "loss": 0.1987,
+      "step": 12528
+    },
+    {
+      "epoch": 0.200464,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.060000000000001e-05,
+      "loss": 0.1578,
+      "step": 12529
+    },
+    {
+      "epoch": 0.20048,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.05983870967742e-05,
+      "loss": 0.1732,
+      "step": 12530
+    },
+    {
+      "epoch": 0.200496,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.059677419354838e-05,
+      "loss": 0.1487,
+      "step": 12531
+    },
+    {
+      "epoch": 0.200512,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.059516129032258e-05,
+      "loss": 0.1657,
+      "step": 12532
+    },
+    {
+      "epoch": 0.200528,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.059354838709678e-05,
+      "loss": 0.16,
+      "step": 12533
+    },
+    {
+      "epoch": 0.200544,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.059193548387098e-05,
+      "loss": 0.1835,
+      "step": 12534
+    },
+    {
+      "epoch": 0.20056,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.059032258064516e-05,
+      "loss": 0.1322,
+      "step": 12535
+    },
+    {
+      "epoch": 0.200576,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.058870967741936e-05,
+      "loss": 0.1546,
+      "step": 12536
+    },
+    {
+      "epoch": 0.200592,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.058709677419355e-05,
+      "loss": 0.1657,
+      "step": 12537
+    },
+    {
+      "epoch": 0.200608,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.058548387096775e-05,
+      "loss": 0.1967,
+      "step": 12538
+    },
+    {
+      "epoch": 0.200624,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.058387096774194e-05,
+      "loss": 0.1454,
+      "step": 12539
+    },
+    {
+      "epoch": 0.20064,
+      "grad_norm": 1.0,
+      "learning_rate": 8.058225806451614e-05,
+      "loss": 0.2017,
+      "step": 12540
+    },
+    {
+      "epoch": 0.200656,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.058064516129032e-05,
+      "loss": 0.1495,
+      "step": 12541
+    },
+    {
+      "epoch": 0.200672,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.057903225806452e-05,
+      "loss": 0.1659,
+      "step": 12542
+    },
+    {
+      "epoch": 0.200688,
+      "grad_norm": 1.3203125,
+      "learning_rate": 8.057741935483871e-05,
+      "loss": 0.2032,
+      "step": 12543
+    },
+    {
+      "epoch": 0.200704,
+      "grad_norm": 0.490234375,
+      "learning_rate": 8.057580645161291e-05,
+      "loss": 0.1578,
+      "step": 12544
+    },
+    {
+      "epoch": 0.20072,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.057419354838711e-05,
+      "loss": 0.2187,
+      "step": 12545
+    },
+    {
+      "epoch": 0.200736,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.05725806451613e-05,
+      "loss": 0.1646,
+      "step": 12546
+    },
+    {
+      "epoch": 0.200752,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.057096774193549e-05,
+      "loss": 0.1799,
+      "step": 12547
+    },
+    {
+      "epoch": 0.200768,
+      "grad_norm": 1.296875,
+      "learning_rate": 8.056935483870968e-05,
+      "loss": 0.1675,
+      "step": 12548
+    },
+    {
+      "epoch": 0.200784,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.056774193548388e-05,
+      "loss": 0.1756,
+      "step": 12549
+    },
+    {
+      "epoch": 0.2008,
+      "grad_norm": 1.0,
+      "learning_rate": 8.056612903225806e-05,
+      "loss": 0.1555,
+      "step": 12550
+    },
+    {
+      "epoch": 0.200816,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.056451612903226e-05,
+      "loss": 0.1364,
+      "step": 12551
+    },
+    {
+      "epoch": 0.200832,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.056290322580645e-05,
+      "loss": 0.1849,
+      "step": 12552
+    },
+    {
+      "epoch": 0.200848,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.056129032258065e-05,
+      "loss": 0.1956,
+      "step": 12553
+    },
+    {
+      "epoch": 0.200864,
+      "grad_norm": 1.4609375,
+      "learning_rate": 8.055967741935484e-05,
+      "loss": 0.2034,
+      "step": 12554
+    },
+    {
+      "epoch": 0.20088,
+      "grad_norm": 1.515625,
+      "learning_rate": 8.055806451612904e-05,
+      "loss": 0.1775,
+      "step": 12555
+    },
+    {
+      "epoch": 0.200896,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.055645161290322e-05,
+      "loss": 0.1602,
+      "step": 12556
+    },
+    {
+      "epoch": 0.200912,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.055483870967742e-05,
+      "loss": 0.177,
+      "step": 12557
+    },
+    {
+      "epoch": 0.200928,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.055322580645162e-05,
+      "loss": 0.1491,
+      "step": 12558
+    },
+    {
+      "epoch": 0.200944,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.055161290322582e-05,
+      "loss": 0.2127,
+      "step": 12559
+    },
+    {
+      "epoch": 0.20096,
+      "grad_norm": 1.125,
+      "learning_rate": 8.055e-05,
+      "loss": 0.176,
+      "step": 12560
+    },
+    {
+      "epoch": 0.200976,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.05483870967742e-05,
+      "loss": 0.1706,
+      "step": 12561
+    },
+    {
+      "epoch": 0.200992,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.054677419354839e-05,
+      "loss": 0.1622,
+      "step": 12562
+    },
+    {
+      "epoch": 0.201008,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.054516129032258e-05,
+      "loss": 0.1528,
+      "step": 12563
+    },
+    {
+      "epoch": 0.201024,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.054354838709678e-05,
+      "loss": 0.1746,
+      "step": 12564
+    },
+    {
+      "epoch": 0.20104,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.054193548387096e-05,
+      "loss": 0.179,
+      "step": 12565
+    },
+    {
+      "epoch": 0.201056,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.054032258064516e-05,
+      "loss": 0.2006,
+      "step": 12566
+    },
+    {
+      "epoch": 0.201072,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.053870967741935e-05,
+      "loss": 0.2142,
+      "step": 12567
+    },
+    {
+      "epoch": 0.201088,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.053709677419355e-05,
+      "loss": 0.1736,
+      "step": 12568
+    },
+    {
+      "epoch": 0.201104,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.053548387096775e-05,
+      "loss": 0.1884,
+      "step": 12569
+    },
+    {
+      "epoch": 0.20112,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.053387096774195e-05,
+      "loss": 0.1342,
+      "step": 12570
+    },
+    {
+      "epoch": 0.201136,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.053225806451613e-05,
+      "loss": 0.1432,
+      "step": 12571
+    },
+    {
+      "epoch": 0.201152,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.053064516129033e-05,
+      "loss": 0.1878,
+      "step": 12572
+    },
+    {
+      "epoch": 0.201168,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.052903225806452e-05,
+      "loss": 0.1658,
+      "step": 12573
+    },
+    {
+      "epoch": 0.201184,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.052741935483872e-05,
+      "loss": 0.1546,
+      "step": 12574
+    },
+    {
+      "epoch": 0.2012,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.05258064516129e-05,
+      "loss": 0.1525,
+      "step": 12575
+    },
+    {
+      "epoch": 0.201216,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.05241935483871e-05,
+      "loss": 0.1493,
+      "step": 12576
+    },
+    {
+      "epoch": 0.201232,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.052258064516129e-05,
+      "loss": 0.1749,
+      "step": 12577
+    },
+    {
+      "epoch": 0.201248,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.052096774193548e-05,
+      "loss": 0.1613,
+      "step": 12578
+    },
+    {
+      "epoch": 0.201264,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.051935483870968e-05,
+      "loss": 0.1294,
+      "step": 12579
+    },
+    {
+      "epoch": 0.20128,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.051774193548388e-05,
+      "loss": 0.1859,
+      "step": 12580
+    },
+    {
+      "epoch": 0.201296,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.051612903225806e-05,
+      "loss": 0.207,
+      "step": 12581
+    },
+    {
+      "epoch": 0.201312,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.051451612903226e-05,
+      "loss": 0.1575,
+      "step": 12582
+    },
+    {
+      "epoch": 0.201328,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.051290322580646e-05,
+      "loss": 0.1427,
+      "step": 12583
+    },
+    {
+      "epoch": 0.201344,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.051129032258065e-05,
+      "loss": 0.117,
+      "step": 12584
+    },
+    {
+      "epoch": 0.20136,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.050967741935485e-05,
+      "loss": 0.1644,
+      "step": 12585
+    },
+    {
+      "epoch": 0.201376,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.050806451612903e-05,
+      "loss": 0.1588,
+      "step": 12586
+    },
+    {
+      "epoch": 0.201392,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.050645161290323e-05,
+      "loss": 0.1696,
+      "step": 12587
+    },
+    {
+      "epoch": 0.201408,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.050483870967742e-05,
+      "loss": 0.2027,
+      "step": 12588
+    },
+    {
+      "epoch": 0.201424,
+      "grad_norm": 0.95703125,
+      "learning_rate": 8.050322580645162e-05,
+      "loss": 0.158,
+      "step": 12589
+    },
+    {
+      "epoch": 0.20144,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.05016129032258e-05,
+      "loss": 0.1419,
+      "step": 12590
+    },
+    {
+      "epoch": 0.201456,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.05e-05,
+      "loss": 0.1692,
+      "step": 12591
+    },
+    {
+      "epoch": 0.201472,
+      "grad_norm": 0.478515625,
+      "learning_rate": 8.049838709677419e-05,
+      "loss": 0.135,
+      "step": 12592
+    },
+    {
+      "epoch": 0.201488,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.049677419354839e-05,
+      "loss": 0.1347,
+      "step": 12593
+    },
+    {
+      "epoch": 0.201504,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.049516129032259e-05,
+      "loss": 0.1646,
+      "step": 12594
+    },
+    {
+      "epoch": 0.20152,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.049354838709678e-05,
+      "loss": 0.1656,
+      "step": 12595
+    },
+    {
+      "epoch": 0.201536,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.049193548387098e-05,
+      "loss": 0.1722,
+      "step": 12596
+    },
+    {
+      "epoch": 0.201552,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.049032258064516e-05,
+      "loss": 0.1901,
+      "step": 12597
+    },
+    {
+      "epoch": 0.201568,
+      "grad_norm": 1.2734375,
+      "learning_rate": 8.048870967741936e-05,
+      "loss": 0.1839,
+      "step": 12598
+    },
+    {
+      "epoch": 0.201584,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.048709677419355e-05,
+      "loss": 0.1336,
+      "step": 12599
+    },
+    {
+      "epoch": 0.2016,
+      "grad_norm": 0.48828125,
+      "learning_rate": 8.048548387096775e-05,
+      "loss": 0.1451,
+      "step": 12600
+    },
+    {
+      "epoch": 0.201616,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.048387096774193e-05,
+      "loss": 0.1964,
+      "step": 12601
+    },
+    {
+      "epoch": 0.201632,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.048225806451613e-05,
+      "loss": 0.1946,
+      "step": 12602
+    },
+    {
+      "epoch": 0.201648,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.048064516129032e-05,
+      "loss": 0.1614,
+      "step": 12603
+    },
+    {
+      "epoch": 0.201664,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.047903225806452e-05,
+      "loss": 0.202,
+      "step": 12604
+    },
+    {
+      "epoch": 0.20168,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.047741935483872e-05,
+      "loss": 0.19,
+      "step": 12605
+    },
+    {
+      "epoch": 0.201696,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.047580645161292e-05,
+      "loss": 0.2026,
+      "step": 12606
+    },
+    {
+      "epoch": 0.201712,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.04741935483871e-05,
+      "loss": 0.2054,
+      "step": 12607
+    },
+    {
+      "epoch": 0.201728,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.04725806451613e-05,
+      "loss": 0.195,
+      "step": 12608
+    },
+    {
+      "epoch": 0.201744,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.047096774193549e-05,
+      "loss": 0.1447,
+      "step": 12609
+    },
+    {
+      "epoch": 0.20176,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.046935483870968e-05,
+      "loss": 0.169,
+      "step": 12610
+    },
+    {
+      "epoch": 0.201776,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.046774193548388e-05,
+      "loss": 0.1705,
+      "step": 12611
+    },
+    {
+      "epoch": 0.201792,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.046612903225806e-05,
+      "loss": 0.1617,
+      "step": 12612
+    },
+    {
+      "epoch": 0.201808,
+      "grad_norm": 1.3125,
+      "learning_rate": 8.046451612903226e-05,
+      "loss": 0.1972,
+      "step": 12613
+    },
+    {
+      "epoch": 0.201824,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.046290322580645e-05,
+      "loss": 0.1567,
+      "step": 12614
+    },
+    {
+      "epoch": 0.20184,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.046129032258065e-05,
+      "loss": 0.1466,
+      "step": 12615
+    },
+    {
+      "epoch": 0.201856,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.045967741935483e-05,
+      "loss": 0.1901,
+      "step": 12616
+    },
+    {
+      "epoch": 0.201872,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.045806451612903e-05,
+      "loss": 0.178,
+      "step": 12617
+    },
+    {
+      "epoch": 0.201888,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.045645161290323e-05,
+      "loss": 0.153,
+      "step": 12618
+    },
+    {
+      "epoch": 0.201904,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.045483870967743e-05,
+      "loss": 0.1943,
+      "step": 12619
+    },
+    {
+      "epoch": 0.20192,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.045322580645162e-05,
+      "loss": 0.1874,
+      "step": 12620
+    },
+    {
+      "epoch": 0.201936,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.045161290322582e-05,
+      "loss": 0.1639,
+      "step": 12621
+    },
+    {
+      "epoch": 0.201952,
+      "grad_norm": 0.625,
+      "learning_rate": 8.045e-05,
+      "loss": 0.1977,
+      "step": 12622
+    },
+    {
+      "epoch": 0.201968,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.04483870967742e-05,
+      "loss": 0.1933,
+      "step": 12623
+    },
+    {
+      "epoch": 0.201984,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.044677419354839e-05,
+      "loss": 0.1522,
+      "step": 12624
+    },
+    {
+      "epoch": 0.202,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.044516129032258e-05,
+      "loss": 0.1703,
+      "step": 12625
+    },
+    {
+      "epoch": 0.202016,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.044354838709678e-05,
+      "loss": 0.1366,
+      "step": 12626
+    },
+    {
+      "epoch": 0.202032,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.044193548387096e-05,
+      "loss": 0.1814,
+      "step": 12627
+    },
+    {
+      "epoch": 0.202048,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.044032258064516e-05,
+      "loss": 0.1729,
+      "step": 12628
+    },
+    {
+      "epoch": 0.202064,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.043870967741936e-05,
+      "loss": 0.2214,
+      "step": 12629
+    },
+    {
+      "epoch": 0.20208,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.043709677419356e-05,
+      "loss": 0.1704,
+      "step": 12630
+    },
+    {
+      "epoch": 0.202096,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.043548387096775e-05,
+      "loss": 0.1451,
+      "step": 12631
+    },
+    {
+      "epoch": 0.202112,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.043387096774195e-05,
+      "loss": 0.1518,
+      "step": 12632
+    },
+    {
+      "epoch": 0.202128,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.043225806451613e-05,
+      "loss": 0.1558,
+      "step": 12633
+    },
+    {
+      "epoch": 0.202144,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.043064516129033e-05,
+      "loss": 0.1665,
+      "step": 12634
+    },
+    {
+      "epoch": 0.20216,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.042903225806452e-05,
+      "loss": 0.187,
+      "step": 12635
+    },
+    {
+      "epoch": 0.202176,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.042741935483872e-05,
+      "loss": 0.1963,
+      "step": 12636
+    },
+    {
+      "epoch": 0.202192,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.04258064516129e-05,
+      "loss": 0.1584,
+      "step": 12637
+    },
+    {
+      "epoch": 0.202208,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.04241935483871e-05,
+      "loss": 0.1637,
+      "step": 12638
+    },
+    {
+      "epoch": 0.202224,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.042258064516129e-05,
+      "loss": 0.1649,
+      "step": 12639
+    },
+    {
+      "epoch": 0.20224,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.042096774193549e-05,
+      "loss": 0.1794,
+      "step": 12640
+    },
+    {
+      "epoch": 0.202256,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.041935483870969e-05,
+      "loss": 0.1852,
+      "step": 12641
+    },
+    {
+      "epoch": 0.202272,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.041774193548387e-05,
+      "loss": 0.1336,
+      "step": 12642
+    },
+    {
+      "epoch": 0.202288,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.041612903225807e-05,
+      "loss": 0.1564,
+      "step": 12643
+    },
+    {
+      "epoch": 0.202304,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.041451612903226e-05,
+      "loss": 0.1646,
+      "step": 12644
+    },
+    {
+      "epoch": 0.20232,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.041290322580646e-05,
+      "loss": 0.1583,
+      "step": 12645
+    },
+    {
+      "epoch": 0.202336,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.041129032258065e-05,
+      "loss": 0.1687,
+      "step": 12646
+    },
+    {
+      "epoch": 0.202352,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.040967741935485e-05,
+      "loss": 0.2053,
+      "step": 12647
+    },
+    {
+      "epoch": 0.202368,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.040806451612903e-05,
+      "loss": 0.1589,
+      "step": 12648
+    },
+    {
+      "epoch": 0.202384,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.040645161290323e-05,
+      "loss": 0.1694,
+      "step": 12649
+    },
+    {
+      "epoch": 0.2024,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.040483870967742e-05,
+      "loss": 0.177,
+      "step": 12650
+    },
+    {
+      "epoch": 0.202416,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.040322580645162e-05,
+      "loss": 0.216,
+      "step": 12651
+    },
+    {
+      "epoch": 0.202432,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.04016129032258e-05,
+      "loss": 0.1975,
+      "step": 12652
+    },
+    {
+      "epoch": 0.202448,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.04e-05,
+      "loss": 0.1629,
+      "step": 12653
+    },
+    {
+      "epoch": 0.202464,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.03983870967742e-05,
+      "loss": 0.17,
+      "step": 12654
+    },
+    {
+      "epoch": 0.20248,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.03967741935484e-05,
+      "loss": 0.1636,
+      "step": 12655
+    },
+    {
+      "epoch": 0.202496,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.039516129032259e-05,
+      "loss": 0.1563,
+      "step": 12656
+    },
+    {
+      "epoch": 0.202512,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.039354838709677e-05,
+      "loss": 0.188,
+      "step": 12657
+    },
+    {
+      "epoch": 0.202528,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.039193548387097e-05,
+      "loss": 0.1713,
+      "step": 12658
+    },
+    {
+      "epoch": 0.202544,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.039032258064516e-05,
+      "loss": 0.1552,
+      "step": 12659
+    },
+    {
+      "epoch": 0.20256,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.038870967741936e-05,
+      "loss": 0.1784,
+      "step": 12660
+    },
+    {
+      "epoch": 0.202576,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.038709677419355e-05,
+      "loss": 0.2002,
+      "step": 12661
+    },
+    {
+      "epoch": 0.202592,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.038548387096775e-05,
+      "loss": 0.175,
+      "step": 12662
+    },
+    {
+      "epoch": 0.202608,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.038387096774193e-05,
+      "loss": 0.1853,
+      "step": 12663
+    },
+    {
+      "epoch": 0.202624,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.038225806451613e-05,
+      "loss": 0.1727,
+      "step": 12664
+    },
+    {
+      "epoch": 0.20264,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.038064516129033e-05,
+      "loss": 0.1488,
+      "step": 12665
+    },
+    {
+      "epoch": 0.202656,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.037903225806453e-05,
+      "loss": 0.1599,
+      "step": 12666
+    },
+    {
+      "epoch": 0.202672,
+      "grad_norm": 0.462890625,
+      "learning_rate": 8.037741935483872e-05,
+      "loss": 0.1498,
+      "step": 12667
+    },
+    {
+      "epoch": 0.202688,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.037580645161292e-05,
+      "loss": 0.1879,
+      "step": 12668
+    },
+    {
+      "epoch": 0.202704,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.03741935483871e-05,
+      "loss": 0.193,
+      "step": 12669
+    },
+    {
+      "epoch": 0.20272,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.03725806451613e-05,
+      "loss": 0.1643,
+      "step": 12670
+    },
+    {
+      "epoch": 0.202736,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.037096774193549e-05,
+      "loss": 0.1866,
+      "step": 12671
+    },
+    {
+      "epoch": 0.202752,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.036935483870967e-05,
+      "loss": 0.187,
+      "step": 12672
+    },
+    {
+      "epoch": 0.202768,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.036774193548387e-05,
+      "loss": 0.1616,
+      "step": 12673
+    },
+    {
+      "epoch": 0.202784,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.036612903225806e-05,
+      "loss": 0.1662,
+      "step": 12674
+    },
+    {
+      "epoch": 0.2028,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.036451612903226e-05,
+      "loss": 0.1678,
+      "step": 12675
+    },
+    {
+      "epoch": 0.202816,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.036290322580645e-05,
+      "loss": 0.1623,
+      "step": 12676
+    },
+    {
+      "epoch": 0.202832,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.036129032258064e-05,
+      "loss": 0.1662,
+      "step": 12677
+    },
+    {
+      "epoch": 0.202848,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.035967741935484e-05,
+      "loss": 0.1534,
+      "step": 12678
+    },
+    {
+      "epoch": 0.202864,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.035806451612904e-05,
+      "loss": 0.1993,
+      "step": 12679
+    },
+    {
+      "epoch": 0.20288,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.035645161290323e-05,
+      "loss": 0.1473,
+      "step": 12680
+    },
+    {
+      "epoch": 0.202896,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.035483870967743e-05,
+      "loss": 0.2039,
+      "step": 12681
+    },
+    {
+      "epoch": 0.202912,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.035322580645162e-05,
+      "loss": 0.1806,
+      "step": 12682
+    },
+    {
+      "epoch": 0.202928,
+      "grad_norm": 0.75,
+      "learning_rate": 8.035161290322582e-05,
+      "loss": 0.1455,
+      "step": 12683
+    },
+    {
+      "epoch": 0.202944,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.035e-05,
+      "loss": 0.1893,
+      "step": 12684
+    },
+    {
+      "epoch": 0.20296,
+      "grad_norm": 1.2109375,
+      "learning_rate": 8.03483870967742e-05,
+      "loss": 0.1694,
+      "step": 12685
+    },
+    {
+      "epoch": 0.202976,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.034677419354839e-05,
+      "loss": 0.1612,
+      "step": 12686
+    },
+    {
+      "epoch": 0.202992,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.034516129032257e-05,
+      "loss": 0.1507,
+      "step": 12687
+    },
+    {
+      "epoch": 0.203008,
+      "grad_norm": 0.875,
+      "learning_rate": 8.034354838709677e-05,
+      "loss": 0.175,
+      "step": 12688
+    },
+    {
+      "epoch": 0.203024,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.034193548387097e-05,
+      "loss": 0.2115,
+      "step": 12689
+    },
+    {
+      "epoch": 0.20304,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.034032258064517e-05,
+      "loss": 0.1553,
+      "step": 12690
+    },
+    {
+      "epoch": 0.203056,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.033870967741936e-05,
+      "loss": 0.1666,
+      "step": 12691
+    },
+    {
+      "epoch": 0.203072,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.033709677419356e-05,
+      "loss": 0.1458,
+      "step": 12692
+    },
+    {
+      "epoch": 0.203088,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.033548387096774e-05,
+      "loss": 0.1236,
+      "step": 12693
+    },
+    {
+      "epoch": 0.203104,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.033387096774194e-05,
+      "loss": 0.2036,
+      "step": 12694
+    },
+    {
+      "epoch": 0.20312,
+      "grad_norm": 1.34375,
+      "learning_rate": 8.033225806451613e-05,
+      "loss": 0.1949,
+      "step": 12695
+    },
+    {
+      "epoch": 0.203136,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.033064516129033e-05,
+      "loss": 0.1727,
+      "step": 12696
+    },
+    {
+      "epoch": 0.203152,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.032903225806452e-05,
+      "loss": 0.1812,
+      "step": 12697
+    },
+    {
+      "epoch": 0.203168,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.032741935483871e-05,
+      "loss": 0.2003,
+      "step": 12698
+    },
+    {
+      "epoch": 0.203184,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.03258064516129e-05,
+      "loss": 0.1565,
+      "step": 12699
+    },
+    {
+      "epoch": 0.2032,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.03241935483871e-05,
+      "loss": 0.186,
+      "step": 12700
+    },
+    {
+      "epoch": 0.203216,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.03225806451613e-05,
+      "loss": 0.1816,
+      "step": 12701
+    },
+    {
+      "epoch": 0.203232,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.032096774193549e-05,
+      "loss": 0.165,
+      "step": 12702
+    },
+    {
+      "epoch": 0.203248,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.031935483870969e-05,
+      "loss": 0.1588,
+      "step": 12703
+    },
+    {
+      "epoch": 0.203264,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.031774193548387e-05,
+      "loss": 0.1875,
+      "step": 12704
+    },
+    {
+      "epoch": 0.20328,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.031612903225807e-05,
+      "loss": 0.1626,
+      "step": 12705
+    },
+    {
+      "epoch": 0.203296,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.031451612903226e-05,
+      "loss": 0.1612,
+      "step": 12706
+    },
+    {
+      "epoch": 0.203312,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.031290322580646e-05,
+      "loss": 0.187,
+      "step": 12707
+    },
+    {
+      "epoch": 0.203328,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.031129032258064e-05,
+      "loss": 0.1554,
+      "step": 12708
+    },
+    {
+      "epoch": 0.203344,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.030967741935484e-05,
+      "loss": 0.1244,
+      "step": 12709
+    },
+    {
+      "epoch": 0.20336,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.030806451612903e-05,
+      "loss": 0.1863,
+      "step": 12710
+    },
+    {
+      "epoch": 0.203376,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.030645161290323e-05,
+      "loss": 0.1735,
+      "step": 12711
+    },
+    {
+      "epoch": 0.203392,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.030483870967741e-05,
+      "loss": 0.2022,
+      "step": 12712
+    },
+    {
+      "epoch": 0.203408,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.030322580645161e-05,
+      "loss": 0.1609,
+      "step": 12713
+    },
+    {
+      "epoch": 0.203424,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.030161290322581e-05,
+      "loss": 0.1808,
+      "step": 12714
+    },
+    {
+      "epoch": 0.20344,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.030000000000001e-05,
+      "loss": 0.1895,
+      "step": 12715
+    },
+    {
+      "epoch": 0.203456,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.02983870967742e-05,
+      "loss": 0.1827,
+      "step": 12716
+    },
+    {
+      "epoch": 0.203472,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.02967741935484e-05,
+      "loss": 0.177,
+      "step": 12717
+    },
+    {
+      "epoch": 0.203488,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.029516129032259e-05,
+      "loss": 0.1724,
+      "step": 12718
+    },
+    {
+      "epoch": 0.203504,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.029354838709677e-05,
+      "loss": 0.1667,
+      "step": 12719
+    },
+    {
+      "epoch": 0.20352,
+      "grad_norm": 1.9453125,
+      "learning_rate": 8.029193548387097e-05,
+      "loss": 0.1923,
+      "step": 12720
+    },
+    {
+      "epoch": 0.203536,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.029032258064516e-05,
+      "loss": 0.1529,
+      "step": 12721
+    },
+    {
+      "epoch": 0.203552,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.028870967741936e-05,
+      "loss": 0.1749,
+      "step": 12722
+    },
+    {
+      "epoch": 0.203568,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.028709677419354e-05,
+      "loss": 0.1619,
+      "step": 12723
+    },
+    {
+      "epoch": 0.203584,
+      "grad_norm": 0.455078125,
+      "learning_rate": 8.028548387096774e-05,
+      "loss": 0.159,
+      "step": 12724
+    },
+    {
+      "epoch": 0.2036,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.028387096774194e-05,
+      "loss": 0.1619,
+      "step": 12725
+    },
+    {
+      "epoch": 0.203616,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.028225806451614e-05,
+      "loss": 0.1652,
+      "step": 12726
+    },
+    {
+      "epoch": 0.203632,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.028064516129033e-05,
+      "loss": 0.1801,
+      "step": 12727
+    },
+    {
+      "epoch": 0.203648,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.027903225806453e-05,
+      "loss": 0.1406,
+      "step": 12728
+    },
+    {
+      "epoch": 0.203664,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.027741935483871e-05,
+      "loss": 0.2058,
+      "step": 12729
+    },
+    {
+      "epoch": 0.20368,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.027580645161291e-05,
+      "loss": 0.189,
+      "step": 12730
+    },
+    {
+      "epoch": 0.203696,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.02741935483871e-05,
+      "loss": 0.1661,
+      "step": 12731
+    },
+    {
+      "epoch": 0.203712,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.02725806451613e-05,
+      "loss": 0.1434,
+      "step": 12732
+    },
+    {
+      "epoch": 0.203728,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.027096774193549e-05,
+      "loss": 0.1419,
+      "step": 12733
+    },
+    {
+      "epoch": 0.203744,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.026935483870967e-05,
+      "loss": 0.165,
+      "step": 12734
+    },
+    {
+      "epoch": 0.20376,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.026774193548387e-05,
+      "loss": 0.1448,
+      "step": 12735
+    },
+    {
+      "epoch": 0.203776,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.026612903225807e-05,
+      "loss": 0.1686,
+      "step": 12736
+    },
+    {
+      "epoch": 0.203792,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.026451612903226e-05,
+      "loss": 0.1521,
+      "step": 12737
+    },
+    {
+      "epoch": 0.203808,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.026290322580646e-05,
+      "loss": 0.1572,
+      "step": 12738
+    },
+    {
+      "epoch": 0.203824,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.026129032258066e-05,
+      "loss": 0.1154,
+      "step": 12739
+    },
+    {
+      "epoch": 0.20384,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.025967741935484e-05,
+      "loss": 0.1964,
+      "step": 12740
+    },
+    {
+      "epoch": 0.203856,
+      "grad_norm": 0.875,
+      "learning_rate": 8.025806451612904e-05,
+      "loss": 0.174,
+      "step": 12741
+    },
+    {
+      "epoch": 0.203872,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.025645161290323e-05,
+      "loss": 0.1504,
+      "step": 12742
+    },
+    {
+      "epoch": 0.203888,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.025483870967743e-05,
+      "loss": 0.1446,
+      "step": 12743
+    },
+    {
+      "epoch": 0.203904,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.025322580645161e-05,
+      "loss": 0.1658,
+      "step": 12744
+    },
+    {
+      "epoch": 0.20392,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.025161290322581e-05,
+      "loss": 0.1726,
+      "step": 12745
+    },
+    {
+      "epoch": 0.203936,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.025e-05,
+      "loss": 0.1699,
+      "step": 12746
+    },
+    {
+      "epoch": 0.203952,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.02483870967742e-05,
+      "loss": 0.1826,
+      "step": 12747
+    },
+    {
+      "epoch": 0.203968,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.024677419354838e-05,
+      "loss": 0.1864,
+      "step": 12748
+    },
+    {
+      "epoch": 0.203984,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.024516129032258e-05,
+      "loss": 0.1721,
+      "step": 12749
+    },
+    {
+      "epoch": 0.204,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.024354838709678e-05,
+      "loss": 0.1465,
+      "step": 12750
+    },
+    {
+      "epoch": 0.204016,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.024193548387097e-05,
+      "loss": 0.1429,
+      "step": 12751
+    },
+    {
+      "epoch": 0.204032,
+      "grad_norm": 0.875,
+      "learning_rate": 8.024032258064517e-05,
+      "loss": 0.1648,
+      "step": 12752
+    },
+    {
+      "epoch": 0.204048,
+      "grad_norm": 1.3359375,
+      "learning_rate": 8.023870967741936e-05,
+      "loss": 0.1822,
+      "step": 12753
+    },
+    {
+      "epoch": 0.204064,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.023709677419356e-05,
+      "loss": 0.1632,
+      "step": 12754
+    },
+    {
+      "epoch": 0.20408,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.023548387096774e-05,
+      "loss": 0.1676,
+      "step": 12755
+    },
+    {
+      "epoch": 0.204096,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.023387096774194e-05,
+      "loss": 0.1637,
+      "step": 12756
+    },
+    {
+      "epoch": 0.204112,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.023225806451613e-05,
+      "loss": 0.2016,
+      "step": 12757
+    },
+    {
+      "epoch": 0.204128,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.023064516129033e-05,
+      "loss": 0.2055,
+      "step": 12758
+    },
+    {
+      "epoch": 0.204144,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.022903225806451e-05,
+      "loss": 0.1437,
+      "step": 12759
+    },
+    {
+      "epoch": 0.20416,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.022741935483871e-05,
+      "loss": 0.1578,
+      "step": 12760
+    },
+    {
+      "epoch": 0.204176,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.022580645161291e-05,
+      "loss": 0.1961,
+      "step": 12761
+    },
+    {
+      "epoch": 0.204192,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.022419354838711e-05,
+      "loss": 0.1177,
+      "step": 12762
+    },
+    {
+      "epoch": 0.204208,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.02225806451613e-05,
+      "loss": 0.2154,
+      "step": 12763
+    },
+    {
+      "epoch": 0.204224,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.02209677419355e-05,
+      "loss": 0.2099,
+      "step": 12764
+    },
+    {
+      "epoch": 0.20424,
+      "grad_norm": 1.3046875,
+      "learning_rate": 8.021935483870968e-05,
+      "loss": 0.1886,
+      "step": 12765
+    },
+    {
+      "epoch": 0.204256,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.021774193548387e-05,
+      "loss": 0.187,
+      "step": 12766
+    },
+    {
+      "epoch": 0.204272,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.021612903225807e-05,
+      "loss": 0.145,
+      "step": 12767
+    },
+    {
+      "epoch": 0.204288,
+      "grad_norm": 1.359375,
+      "learning_rate": 8.021451612903226e-05,
+      "loss": 0.2227,
+      "step": 12768
+    },
+    {
+      "epoch": 0.204304,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.021290322580646e-05,
+      "loss": 0.1529,
+      "step": 12769
+    },
+    {
+      "epoch": 0.20432,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.021129032258064e-05,
+      "loss": 0.2326,
+      "step": 12770
+    },
+    {
+      "epoch": 0.204336,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.020967741935484e-05,
+      "loss": 0.2202,
+      "step": 12771
+    },
+    {
+      "epoch": 0.204352,
+      "grad_norm": 1.296875,
+      "learning_rate": 8.020806451612903e-05,
+      "loss": 0.173,
+      "step": 12772
+    },
+    {
+      "epoch": 0.204368,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.020645161290323e-05,
+      "loss": 0.1749,
+      "step": 12773
+    },
+    {
+      "epoch": 0.204384,
+      "grad_norm": 1.125,
+      "learning_rate": 8.020483870967743e-05,
+      "loss": 0.1595,
+      "step": 12774
+    },
+    {
+      "epoch": 0.2044,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.020322580645163e-05,
+      "loss": 0.1838,
+      "step": 12775
+    },
+    {
+      "epoch": 0.204416,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.020161290322581e-05,
+      "loss": 0.1759,
+      "step": 12776
+    },
+    {
+      "epoch": 0.204432,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.020000000000001e-05,
+      "loss": 0.1805,
+      "step": 12777
+    },
+    {
+      "epoch": 0.204448,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.01983870967742e-05,
+      "loss": 0.1877,
+      "step": 12778
+    },
+    {
+      "epoch": 0.204464,
+      "grad_norm": 0.625,
+      "learning_rate": 8.01967741935484e-05,
+      "loss": 0.1705,
+      "step": 12779
+    },
+    {
+      "epoch": 0.20448,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.019516129032258e-05,
+      "loss": 0.1701,
+      "step": 12780
+    },
+    {
+      "epoch": 0.204496,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.019354838709677e-05,
+      "loss": 0.1625,
+      "step": 12781
+    },
+    {
+      "epoch": 0.204512,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.019193548387097e-05,
+      "loss": 0.1849,
+      "step": 12782
+    },
+    {
+      "epoch": 0.204528,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.019032258064515e-05,
+      "loss": 0.1491,
+      "step": 12783
+    },
+    {
+      "epoch": 0.204544,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.018870967741935e-05,
+      "loss": 0.145,
+      "step": 12784
+    },
+    {
+      "epoch": 0.20456,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.018709677419355e-05,
+      "loss": 0.2006,
+      "step": 12785
+    },
+    {
+      "epoch": 0.204576,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.018548387096775e-05,
+      "loss": 0.216,
+      "step": 12786
+    },
+    {
+      "epoch": 0.204592,
+      "grad_norm": 1.8125,
+      "learning_rate": 8.018387096774194e-05,
+      "loss": 0.1993,
+      "step": 12787
+    },
+    {
+      "epoch": 0.204608,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.018225806451614e-05,
+      "loss": 0.198,
+      "step": 12788
+    },
+    {
+      "epoch": 0.204624,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.018064516129033e-05,
+      "loss": 0.1512,
+      "step": 12789
+    },
+    {
+      "epoch": 0.20464,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.017903225806453e-05,
+      "loss": 0.1415,
+      "step": 12790
+    },
+    {
+      "epoch": 0.204656,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.017741935483871e-05,
+      "loss": 0.1829,
+      "step": 12791
+    },
+    {
+      "epoch": 0.204672,
+      "grad_norm": 1.3046875,
+      "learning_rate": 8.017580645161291e-05,
+      "loss": 0.1797,
+      "step": 12792
+    },
+    {
+      "epoch": 0.204688,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.01741935483871e-05,
+      "loss": 0.2679,
+      "step": 12793
+    },
+    {
+      "epoch": 0.204704,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.01725806451613e-05,
+      "loss": 0.1528,
+      "step": 12794
+    },
+    {
+      "epoch": 0.20472,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.017096774193548e-05,
+      "loss": 0.166,
+      "step": 12795
+    },
+    {
+      "epoch": 0.204736,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.016935483870968e-05,
+      "loss": 0.2063,
+      "step": 12796
+    },
+    {
+      "epoch": 0.204752,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.016774193548388e-05,
+      "loss": 0.194,
+      "step": 12797
+    },
+    {
+      "epoch": 0.204768,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.016612903225807e-05,
+      "loss": 0.2042,
+      "step": 12798
+    },
+    {
+      "epoch": 0.204784,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.016451612903227e-05,
+      "loss": 0.1551,
+      "step": 12799
+    },
+    {
+      "epoch": 0.2048,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.016290322580645e-05,
+      "loss": 0.1683,
+      "step": 12800
+    },
+    {
+      "epoch": 0.204816,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.016129032258065e-05,
+      "loss": 0.1867,
+      "step": 12801
+    },
+    {
+      "epoch": 0.204832,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.015967741935484e-05,
+      "loss": 0.1973,
+      "step": 12802
+    },
+    {
+      "epoch": 0.204848,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.015806451612904e-05,
+      "loss": 0.1457,
+      "step": 12803
+    },
+    {
+      "epoch": 0.204864,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.015645161290323e-05,
+      "loss": 0.1833,
+      "step": 12804
+    },
+    {
+      "epoch": 0.20488,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.015483870967742e-05,
+      "loss": 0.1661,
+      "step": 12805
+    },
+    {
+      "epoch": 0.204896,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.015322580645161e-05,
+      "loss": 0.1617,
+      "step": 12806
+    },
+    {
+      "epoch": 0.204912,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.015161290322581e-05,
+      "loss": 0.2176,
+      "step": 12807
+    },
+    {
+      "epoch": 0.204928,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.015e-05,
+      "loss": 0.1848,
+      "step": 12808
+    },
+    {
+      "epoch": 0.204944,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.01483870967742e-05,
+      "loss": 0.2047,
+      "step": 12809
+    },
+    {
+      "epoch": 0.20496,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.01467741935484e-05,
+      "loss": 0.1873,
+      "step": 12810
+    },
+    {
+      "epoch": 0.204976,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.014516129032258e-05,
+      "loss": 0.221,
+      "step": 12811
+    },
+    {
+      "epoch": 0.204992,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.014354838709678e-05,
+      "loss": 0.1996,
+      "step": 12812
+    },
+    {
+      "epoch": 0.205008,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.014193548387097e-05,
+      "loss": 0.1875,
+      "step": 12813
+    },
+    {
+      "epoch": 0.205024,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.014032258064517e-05,
+      "loss": 0.142,
+      "step": 12814
+    },
+    {
+      "epoch": 0.20504,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.013870967741935e-05,
+      "loss": 0.1473,
+      "step": 12815
+    },
+    {
+      "epoch": 0.205056,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.013709677419355e-05,
+      "loss": 0.1455,
+      "step": 12816
+    },
+    {
+      "epoch": 0.205072,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.013548387096774e-05,
+      "loss": 0.1046,
+      "step": 12817
+    },
+    {
+      "epoch": 0.205088,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.013387096774194e-05,
+      "loss": 0.1948,
+      "step": 12818
+    },
+    {
+      "epoch": 0.205104,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.013225806451612e-05,
+      "loss": 0.1553,
+      "step": 12819
+    },
+    {
+      "epoch": 0.20512,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.013064516129032e-05,
+      "loss": 0.1366,
+      "step": 12820
+    },
+    {
+      "epoch": 0.205136,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.012903225806452e-05,
+      "loss": 0.1841,
+      "step": 12821
+    },
+    {
+      "epoch": 0.205152,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.012741935483872e-05,
+      "loss": 0.1651,
+      "step": 12822
+    },
+    {
+      "epoch": 0.205168,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.012580645161291e-05,
+      "loss": 0.1756,
+      "step": 12823
+    },
+    {
+      "epoch": 0.205184,
+      "grad_norm": 0.95703125,
+      "learning_rate": 8.012419354838711e-05,
+      "loss": 0.1473,
+      "step": 12824
+    },
+    {
+      "epoch": 0.2052,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.01225806451613e-05,
+      "loss": 0.1861,
+      "step": 12825
+    },
+    {
+      "epoch": 0.205216,
+      "grad_norm": 0.46484375,
+      "learning_rate": 8.01209677419355e-05,
+      "loss": 0.1236,
+      "step": 12826
+    },
+    {
+      "epoch": 0.205232,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.011935483870968e-05,
+      "loss": 0.2089,
+      "step": 12827
+    },
+    {
+      "epoch": 0.205248,
+      "grad_norm": 1.0,
+      "learning_rate": 8.011774193548387e-05,
+      "loss": 0.1782,
+      "step": 12828
+    },
+    {
+      "epoch": 0.205264,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.011612903225807e-05,
+      "loss": 0.183,
+      "step": 12829
+    },
+    {
+      "epoch": 0.20528,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.011451612903225e-05,
+      "loss": 0.2111,
+      "step": 12830
+    },
+    {
+      "epoch": 0.205296,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.011290322580645e-05,
+      "loss": 0.1411,
+      "step": 12831
+    },
+    {
+      "epoch": 0.205312,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.011129032258065e-05,
+      "loss": 0.1761,
+      "step": 12832
+    },
+    {
+      "epoch": 0.205328,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.010967741935484e-05,
+      "loss": 0.1836,
+      "step": 12833
+    },
+    {
+      "epoch": 0.205344,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.010806451612904e-05,
+      "loss": 0.1379,
+      "step": 12834
+    },
+    {
+      "epoch": 0.20536,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.010645161290324e-05,
+      "loss": 0.1538,
+      "step": 12835
+    },
+    {
+      "epoch": 0.205376,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.010483870967742e-05,
+      "loss": 0.1742,
+      "step": 12836
+    },
+    {
+      "epoch": 0.205392,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.010322580645162e-05,
+      "loss": 0.1936,
+      "step": 12837
+    },
+    {
+      "epoch": 0.205408,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.010161290322581e-05,
+      "loss": 0.1738,
+      "step": 12838
+    },
+    {
+      "epoch": 0.205424,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.010000000000001e-05,
+      "loss": 0.1701,
+      "step": 12839
+    },
+    {
+      "epoch": 0.20544,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.00983870967742e-05,
+      "loss": 0.1551,
+      "step": 12840
+    },
+    {
+      "epoch": 0.205456,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.00967741935484e-05,
+      "loss": 0.1968,
+      "step": 12841
+    },
+    {
+      "epoch": 0.205472,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.009516129032258e-05,
+      "loss": 0.2098,
+      "step": 12842
+    },
+    {
+      "epoch": 0.205488,
+      "grad_norm": 0.95703125,
+      "learning_rate": 8.009354838709677e-05,
+      "loss": 0.1911,
+      "step": 12843
+    },
+    {
+      "epoch": 0.205504,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.009193548387097e-05,
+      "loss": 0.1885,
+      "step": 12844
+    },
+    {
+      "epoch": 0.20552,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.009032258064517e-05,
+      "loss": 0.1485,
+      "step": 12845
+    },
+    {
+      "epoch": 0.205536,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.008870967741937e-05,
+      "loss": 0.1287,
+      "step": 12846
+    },
+    {
+      "epoch": 0.205552,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.008709677419355e-05,
+      "loss": 0.1509,
+      "step": 12847
+    },
+    {
+      "epoch": 0.205568,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.008548387096775e-05,
+      "loss": 0.1651,
+      "step": 12848
+    },
+    {
+      "epoch": 0.205584,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.008387096774194e-05,
+      "loss": 0.1702,
+      "step": 12849
+    },
+    {
+      "epoch": 0.2056,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.008225806451614e-05,
+      "loss": 0.1706,
+      "step": 12850
+    },
+    {
+      "epoch": 0.205616,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.008064516129032e-05,
+      "loss": 0.1763,
+      "step": 12851
+    },
+    {
+      "epoch": 0.205632,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.007903225806452e-05,
+      "loss": 0.1803,
+      "step": 12852
+    },
+    {
+      "epoch": 0.205648,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.007741935483871e-05,
+      "loss": 0.157,
+      "step": 12853
+    },
+    {
+      "epoch": 0.205664,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.007580645161291e-05,
+      "loss": 0.1718,
+      "step": 12854
+    },
+    {
+      "epoch": 0.20568,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.00741935483871e-05,
+      "loss": 0.1851,
+      "step": 12855
+    },
+    {
+      "epoch": 0.205696,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.00725806451613e-05,
+      "loss": 0.1995,
+      "step": 12856
+    },
+    {
+      "epoch": 0.205712,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.00709677419355e-05,
+      "loss": 0.1747,
+      "step": 12857
+    },
+    {
+      "epoch": 0.205728,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.006935483870968e-05,
+      "loss": 0.1629,
+      "step": 12858
+    },
+    {
+      "epoch": 0.205744,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.006774193548388e-05,
+      "loss": 0.1857,
+      "step": 12859
+    },
+    {
+      "epoch": 0.20576,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.006612903225807e-05,
+      "loss": 0.1876,
+      "step": 12860
+    },
+    {
+      "epoch": 0.205776,
+      "grad_norm": 0.625,
+      "learning_rate": 8.006451612903227e-05,
+      "loss": 0.1799,
+      "step": 12861
+    },
+    {
+      "epoch": 0.205792,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.006290322580645e-05,
+      "loss": 0.1846,
+      "step": 12862
+    },
+    {
+      "epoch": 0.205808,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.006129032258065e-05,
+      "loss": 0.2069,
+      "step": 12863
+    },
+    {
+      "epoch": 0.205824,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.005967741935484e-05,
+      "loss": 0.1562,
+      "step": 12864
+    },
+    {
+      "epoch": 0.20584,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.005806451612904e-05,
+      "loss": 0.2164,
+      "step": 12865
+    },
+    {
+      "epoch": 0.205856,
+      "grad_norm": 1.0,
+      "learning_rate": 8.005645161290322e-05,
+      "loss": 0.1719,
+      "step": 12866
+    },
+    {
+      "epoch": 0.205872,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.005483870967742e-05,
+      "loss": 0.1645,
+      "step": 12867
+    },
+    {
+      "epoch": 0.205888,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.005322580645161e-05,
+      "loss": 0.152,
+      "step": 12868
+    },
+    {
+      "epoch": 0.205904,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.005161290322581e-05,
+      "loss": 0.2371,
+      "step": 12869
+    },
+    {
+      "epoch": 0.20592,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.005000000000001e-05,
+      "loss": 0.1333,
+      "step": 12870
+    },
+    {
+      "epoch": 0.205936,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.004838709677421e-05,
+      "loss": 0.1417,
+      "step": 12871
+    },
+    {
+      "epoch": 0.205952,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.00467741935484e-05,
+      "loss": 0.1728,
+      "step": 12872
+    },
+    {
+      "epoch": 0.205968,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.004516129032258e-05,
+      "loss": 0.1585,
+      "step": 12873
+    },
+    {
+      "epoch": 0.205984,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.004354838709678e-05,
+      "loss": 0.1723,
+      "step": 12874
+    },
+    {
+      "epoch": 0.206,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.004193548387097e-05,
+      "loss": 0.1797,
+      "step": 12875
+    },
+    {
+      "epoch": 0.206016,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.004032258064516e-05,
+      "loss": 0.1648,
+      "step": 12876
+    },
+    {
+      "epoch": 0.206032,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.003870967741935e-05,
+      "loss": 0.1446,
+      "step": 12877
+    },
+    {
+      "epoch": 0.206048,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.003709677419355e-05,
+      "loss": 0.1403,
+      "step": 12878
+    },
+    {
+      "epoch": 0.206064,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.003548387096774e-05,
+      "loss": 0.1489,
+      "step": 12879
+    },
+    {
+      "epoch": 0.20608,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.003387096774194e-05,
+      "loss": 0.1275,
+      "step": 12880
+    },
+    {
+      "epoch": 0.206096,
+      "grad_norm": 0.49609375,
+      "learning_rate": 8.003225806451614e-05,
+      "loss": 0.1253,
+      "step": 12881
+    },
+    {
+      "epoch": 0.206112,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.003064516129034e-05,
+      "loss": 0.1392,
+      "step": 12882
+    },
+    {
+      "epoch": 0.206128,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.002903225806452e-05,
+      "loss": 0.1597,
+      "step": 12883
+    },
+    {
+      "epoch": 0.206144,
+      "grad_norm": 0.482421875,
+      "learning_rate": 8.002741935483872e-05,
+      "loss": 0.12,
+      "step": 12884
+    },
+    {
+      "epoch": 0.20616,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.002580645161291e-05,
+      "loss": 0.1748,
+      "step": 12885
+    },
+    {
+      "epoch": 0.206176,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.002419354838711e-05,
+      "loss": 0.181,
+      "step": 12886
+    },
+    {
+      "epoch": 0.206192,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.002258064516129e-05,
+      "loss": 0.1383,
+      "step": 12887
+    },
+    {
+      "epoch": 0.206208,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.002096774193549e-05,
+      "loss": 0.116,
+      "step": 12888
+    },
+    {
+      "epoch": 0.206224,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.001935483870968e-05,
+      "loss": 0.1461,
+      "step": 12889
+    },
+    {
+      "epoch": 0.20624,
+      "grad_norm": 0.75,
+      "learning_rate": 8.001774193548386e-05,
+      "loss": 0.177,
+      "step": 12890
+    },
+    {
+      "epoch": 0.206256,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.001612903225806e-05,
+      "loss": 0.135,
+      "step": 12891
+    },
+    {
+      "epoch": 0.206272,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.001451612903226e-05,
+      "loss": 0.1851,
+      "step": 12892
+    },
+    {
+      "epoch": 0.206288,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.001290322580646e-05,
+      "loss": 0.154,
+      "step": 12893
+    },
+    {
+      "epoch": 0.206304,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.001129032258065e-05,
+      "loss": 0.1645,
+      "step": 12894
+    },
+    {
+      "epoch": 0.20632,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.000967741935485e-05,
+      "loss": 0.142,
+      "step": 12895
+    },
+    {
+      "epoch": 0.206336,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.000806451612904e-05,
+      "loss": 0.1947,
+      "step": 12896
+    },
+    {
+      "epoch": 0.206352,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.000645161290324e-05,
+      "loss": 0.1467,
+      "step": 12897
+    },
+    {
+      "epoch": 0.206368,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.000483870967742e-05,
+      "loss": 0.2129,
+      "step": 12898
+    },
+    {
+      "epoch": 0.206384,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.000322580645162e-05,
+      "loss": 0.1632,
+      "step": 12899
+    },
+    {
+      "epoch": 0.2064,
+      "grad_norm": 1.5,
+      "learning_rate": 8.000161290322581e-05,
+      "loss": 0.2008,
+      "step": 12900
+    },
+    {
+      "epoch": 0.206416,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8e-05,
+      "loss": 0.2057,
+      "step": 12901
+    },
+    {
+      "epoch": 0.206432,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.999838709677419e-05,
+      "loss": 0.1648,
+      "step": 12902
+    },
+    {
+      "epoch": 0.206448,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.999677419354839e-05,
+      "loss": 0.2077,
+      "step": 12903
+    },
+    {
+      "epoch": 0.206464,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.999516129032258e-05,
+      "loss": 0.1449,
+      "step": 12904
+    },
+    {
+      "epoch": 0.20648,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.999354838709678e-05,
+      "loss": 0.1579,
+      "step": 12905
+    },
+    {
+      "epoch": 0.206496,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.999193548387098e-05,
+      "loss": 0.1627,
+      "step": 12906
+    },
+    {
+      "epoch": 0.206512,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.999032258064516e-05,
+      "loss": 0.18,
+      "step": 12907
+    },
+    {
+      "epoch": 0.206528,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.998870967741936e-05,
+      "loss": 0.149,
+      "step": 12908
+    },
+    {
+      "epoch": 0.206544,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.998709677419355e-05,
+      "loss": 0.1618,
+      "step": 12909
+    },
+    {
+      "epoch": 0.20656,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.998548387096775e-05,
+      "loss": 0.155,
+      "step": 12910
+    },
+    {
+      "epoch": 0.206576,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.998387096774194e-05,
+      "loss": 0.1875,
+      "step": 12911
+    },
+    {
+      "epoch": 0.206592,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.998225806451613e-05,
+      "loss": 0.1857,
+      "step": 12912
+    },
+    {
+      "epoch": 0.206608,
+      "grad_norm": 1.3671875,
+      "learning_rate": 7.998064516129032e-05,
+      "loss": 0.2197,
+      "step": 12913
+    },
+    {
+      "epoch": 0.206624,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.997903225806452e-05,
+      "loss": 0.194,
+      "step": 12914
+    },
+    {
+      "epoch": 0.20664,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.99774193548387e-05,
+      "loss": 0.1468,
+      "step": 12915
+    },
+    {
+      "epoch": 0.206656,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.99758064516129e-05,
+      "loss": 0.1348,
+      "step": 12916
+    },
+    {
+      "epoch": 0.206672,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.99741935483871e-05,
+      "loss": 0.1701,
+      "step": 12917
+    },
+    {
+      "epoch": 0.206688,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.99725806451613e-05,
+      "loss": 0.1612,
+      "step": 12918
+    },
+    {
+      "epoch": 0.206704,
+      "grad_norm": 1.4453125,
+      "learning_rate": 7.997096774193549e-05,
+      "loss": 0.1735,
+      "step": 12919
+    },
+    {
+      "epoch": 0.20672,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.996935483870968e-05,
+      "loss": 0.144,
+      "step": 12920
+    },
+    {
+      "epoch": 0.206736,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.996774193548388e-05,
+      "loss": 0.1471,
+      "step": 12921
+    },
+    {
+      "epoch": 0.206752,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.996612903225806e-05,
+      "loss": 0.1852,
+      "step": 12922
+    },
+    {
+      "epoch": 0.206768,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.996451612903226e-05,
+      "loss": 0.1726,
+      "step": 12923
+    },
+    {
+      "epoch": 0.206784,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.996290322580645e-05,
+      "loss": 0.2093,
+      "step": 12924
+    },
+    {
+      "epoch": 0.2068,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.996129032258065e-05,
+      "loss": 0.1849,
+      "step": 12925
+    },
+    {
+      "epoch": 0.206816,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.995967741935483e-05,
+      "loss": 0.1944,
+      "step": 12926
+    },
+    {
+      "epoch": 0.206832,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.995806451612903e-05,
+      "loss": 0.1761,
+      "step": 12927
+    },
+    {
+      "epoch": 0.206848,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.995645161290322e-05,
+      "loss": 0.1816,
+      "step": 12928
+    },
+    {
+      "epoch": 0.206864,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.995483870967742e-05,
+      "loss": 0.1657,
+      "step": 12929
+    },
+    {
+      "epoch": 0.20688,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.995322580645162e-05,
+      "loss": 0.2071,
+      "step": 12930
+    },
+    {
+      "epoch": 0.206896,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.995161290322582e-05,
+      "loss": 0.16,
+      "step": 12931
+    },
+    {
+      "epoch": 0.206912,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.995e-05,
+      "loss": 0.1838,
+      "step": 12932
+    },
+    {
+      "epoch": 0.206928,
+      "grad_norm": 1.53125,
+      "learning_rate": 7.99483870967742e-05,
+      "loss": 0.2112,
+      "step": 12933
+    },
+    {
+      "epoch": 0.206944,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.994677419354839e-05,
+      "loss": 0.1597,
+      "step": 12934
+    },
+    {
+      "epoch": 0.20696,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.994516129032259e-05,
+      "loss": 0.1761,
+      "step": 12935
+    },
+    {
+      "epoch": 0.206976,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.994354838709678e-05,
+      "loss": 0.1768,
+      "step": 12936
+    },
+    {
+      "epoch": 0.206992,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.994193548387096e-05,
+      "loss": 0.1846,
+      "step": 12937
+    },
+    {
+      "epoch": 0.207008,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.994032258064516e-05,
+      "loss": 0.1564,
+      "step": 12938
+    },
+    {
+      "epoch": 0.207024,
+      "grad_norm": 1.515625,
+      "learning_rate": 7.993870967741935e-05,
+      "loss": 0.1769,
+      "step": 12939
+    },
+    {
+      "epoch": 0.20704,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.993709677419355e-05,
+      "loss": 0.1749,
+      "step": 12940
+    },
+    {
+      "epoch": 0.207056,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.993548387096775e-05,
+      "loss": 0.2006,
+      "step": 12941
+    },
+    {
+      "epoch": 0.207072,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.993387096774195e-05,
+      "loss": 0.1855,
+      "step": 12942
+    },
+    {
+      "epoch": 0.207088,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.993225806451613e-05,
+      "loss": 0.1701,
+      "step": 12943
+    },
+    {
+      "epoch": 0.207104,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.993064516129033e-05,
+      "loss": 0.2015,
+      "step": 12944
+    },
+    {
+      "epoch": 0.20712,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.992903225806452e-05,
+      "loss": 0.1762,
+      "step": 12945
+    },
+    {
+      "epoch": 0.207136,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.992741935483872e-05,
+      "loss": 0.1836,
+      "step": 12946
+    },
+    {
+      "epoch": 0.207152,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.99258064516129e-05,
+      "loss": 0.1564,
+      "step": 12947
+    },
+    {
+      "epoch": 0.207168,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.99241935483871e-05,
+      "loss": 0.1738,
+      "step": 12948
+    },
+    {
+      "epoch": 0.207184,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.992258064516129e-05,
+      "loss": 0.2304,
+      "step": 12949
+    },
+    {
+      "epoch": 0.2072,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.992096774193549e-05,
+      "loss": 0.1941,
+      "step": 12950
+    },
+    {
+      "epoch": 0.207216,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.991935483870968e-05,
+      "loss": 0.1845,
+      "step": 12951
+    },
+    {
+      "epoch": 0.207232,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.991774193548388e-05,
+      "loss": 0.2174,
+      "step": 12952
+    },
+    {
+      "epoch": 0.207248,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.991612903225808e-05,
+      "loss": 0.1431,
+      "step": 12953
+    },
+    {
+      "epoch": 0.207264,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.991451612903226e-05,
+      "loss": 0.1955,
+      "step": 12954
+    },
+    {
+      "epoch": 0.20728,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.991290322580646e-05,
+      "loss": 0.1576,
+      "step": 12955
+    },
+    {
+      "epoch": 0.207296,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.991129032258065e-05,
+      "loss": 0.1531,
+      "step": 12956
+    },
+    {
+      "epoch": 0.207312,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.990967741935485e-05,
+      "loss": 0.1768,
+      "step": 12957
+    },
+    {
+      "epoch": 0.207328,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.990806451612903e-05,
+      "loss": 0.2144,
+      "step": 12958
+    },
+    {
+      "epoch": 0.207344,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.990645161290323e-05,
+      "loss": 0.1666,
+      "step": 12959
+    },
+    {
+      "epoch": 0.20736,
+      "grad_norm": 1.28125,
+      "learning_rate": 7.990483870967742e-05,
+      "loss": 0.1785,
+      "step": 12960
+    },
+    {
+      "epoch": 0.207376,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.990322580645162e-05,
+      "loss": 0.175,
+      "step": 12961
+    },
+    {
+      "epoch": 0.207392,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.99016129032258e-05,
+      "loss": 0.19,
+      "step": 12962
+    },
+    {
+      "epoch": 0.207408,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.99e-05,
+      "loss": 0.1331,
+      "step": 12963
+    },
+    {
+      "epoch": 0.207424,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.989838709677419e-05,
+      "loss": 0.1946,
+      "step": 12964
+    },
+    {
+      "epoch": 0.20744,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.989677419354839e-05,
+      "loss": 0.1656,
+      "step": 12965
+    },
+    {
+      "epoch": 0.207456,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.989516129032259e-05,
+      "loss": 0.1769,
+      "step": 12966
+    },
+    {
+      "epoch": 0.207472,
+      "grad_norm": 2.125,
+      "learning_rate": 7.989354838709678e-05,
+      "loss": 0.1867,
+      "step": 12967
+    },
+    {
+      "epoch": 0.207488,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.989193548387098e-05,
+      "loss": 0.1344,
+      "step": 12968
+    },
+    {
+      "epoch": 0.207504,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.989032258064516e-05,
+      "loss": 0.1472,
+      "step": 12969
+    },
+    {
+      "epoch": 0.20752,
+      "grad_norm": 1.3984375,
+      "learning_rate": 7.988870967741936e-05,
+      "loss": 0.1477,
+      "step": 12970
+    },
+    {
+      "epoch": 0.207536,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.988709677419355e-05,
+      "loss": 0.1714,
+      "step": 12971
+    },
+    {
+      "epoch": 0.207552,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.988548387096775e-05,
+      "loss": 0.1684,
+      "step": 12972
+    },
+    {
+      "epoch": 0.207568,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.988387096774193e-05,
+      "loss": 0.1528,
+      "step": 12973
+    },
+    {
+      "epoch": 0.207584,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.988225806451613e-05,
+      "loss": 0.2059,
+      "step": 12974
+    },
+    {
+      "epoch": 0.2076,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.988064516129032e-05,
+      "loss": 0.1905,
+      "step": 12975
+    },
+    {
+      "epoch": 0.207616,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.987903225806452e-05,
+      "loss": 0.1827,
+      "step": 12976
+    },
+    {
+      "epoch": 0.207632,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.987741935483872e-05,
+      "loss": 0.1714,
+      "step": 12977
+    },
+    {
+      "epoch": 0.207648,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.987580645161292e-05,
+      "loss": 0.1841,
+      "step": 12978
+    },
+    {
+      "epoch": 0.207664,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.98741935483871e-05,
+      "loss": 0.1613,
+      "step": 12979
+    },
+    {
+      "epoch": 0.20768,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.98725806451613e-05,
+      "loss": 0.1501,
+      "step": 12980
+    },
+    {
+      "epoch": 0.207696,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.987096774193549e-05,
+      "loss": 0.1961,
+      "step": 12981
+    },
+    {
+      "epoch": 0.207712,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.986935483870968e-05,
+      "loss": 0.1945,
+      "step": 12982
+    },
+    {
+      "epoch": 0.207728,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.986774193548387e-05,
+      "loss": 0.192,
+      "step": 12983
+    },
+    {
+      "epoch": 0.207744,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.986612903225806e-05,
+      "loss": 0.1881,
+      "step": 12984
+    },
+    {
+      "epoch": 0.20776,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.986451612903226e-05,
+      "loss": 0.1948,
+      "step": 12985
+    },
+    {
+      "epoch": 0.207776,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.986290322580645e-05,
+      "loss": 0.1664,
+      "step": 12986
+    },
+    {
+      "epoch": 0.207792,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.986129032258065e-05,
+      "loss": 0.1786,
+      "step": 12987
+    },
+    {
+      "epoch": 0.207808,
+      "grad_norm": 1.484375,
+      "learning_rate": 7.985967741935485e-05,
+      "loss": 0.1619,
+      "step": 12988
+    },
+    {
+      "epoch": 0.207824,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.985806451612903e-05,
+      "loss": 0.1985,
+      "step": 12989
+    },
+    {
+      "epoch": 0.20784,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.985645161290323e-05,
+      "loss": 0.188,
+      "step": 12990
+    },
+    {
+      "epoch": 0.207856,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.985483870967743e-05,
+      "loss": 0.1332,
+      "step": 12991
+    },
+    {
+      "epoch": 0.207872,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.985322580645162e-05,
+      "loss": 0.1705,
+      "step": 12992
+    },
+    {
+      "epoch": 0.207888,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.985161290322582e-05,
+      "loss": 0.1933,
+      "step": 12993
+    },
+    {
+      "epoch": 0.207904,
+      "grad_norm": 1.390625,
+      "learning_rate": 7.985e-05,
+      "loss": 0.1947,
+      "step": 12994
+    },
+    {
+      "epoch": 0.20792,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.98483870967742e-05,
+      "loss": 0.1744,
+      "step": 12995
+    },
+    {
+      "epoch": 0.207936,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.984677419354839e-05,
+      "loss": 0.1988,
+      "step": 12996
+    },
+    {
+      "epoch": 0.207952,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.984516129032259e-05,
+      "loss": 0.1879,
+      "step": 12997
+    },
+    {
+      "epoch": 0.207968,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.984354838709677e-05,
+      "loss": 0.1955,
+      "step": 12998
+    },
+    {
+      "epoch": 0.207984,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.984193548387096e-05,
+      "loss": 0.183,
+      "step": 12999
+    },
+    {
+      "epoch": 0.208,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.984032258064516e-05,
+      "loss": 0.162,
+      "step": 13000
+    },
+    {
+      "epoch": 0.208016,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.983870967741936e-05,
+      "loss": 0.1836,
+      "step": 13001
+    },
+    {
+      "epoch": 0.208032,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.983709677419356e-05,
+      "loss": 0.1716,
+      "step": 13002
+    },
+    {
+      "epoch": 0.208048,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.983548387096775e-05,
+      "loss": 0.1633,
+      "step": 13003
+    },
+    {
+      "epoch": 0.208064,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.983387096774194e-05,
+      "loss": 0.1484,
+      "step": 13004
+    },
+    {
+      "epoch": 0.20808,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.983225806451613e-05,
+      "loss": 0.148,
+      "step": 13005
+    },
+    {
+      "epoch": 0.208096,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.983064516129033e-05,
+      "loss": 0.1983,
+      "step": 13006
+    },
+    {
+      "epoch": 0.208112,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.982903225806452e-05,
+      "loss": 0.1988,
+      "step": 13007
+    },
+    {
+      "epoch": 0.208128,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.982741935483872e-05,
+      "loss": 0.1535,
+      "step": 13008
+    },
+    {
+      "epoch": 0.208144,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.98258064516129e-05,
+      "loss": 0.173,
+      "step": 13009
+    },
+    {
+      "epoch": 0.20816,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.98241935483871e-05,
+      "loss": 0.162,
+      "step": 13010
+    },
+    {
+      "epoch": 0.208176,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.982258064516129e-05,
+      "loss": 0.1434,
+      "step": 13011
+    },
+    {
+      "epoch": 0.208192,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.982096774193549e-05,
+      "loss": 0.1779,
+      "step": 13012
+    },
+    {
+      "epoch": 0.208208,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.981935483870969e-05,
+      "loss": 0.1958,
+      "step": 13013
+    },
+    {
+      "epoch": 0.208224,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.981774193548387e-05,
+      "loss": 0.1623,
+      "step": 13014
+    },
+    {
+      "epoch": 0.20824,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.981612903225807e-05,
+      "loss": 0.1919,
+      "step": 13015
+    },
+    {
+      "epoch": 0.208256,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.981451612903226e-05,
+      "loss": 0.1276,
+      "step": 13016
+    },
+    {
+      "epoch": 0.208272,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.981290322580646e-05,
+      "loss": 0.1212,
+      "step": 13017
+    },
+    {
+      "epoch": 0.208288,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.981129032258064e-05,
+      "loss": 0.172,
+      "step": 13018
+    },
+    {
+      "epoch": 0.208304,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.980967741935484e-05,
+      "loss": 0.1822,
+      "step": 13019
+    },
+    {
+      "epoch": 0.20832,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.980806451612903e-05,
+      "loss": 0.1964,
+      "step": 13020
+    },
+    {
+      "epoch": 0.208336,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.980645161290323e-05,
+      "loss": 0.1748,
+      "step": 13021
+    },
+    {
+      "epoch": 0.208352,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.980483870967742e-05,
+      "loss": 0.2067,
+      "step": 13022
+    },
+    {
+      "epoch": 0.208368,
+      "grad_norm": 0.494140625,
+      "learning_rate": 7.980322580645162e-05,
+      "loss": 0.1557,
+      "step": 13023
+    },
+    {
+      "epoch": 0.208384,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.98016129032258e-05,
+      "loss": 0.1881,
+      "step": 13024
+    },
+    {
+      "epoch": 0.2084,
+      "grad_norm": 1.5390625,
+      "learning_rate": 7.98e-05,
+      "loss": 0.198,
+      "step": 13025
+    },
+    {
+      "epoch": 0.208416,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.97983870967742e-05,
+      "loss": 0.1561,
+      "step": 13026
+    },
+    {
+      "epoch": 0.208432,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.97967741935484e-05,
+      "loss": 0.1763,
+      "step": 13027
+    },
+    {
+      "epoch": 0.208448,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.979516129032259e-05,
+      "loss": 0.1951,
+      "step": 13028
+    },
+    {
+      "epoch": 0.208464,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.979354838709677e-05,
+      "loss": 0.1878,
+      "step": 13029
+    },
+    {
+      "epoch": 0.20848,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.979193548387097e-05,
+      "loss": 0.1731,
+      "step": 13030
+    },
+    {
+      "epoch": 0.208496,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.979032258064516e-05,
+      "loss": 0.1935,
+      "step": 13031
+    },
+    {
+      "epoch": 0.208512,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.978870967741936e-05,
+      "loss": 0.1826,
+      "step": 13032
+    },
+    {
+      "epoch": 0.208528,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.978709677419354e-05,
+      "loss": 0.1824,
+      "step": 13033
+    },
+    {
+      "epoch": 0.208544,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.978548387096774e-05,
+      "loss": 0.1538,
+      "step": 13034
+    },
+    {
+      "epoch": 0.20856,
+      "grad_norm": 1.34375,
+      "learning_rate": 7.978387096774193e-05,
+      "loss": 0.1546,
+      "step": 13035
+    },
+    {
+      "epoch": 0.208576,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.978225806451613e-05,
+      "loss": 0.1782,
+      "step": 13036
+    },
+    {
+      "epoch": 0.208592,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.978064516129033e-05,
+      "loss": 0.2026,
+      "step": 13037
+    },
+    {
+      "epoch": 0.208608,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.977903225806453e-05,
+      "loss": 0.1395,
+      "step": 13038
+    },
+    {
+      "epoch": 0.208624,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.977741935483872e-05,
+      "loss": 0.1722,
+      "step": 13039
+    },
+    {
+      "epoch": 0.20864,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.977580645161291e-05,
+      "loss": 0.1531,
+      "step": 13040
+    },
+    {
+      "epoch": 0.208656,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.97741935483871e-05,
+      "loss": 0.2305,
+      "step": 13041
+    },
+    {
+      "epoch": 0.208672,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.97725806451613e-05,
+      "loss": 0.2254,
+      "step": 13042
+    },
+    {
+      "epoch": 0.208688,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.977096774193549e-05,
+      "loss": 0.1838,
+      "step": 13043
+    },
+    {
+      "epoch": 0.208704,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.976935483870969e-05,
+      "loss": 0.1814,
+      "step": 13044
+    },
+    {
+      "epoch": 0.20872,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.976774193548387e-05,
+      "loss": 0.1825,
+      "step": 13045
+    },
+    {
+      "epoch": 0.208736,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.976612903225806e-05,
+      "loss": 0.1529,
+      "step": 13046
+    },
+    {
+      "epoch": 0.208752,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.976451612903226e-05,
+      "loss": 0.1546,
+      "step": 13047
+    },
+    {
+      "epoch": 0.208768,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.976290322580646e-05,
+      "loss": 0.1715,
+      "step": 13048
+    },
+    {
+      "epoch": 0.208784,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.976129032258066e-05,
+      "loss": 0.1657,
+      "step": 13049
+    },
+    {
+      "epoch": 0.2088,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.975967741935484e-05,
+      "loss": 0.2039,
+      "step": 13050
+    },
+    {
+      "epoch": 0.208816,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.975806451612904e-05,
+      "loss": 0.1688,
+      "step": 13051
+    },
+    {
+      "epoch": 0.208832,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.975645161290323e-05,
+      "loss": 0.1405,
+      "step": 13052
+    },
+    {
+      "epoch": 0.208848,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.975483870967743e-05,
+      "loss": 0.1719,
+      "step": 13053
+    },
+    {
+      "epoch": 0.208864,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.975322580645161e-05,
+      "loss": 0.1433,
+      "step": 13054
+    },
+    {
+      "epoch": 0.20888,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.975161290322581e-05,
+      "loss": 0.1689,
+      "step": 13055
+    },
+    {
+      "epoch": 0.208896,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.975e-05,
+      "loss": 0.1911,
+      "step": 13056
+    },
+    {
+      "epoch": 0.208912,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.97483870967742e-05,
+      "loss": 0.1554,
+      "step": 13057
+    },
+    {
+      "epoch": 0.208928,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.974677419354839e-05,
+      "loss": 0.1954,
+      "step": 13058
+    },
+    {
+      "epoch": 0.208944,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.974516129032259e-05,
+      "loss": 0.1977,
+      "step": 13059
+    },
+    {
+      "epoch": 0.20896,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.974354838709677e-05,
+      "loss": 0.1343,
+      "step": 13060
+    },
+    {
+      "epoch": 0.208976,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.974193548387097e-05,
+      "loss": 0.1578,
+      "step": 13061
+    },
+    {
+      "epoch": 0.208992,
+      "grad_norm": 0.49609375,
+      "learning_rate": 7.974032258064517e-05,
+      "loss": 0.1525,
+      "step": 13062
+    },
+    {
+      "epoch": 0.209008,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.973870967741936e-05,
+      "loss": 0.1603,
+      "step": 13063
+    },
+    {
+      "epoch": 0.209024,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.973709677419356e-05,
+      "loss": 0.1884,
+      "step": 13064
+    },
+    {
+      "epoch": 0.20904,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.973548387096774e-05,
+      "loss": 0.1629,
+      "step": 13065
+    },
+    {
+      "epoch": 0.209056,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.973387096774194e-05,
+      "loss": 0.1964,
+      "step": 13066
+    },
+    {
+      "epoch": 0.209072,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.973225806451613e-05,
+      "loss": 0.099,
+      "step": 13067
+    },
+    {
+      "epoch": 0.209088,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.973064516129033e-05,
+      "loss": 0.1381,
+      "step": 13068
+    },
+    {
+      "epoch": 0.209104,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.972903225806451e-05,
+      "loss": 0.1772,
+      "step": 13069
+    },
+    {
+      "epoch": 0.20912,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.972741935483871e-05,
+      "loss": 0.1869,
+      "step": 13070
+    },
+    {
+      "epoch": 0.209136,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.97258064516129e-05,
+      "loss": 0.1726,
+      "step": 13071
+    },
+    {
+      "epoch": 0.209152,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.97241935483871e-05,
+      "loss": 0.1511,
+      "step": 13072
+    },
+    {
+      "epoch": 0.209168,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.97225806451613e-05,
+      "loss": 0.216,
+      "step": 13073
+    },
+    {
+      "epoch": 0.209184,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.97209677419355e-05,
+      "loss": 0.2015,
+      "step": 13074
+    },
+    {
+      "epoch": 0.2092,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.971935483870969e-05,
+      "loss": 0.1858,
+      "step": 13075
+    },
+    {
+      "epoch": 0.209216,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.971774193548387e-05,
+      "loss": 0.202,
+      "step": 13076
+    },
+    {
+      "epoch": 0.209232,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.971612903225807e-05,
+      "loss": 0.1623,
+      "step": 13077
+    },
+    {
+      "epoch": 0.209248,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.971451612903226e-05,
+      "loss": 0.1699,
+      "step": 13078
+    },
+    {
+      "epoch": 0.209264,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.971290322580646e-05,
+      "loss": 0.1914,
+      "step": 13079
+    },
+    {
+      "epoch": 0.20928,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.971129032258064e-05,
+      "loss": 0.1789,
+      "step": 13080
+    },
+    {
+      "epoch": 0.209296,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.970967741935484e-05,
+      "loss": 0.1155,
+      "step": 13081
+    },
+    {
+      "epoch": 0.209312,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.970806451612903e-05,
+      "loss": 0.1665,
+      "step": 13082
+    },
+    {
+      "epoch": 0.209328,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.970645161290323e-05,
+      "loss": 0.1455,
+      "step": 13083
+    },
+    {
+      "epoch": 0.209344,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.970483870967743e-05,
+      "loss": 0.1746,
+      "step": 13084
+    },
+    {
+      "epoch": 0.20936,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.970322580645161e-05,
+      "loss": 0.1901,
+      "step": 13085
+    },
+    {
+      "epoch": 0.209376,
+      "grad_norm": 1.125,
+      "learning_rate": 7.970161290322581e-05,
+      "loss": 0.1664,
+      "step": 13086
+    },
+    {
+      "epoch": 0.209392,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.970000000000001e-05,
+      "loss": 0.1581,
+      "step": 13087
+    },
+    {
+      "epoch": 0.209408,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.96983870967742e-05,
+      "loss": 0.1736,
+      "step": 13088
+    },
+    {
+      "epoch": 0.209424,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.96967741935484e-05,
+      "loss": 0.1931,
+      "step": 13089
+    },
+    {
+      "epoch": 0.20944,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.969516129032258e-05,
+      "loss": 0.1573,
+      "step": 13090
+    },
+    {
+      "epoch": 0.209456,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.969354838709677e-05,
+      "loss": 0.1406,
+      "step": 13091
+    },
+    {
+      "epoch": 0.209472,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.969193548387097e-05,
+      "loss": 0.1547,
+      "step": 13092
+    },
+    {
+      "epoch": 0.209488,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.969032258064516e-05,
+      "loss": 0.1758,
+      "step": 13093
+    },
+    {
+      "epoch": 0.209504,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.968870967741936e-05,
+      "loss": 0.1958,
+      "step": 13094
+    },
+    {
+      "epoch": 0.20952,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.968709677419354e-05,
+      "loss": 0.1832,
+      "step": 13095
+    },
+    {
+      "epoch": 0.209536,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.968548387096774e-05,
+      "loss": 0.1684,
+      "step": 13096
+    },
+    {
+      "epoch": 0.209552,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.968387096774194e-05,
+      "loss": 0.2013,
+      "step": 13097
+    },
+    {
+      "epoch": 0.209568,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.968225806451614e-05,
+      "loss": 0.1528,
+      "step": 13098
+    },
+    {
+      "epoch": 0.209584,
+      "grad_norm": 1.7265625,
+      "learning_rate": 7.968064516129033e-05,
+      "loss": 0.1842,
+      "step": 13099
+    },
+    {
+      "epoch": 0.2096,
+      "grad_norm": 1.2109375,
+      "learning_rate": 7.967903225806453e-05,
+      "loss": 0.189,
+      "step": 13100
+    },
+    {
+      "epoch": 0.209616,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.967741935483871e-05,
+      "loss": 0.1539,
+      "step": 13101
+    },
+    {
+      "epoch": 0.209632,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.967580645161291e-05,
+      "loss": 0.1489,
+      "step": 13102
+    },
+    {
+      "epoch": 0.209648,
+      "grad_norm": 1.3515625,
+      "learning_rate": 7.96741935483871e-05,
+      "loss": 0.1464,
+      "step": 13103
+    },
+    {
+      "epoch": 0.209664,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.96725806451613e-05,
+      "loss": 0.1344,
+      "step": 13104
+    },
+    {
+      "epoch": 0.20968,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.967096774193548e-05,
+      "loss": 0.1804,
+      "step": 13105
+    },
+    {
+      "epoch": 0.209696,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.966935483870968e-05,
+      "loss": 0.2109,
+      "step": 13106
+    },
+    {
+      "epoch": 0.209712,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.966774193548387e-05,
+      "loss": 0.1994,
+      "step": 13107
+    },
+    {
+      "epoch": 0.209728,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.966612903225807e-05,
+      "loss": 0.1702,
+      "step": 13108
+    },
+    {
+      "epoch": 0.209744,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.966451612903227e-05,
+      "loss": 0.1298,
+      "step": 13109
+    },
+    {
+      "epoch": 0.20976,
+      "grad_norm": 0.75,
+      "learning_rate": 7.966290322580646e-05,
+      "loss": 0.1749,
+      "step": 13110
+    },
+    {
+      "epoch": 0.209776,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.966129032258065e-05,
+      "loss": 0.1321,
+      "step": 13111
+    },
+    {
+      "epoch": 0.209792,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.965967741935484e-05,
+      "loss": 0.1478,
+      "step": 13112
+    },
+    {
+      "epoch": 0.209808,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.965806451612904e-05,
+      "loss": 0.1789,
+      "step": 13113
+    },
+    {
+      "epoch": 0.209824,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.965645161290323e-05,
+      "loss": 0.1497,
+      "step": 13114
+    },
+    {
+      "epoch": 0.20984,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.965483870967743e-05,
+      "loss": 0.187,
+      "step": 13115
+    },
+    {
+      "epoch": 0.209856,
+      "grad_norm": 0.625,
+      "learning_rate": 7.965322580645161e-05,
+      "loss": 0.1732,
+      "step": 13116
+    },
+    {
+      "epoch": 0.209872,
+      "grad_norm": 1.3125,
+      "learning_rate": 7.965161290322581e-05,
+      "loss": 0.1956,
+      "step": 13117
+    },
+    {
+      "epoch": 0.209888,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.965e-05,
+      "loss": 0.1881,
+      "step": 13118
+    },
+    {
+      "epoch": 0.209904,
+      "grad_norm": 1.515625,
+      "learning_rate": 7.96483870967742e-05,
+      "loss": 0.1679,
+      "step": 13119
+    },
+    {
+      "epoch": 0.20992,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.964677419354838e-05,
+      "loss": 0.1764,
+      "step": 13120
+    },
+    {
+      "epoch": 0.209936,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.964516129032258e-05,
+      "loss": 0.1909,
+      "step": 13121
+    },
+    {
+      "epoch": 0.209952,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.964354838709678e-05,
+      "loss": 0.1826,
+      "step": 13122
+    },
+    {
+      "epoch": 0.209968,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.964193548387097e-05,
+      "loss": 0.1773,
+      "step": 13123
+    },
+    {
+      "epoch": 0.209984,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.964032258064517e-05,
+      "loss": 0.1723,
+      "step": 13124
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.963870967741935e-05,
+      "loss": 0.2104,
+      "step": 13125
+    },
+    {
+      "epoch": 0.210016,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.963709677419355e-05,
+      "loss": 0.1399,
+      "step": 13126
+    },
+    {
+      "epoch": 0.210032,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.963548387096774e-05,
+      "loss": 0.1155,
+      "step": 13127
+    },
+    {
+      "epoch": 0.210048,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.963387096774194e-05,
+      "loss": 0.1921,
+      "step": 13128
+    },
+    {
+      "epoch": 0.210064,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.963225806451613e-05,
+      "loss": 0.1817,
+      "step": 13129
+    },
+    {
+      "epoch": 0.21008,
+      "grad_norm": 0.75,
+      "learning_rate": 7.963064516129033e-05,
+      "loss": 0.1821,
+      "step": 13130
+    },
+    {
+      "epoch": 0.210096,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.962903225806451e-05,
+      "loss": 0.1283,
+      "step": 13131
+    },
+    {
+      "epoch": 0.210112,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.962741935483871e-05,
+      "loss": 0.1528,
+      "step": 13132
+    },
+    {
+      "epoch": 0.210128,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.962580645161291e-05,
+      "loss": 0.1525,
+      "step": 13133
+    },
+    {
+      "epoch": 0.210144,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.962419354838711e-05,
+      "loss": 0.1798,
+      "step": 13134
+    },
+    {
+      "epoch": 0.21016,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.96225806451613e-05,
+      "loss": 0.133,
+      "step": 13135
+    },
+    {
+      "epoch": 0.210176,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.96209677419355e-05,
+      "loss": 0.1964,
+      "step": 13136
+    },
+    {
+      "epoch": 0.210192,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.961935483870968e-05,
+      "loss": 0.1969,
+      "step": 13137
+    },
+    {
+      "epoch": 0.210208,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.961774193548387e-05,
+      "loss": 0.1823,
+      "step": 13138
+    },
+    {
+      "epoch": 0.210224,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.961612903225807e-05,
+      "loss": 0.1361,
+      "step": 13139
+    },
+    {
+      "epoch": 0.21024,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.961451612903225e-05,
+      "loss": 0.1792,
+      "step": 13140
+    },
+    {
+      "epoch": 0.210256,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.961290322580645e-05,
+      "loss": 0.1897,
+      "step": 13141
+    },
+    {
+      "epoch": 0.210272,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.961129032258064e-05,
+      "loss": 0.1519,
+      "step": 13142
+    },
+    {
+      "epoch": 0.210288,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.960967741935484e-05,
+      "loss": 0.1945,
+      "step": 13143
+    },
+    {
+      "epoch": 0.210304,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.960806451612904e-05,
+      "loss": 0.2006,
+      "step": 13144
+    },
+    {
+      "epoch": 0.21032,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.960645161290324e-05,
+      "loss": 0.1583,
+      "step": 13145
+    },
+    {
+      "epoch": 0.210336,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.960483870967743e-05,
+      "loss": 0.1745,
+      "step": 13146
+    },
+    {
+      "epoch": 0.210352,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.960322580645162e-05,
+      "loss": 0.1762,
+      "step": 13147
+    },
+    {
+      "epoch": 0.210368,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.960161290322581e-05,
+      "loss": 0.1566,
+      "step": 13148
+    },
+    {
+      "epoch": 0.210384,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.960000000000001e-05,
+      "loss": 0.1699,
+      "step": 13149
+    },
+    {
+      "epoch": 0.2104,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.95983870967742e-05,
+      "loss": 0.1567,
+      "step": 13150
+    },
+    {
+      "epoch": 0.210416,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.95967741935484e-05,
+      "loss": 0.1587,
+      "step": 13151
+    },
+    {
+      "epoch": 0.210432,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.959516129032258e-05,
+      "loss": 0.1808,
+      "step": 13152
+    },
+    {
+      "epoch": 0.210448,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.959354838709678e-05,
+      "loss": 0.1845,
+      "step": 13153
+    },
+    {
+      "epoch": 0.210464,
+      "grad_norm": 1.0,
+      "learning_rate": 7.959193548387097e-05,
+      "loss": 0.2037,
+      "step": 13154
+    },
+    {
+      "epoch": 0.21048,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.959032258064515e-05,
+      "loss": 0.1483,
+      "step": 13155
+    },
+    {
+      "epoch": 0.210496,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.958870967741935e-05,
+      "loss": 0.1727,
+      "step": 13156
+    },
+    {
+      "epoch": 0.210512,
+      "grad_norm": 1.296875,
+      "learning_rate": 7.958709677419355e-05,
+      "loss": 0.1741,
+      "step": 13157
+    },
+    {
+      "epoch": 0.210528,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.958548387096775e-05,
+      "loss": 0.1963,
+      "step": 13158
+    },
+    {
+      "epoch": 0.210544,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.958387096774194e-05,
+      "loss": 0.1597,
+      "step": 13159
+    },
+    {
+      "epoch": 0.21056,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.958225806451614e-05,
+      "loss": 0.1732,
+      "step": 13160
+    },
+    {
+      "epoch": 0.210576,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.958064516129032e-05,
+      "loss": 0.1627,
+      "step": 13161
+    },
+    {
+      "epoch": 0.210592,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.957903225806452e-05,
+      "loss": 0.1272,
+      "step": 13162
+    },
+    {
+      "epoch": 0.210608,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.957741935483871e-05,
+      "loss": 0.2051,
+      "step": 13163
+    },
+    {
+      "epoch": 0.210624,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.957580645161291e-05,
+      "loss": 0.1752,
+      "step": 13164
+    },
+    {
+      "epoch": 0.21064,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.95741935483871e-05,
+      "loss": 0.1837,
+      "step": 13165
+    },
+    {
+      "epoch": 0.210656,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.95725806451613e-05,
+      "loss": 0.1798,
+      "step": 13166
+    },
+    {
+      "epoch": 0.210672,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.957096774193548e-05,
+      "loss": 0.2077,
+      "step": 13167
+    },
+    {
+      "epoch": 0.210688,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.956935483870968e-05,
+      "loss": 0.1979,
+      "step": 13168
+    },
+    {
+      "epoch": 0.210704,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.956774193548388e-05,
+      "loss": 0.1769,
+      "step": 13169
+    },
+    {
+      "epoch": 0.21072,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.956612903225807e-05,
+      "loss": 0.2029,
+      "step": 13170
+    },
+    {
+      "epoch": 0.210736,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.956451612903227e-05,
+      "loss": 0.2014,
+      "step": 13171
+    },
+    {
+      "epoch": 0.210752,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.956290322580645e-05,
+      "loss": 0.1964,
+      "step": 13172
+    },
+    {
+      "epoch": 0.210768,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.956129032258065e-05,
+      "loss": 0.1637,
+      "step": 13173
+    },
+    {
+      "epoch": 0.210784,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.955967741935484e-05,
+      "loss": 0.1921,
+      "step": 13174
+    },
+    {
+      "epoch": 0.2108,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.955806451612904e-05,
+      "loss": 0.1632,
+      "step": 13175
+    },
+    {
+      "epoch": 0.210816,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.955645161290322e-05,
+      "loss": 0.1662,
+      "step": 13176
+    },
+    {
+      "epoch": 0.210832,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.955483870967742e-05,
+      "loss": 0.1703,
+      "step": 13177
+    },
+    {
+      "epoch": 0.210848,
+      "grad_norm": 1.390625,
+      "learning_rate": 7.955322580645161e-05,
+      "loss": 0.2124,
+      "step": 13178
+    },
+    {
+      "epoch": 0.210864,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.955161290322581e-05,
+      "loss": 0.1974,
+      "step": 13179
+    },
+    {
+      "epoch": 0.21088,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.955e-05,
+      "loss": 0.1612,
+      "step": 13180
+    },
+    {
+      "epoch": 0.210896,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.95483870967742e-05,
+      "loss": 0.1629,
+      "step": 13181
+    },
+    {
+      "epoch": 0.210912,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.95467741935484e-05,
+      "loss": 0.1834,
+      "step": 13182
+    },
+    {
+      "epoch": 0.210928,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.95451612903226e-05,
+      "loss": 0.1873,
+      "step": 13183
+    },
+    {
+      "epoch": 0.210944,
+      "grad_norm": 0.484375,
+      "learning_rate": 7.954354838709678e-05,
+      "loss": 0.145,
+      "step": 13184
+    },
+    {
+      "epoch": 0.21096,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.954193548387097e-05,
+      "loss": 0.173,
+      "step": 13185
+    },
+    {
+      "epoch": 0.210976,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.954032258064517e-05,
+      "loss": 0.1774,
+      "step": 13186
+    },
+    {
+      "epoch": 0.210992,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.953870967741935e-05,
+      "loss": 0.1819,
+      "step": 13187
+    },
+    {
+      "epoch": 0.211008,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.953709677419355e-05,
+      "loss": 0.1858,
+      "step": 13188
+    },
+    {
+      "epoch": 0.211024,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.953548387096774e-05,
+      "loss": 0.1356,
+      "step": 13189
+    },
+    {
+      "epoch": 0.21104,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.953387096774194e-05,
+      "loss": 0.1546,
+      "step": 13190
+    },
+    {
+      "epoch": 0.211056,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.953225806451612e-05,
+      "loss": 0.1984,
+      "step": 13191
+    },
+    {
+      "epoch": 0.211072,
+      "grad_norm": 1.828125,
+      "learning_rate": 7.953064516129032e-05,
+      "loss": 0.1994,
+      "step": 13192
+    },
+    {
+      "epoch": 0.211088,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.952903225806452e-05,
+      "loss": 0.1806,
+      "step": 13193
+    },
+    {
+      "epoch": 0.211104,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.952741935483872e-05,
+      "loss": 0.1759,
+      "step": 13194
+    },
+    {
+      "epoch": 0.21112,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.952580645161291e-05,
+      "loss": 0.1635,
+      "step": 13195
+    },
+    {
+      "epoch": 0.211136,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.952419354838711e-05,
+      "loss": 0.1914,
+      "step": 13196
+    },
+    {
+      "epoch": 0.211152,
+      "grad_norm": 1.8671875,
+      "learning_rate": 7.95225806451613e-05,
+      "loss": 0.1628,
+      "step": 13197
+    },
+    {
+      "epoch": 0.211168,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.95209677419355e-05,
+      "loss": 0.1872,
+      "step": 13198
+    },
+    {
+      "epoch": 0.211184,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.951935483870968e-05,
+      "loss": 0.2294,
+      "step": 13199
+    },
+    {
+      "epoch": 0.2112,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.951774193548387e-05,
+      "loss": 0.1791,
+      "step": 13200
+    },
+    {
+      "epoch": 0.211216,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.951612903225807e-05,
+      "loss": 0.172,
+      "step": 13201
+    },
+    {
+      "epoch": 0.211232,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.951451612903225e-05,
+      "loss": 0.1523,
+      "step": 13202
+    },
+    {
+      "epoch": 0.211248,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.951290322580645e-05,
+      "loss": 0.1803,
+      "step": 13203
+    },
+    {
+      "epoch": 0.211264,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.951129032258065e-05,
+      "loss": 0.1772,
+      "step": 13204
+    },
+    {
+      "epoch": 0.21128,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.950967741935485e-05,
+      "loss": 0.1771,
+      "step": 13205
+    },
+    {
+      "epoch": 0.211296,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.950806451612904e-05,
+      "loss": 0.1841,
+      "step": 13206
+    },
+    {
+      "epoch": 0.211312,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.950645161290324e-05,
+      "loss": 0.1673,
+      "step": 13207
+    },
+    {
+      "epoch": 0.211328,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.950483870967742e-05,
+      "loss": 0.1844,
+      "step": 13208
+    },
+    {
+      "epoch": 0.211344,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.950322580645162e-05,
+      "loss": 0.156,
+      "step": 13209
+    },
+    {
+      "epoch": 0.21136,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.950161290322581e-05,
+      "loss": 0.1433,
+      "step": 13210
+    },
+    {
+      "epoch": 0.211376,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.950000000000001e-05,
+      "loss": 0.1823,
+      "step": 13211
+    },
+    {
+      "epoch": 0.211392,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.94983870967742e-05,
+      "loss": 0.1895,
+      "step": 13212
+    },
+    {
+      "epoch": 0.211408,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.94967741935484e-05,
+      "loss": 0.1818,
+      "step": 13213
+    },
+    {
+      "epoch": 0.211424,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.949516129032258e-05,
+      "loss": 0.1799,
+      "step": 13214
+    },
+    {
+      "epoch": 0.21144,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.949354838709678e-05,
+      "loss": 0.1646,
+      "step": 13215
+    },
+    {
+      "epoch": 0.211456,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.949193548387097e-05,
+      "loss": 0.1571,
+      "step": 13216
+    },
+    {
+      "epoch": 0.211472,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.949032258064517e-05,
+      "loss": 0.1809,
+      "step": 13217
+    },
+    {
+      "epoch": 0.211488,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.948870967741936e-05,
+      "loss": 0.2,
+      "step": 13218
+    },
+    {
+      "epoch": 0.211504,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.948709677419355e-05,
+      "loss": 0.1431,
+      "step": 13219
+    },
+    {
+      "epoch": 0.21152,
+      "grad_norm": 1.125,
+      "learning_rate": 7.948548387096775e-05,
+      "loss": 0.2132,
+      "step": 13220
+    },
+    {
+      "epoch": 0.211536,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.948387096774194e-05,
+      "loss": 0.1775,
+      "step": 13221
+    },
+    {
+      "epoch": 0.211552,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.948225806451614e-05,
+      "loss": 0.1993,
+      "step": 13222
+    },
+    {
+      "epoch": 0.211568,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.948064516129032e-05,
+      "loss": 0.1794,
+      "step": 13223
+    },
+    {
+      "epoch": 0.211584,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.947903225806452e-05,
+      "loss": 0.1872,
+      "step": 13224
+    },
+    {
+      "epoch": 0.2116,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.947741935483871e-05,
+      "loss": 0.1294,
+      "step": 13225
+    },
+    {
+      "epoch": 0.211616,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.947580645161291e-05,
+      "loss": 0.1996,
+      "step": 13226
+    },
+    {
+      "epoch": 0.211632,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.94741935483871e-05,
+      "loss": 0.1745,
+      "step": 13227
+    },
+    {
+      "epoch": 0.211648,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.94725806451613e-05,
+      "loss": 0.1897,
+      "step": 13228
+    },
+    {
+      "epoch": 0.211664,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.947096774193549e-05,
+      "loss": 0.2069,
+      "step": 13229
+    },
+    {
+      "epoch": 0.21168,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.946935483870969e-05,
+      "loss": 0.1871,
+      "step": 13230
+    },
+    {
+      "epoch": 0.211696,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.946774193548388e-05,
+      "loss": 0.1706,
+      "step": 13231
+    },
+    {
+      "epoch": 0.211712,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.946612903225806e-05,
+      "loss": 0.1887,
+      "step": 13232
+    },
+    {
+      "epoch": 0.211728,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.946451612903226e-05,
+      "loss": 0.193,
+      "step": 13233
+    },
+    {
+      "epoch": 0.211744,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.946290322580645e-05,
+      "loss": 0.1487,
+      "step": 13234
+    },
+    {
+      "epoch": 0.21176,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.946129032258065e-05,
+      "loss": 0.1505,
+      "step": 13235
+    },
+    {
+      "epoch": 0.211776,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.945967741935484e-05,
+      "loss": 0.15,
+      "step": 13236
+    },
+    {
+      "epoch": 0.211792,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.945806451612904e-05,
+      "loss": 0.1557,
+      "step": 13237
+    },
+    {
+      "epoch": 0.211808,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.945645161290322e-05,
+      "loss": 0.1997,
+      "step": 13238
+    },
+    {
+      "epoch": 0.211824,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.945483870967742e-05,
+      "loss": 0.2028,
+      "step": 13239
+    },
+    {
+      "epoch": 0.21184,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.945322580645162e-05,
+      "loss": 0.2035,
+      "step": 13240
+    },
+    {
+      "epoch": 0.211856,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.945161290322581e-05,
+      "loss": 0.2054,
+      "step": 13241
+    },
+    {
+      "epoch": 0.211872,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.945e-05,
+      "loss": 0.2268,
+      "step": 13242
+    },
+    {
+      "epoch": 0.211888,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.94483870967742e-05,
+      "loss": 0.1893,
+      "step": 13243
+    },
+    {
+      "epoch": 0.211904,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.944677419354839e-05,
+      "loss": 0.1865,
+      "step": 13244
+    },
+    {
+      "epoch": 0.21192,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.944516129032259e-05,
+      "loss": 0.1638,
+      "step": 13245
+    },
+    {
+      "epoch": 0.211936,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.944354838709678e-05,
+      "loss": 0.1513,
+      "step": 13246
+    },
+    {
+      "epoch": 0.211952,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.944193548387096e-05,
+      "loss": 0.1896,
+      "step": 13247
+    },
+    {
+      "epoch": 0.211968,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.944032258064516e-05,
+      "loss": 0.186,
+      "step": 13248
+    },
+    {
+      "epoch": 0.211984,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.943870967741935e-05,
+      "loss": 0.1495,
+      "step": 13249
+    },
+    {
+      "epoch": 0.212,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.943709677419355e-05,
+      "loss": 0.1642,
+      "step": 13250
+    },
+    {
+      "epoch": 0.212016,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.943548387096774e-05,
+      "loss": 0.167,
+      "step": 13251
+    },
+    {
+      "epoch": 0.212032,
+      "grad_norm": 1.2109375,
+      "learning_rate": 7.943387096774194e-05,
+      "loss": 0.1513,
+      "step": 13252
+    },
+    {
+      "epoch": 0.212048,
+      "grad_norm": 0.5,
+      "learning_rate": 7.943225806451613e-05,
+      "loss": 0.143,
+      "step": 13253
+    },
+    {
+      "epoch": 0.212064,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.943064516129033e-05,
+      "loss": 0.1622,
+      "step": 13254
+    },
+    {
+      "epoch": 0.21208,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.942903225806452e-05,
+      "loss": 0.1689,
+      "step": 13255
+    },
+    {
+      "epoch": 0.212096,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.942741935483872e-05,
+      "loss": 0.1865,
+      "step": 13256
+    },
+    {
+      "epoch": 0.212112,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.94258064516129e-05,
+      "loss": 0.1813,
+      "step": 13257
+    },
+    {
+      "epoch": 0.212128,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.94241935483871e-05,
+      "loss": 0.1563,
+      "step": 13258
+    },
+    {
+      "epoch": 0.212144,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.942258064516129e-05,
+      "loss": 0.163,
+      "step": 13259
+    },
+    {
+      "epoch": 0.21216,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.942096774193549e-05,
+      "loss": 0.1811,
+      "step": 13260
+    },
+    {
+      "epoch": 0.212176,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.941935483870968e-05,
+      "loss": 0.2008,
+      "step": 13261
+    },
+    {
+      "epoch": 0.212192,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.941774193548388e-05,
+      "loss": 0.1867,
+      "step": 13262
+    },
+    {
+      "epoch": 0.212208,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.941612903225806e-05,
+      "loss": 0.1711,
+      "step": 13263
+    },
+    {
+      "epoch": 0.212224,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.941451612903226e-05,
+      "loss": 0.174,
+      "step": 13264
+    },
+    {
+      "epoch": 0.21224,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.941290322580646e-05,
+      "loss": 0.1746,
+      "step": 13265
+    },
+    {
+      "epoch": 0.212256,
+      "grad_norm": 0.875,
+      "learning_rate": 7.941129032258065e-05,
+      "loss": 0.1232,
+      "step": 13266
+    },
+    {
+      "epoch": 0.212272,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.940967741935485e-05,
+      "loss": 0.1489,
+      "step": 13267
+    },
+    {
+      "epoch": 0.212288,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.940806451612903e-05,
+      "loss": 0.1389,
+      "step": 13268
+    },
+    {
+      "epoch": 0.212304,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.940645161290323e-05,
+      "loss": 0.1766,
+      "step": 13269
+    },
+    {
+      "epoch": 0.21232,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.940483870967742e-05,
+      "loss": 0.2064,
+      "step": 13270
+    },
+    {
+      "epoch": 0.212336,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.940322580645162e-05,
+      "loss": 0.1771,
+      "step": 13271
+    },
+    {
+      "epoch": 0.212352,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.94016129032258e-05,
+      "loss": 0.1619,
+      "step": 13272
+    },
+    {
+      "epoch": 0.212368,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.94e-05,
+      "loss": 0.1905,
+      "step": 13273
+    },
+    {
+      "epoch": 0.212384,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.939838709677419e-05,
+      "loss": 0.1648,
+      "step": 13274
+    },
+    {
+      "epoch": 0.2124,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.939677419354839e-05,
+      "loss": 0.1385,
+      "step": 13275
+    },
+    {
+      "epoch": 0.212416,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.939516129032258e-05,
+      "loss": 0.1689,
+      "step": 13276
+    },
+    {
+      "epoch": 0.212432,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.939354838709678e-05,
+      "loss": 0.1587,
+      "step": 13277
+    },
+    {
+      "epoch": 0.212448,
+      "grad_norm": 1.8984375,
+      "learning_rate": 7.939193548387098e-05,
+      "loss": 0.1995,
+      "step": 13278
+    },
+    {
+      "epoch": 0.212464,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.939032258064516e-05,
+      "loss": 0.1387,
+      "step": 13279
+    },
+    {
+      "epoch": 0.21248,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.938870967741936e-05,
+      "loss": 0.1539,
+      "step": 13280
+    },
+    {
+      "epoch": 0.212496,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.938709677419355e-05,
+      "loss": 0.1641,
+      "step": 13281
+    },
+    {
+      "epoch": 0.212512,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.938548387096775e-05,
+      "loss": 0.1609,
+      "step": 13282
+    },
+    {
+      "epoch": 0.212528,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.938387096774193e-05,
+      "loss": 0.1687,
+      "step": 13283
+    },
+    {
+      "epoch": 0.212544,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.938225806451613e-05,
+      "loss": 0.1752,
+      "step": 13284
+    },
+    {
+      "epoch": 0.21256,
+      "grad_norm": 1.375,
+      "learning_rate": 7.938064516129032e-05,
+      "loss": 0.2033,
+      "step": 13285
+    },
+    {
+      "epoch": 0.212576,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.937903225806452e-05,
+      "loss": 0.1761,
+      "step": 13286
+    },
+    {
+      "epoch": 0.212592,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.93774193548387e-05,
+      "loss": 0.1207,
+      "step": 13287
+    },
+    {
+      "epoch": 0.212608,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.93758064516129e-05,
+      "loss": 0.1766,
+      "step": 13288
+    },
+    {
+      "epoch": 0.212624,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.93741935483871e-05,
+      "loss": 0.1707,
+      "step": 13289
+    },
+    {
+      "epoch": 0.21264,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.93725806451613e-05,
+      "loss": 0.1621,
+      "step": 13290
+    },
+    {
+      "epoch": 0.212656,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.937096774193549e-05,
+      "loss": 0.1917,
+      "step": 13291
+    },
+    {
+      "epoch": 0.212672,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.936935483870969e-05,
+      "loss": 0.1568,
+      "step": 13292
+    },
+    {
+      "epoch": 0.212688,
+      "grad_norm": 1.28125,
+      "learning_rate": 7.936774193548388e-05,
+      "loss": 0.1773,
+      "step": 13293
+    },
+    {
+      "epoch": 0.212704,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.936612903225806e-05,
+      "loss": 0.213,
+      "step": 13294
+    },
+    {
+      "epoch": 0.21272,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.936451612903226e-05,
+      "loss": 0.1526,
+      "step": 13295
+    },
+    {
+      "epoch": 0.212736,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.936290322580645e-05,
+      "loss": 0.1731,
+      "step": 13296
+    },
+    {
+      "epoch": 0.212752,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.936129032258065e-05,
+      "loss": 0.1653,
+      "step": 13297
+    },
+    {
+      "epoch": 0.212768,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.935967741935483e-05,
+      "loss": 0.1828,
+      "step": 13298
+    },
+    {
+      "epoch": 0.212784,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.935806451612903e-05,
+      "loss": 0.1208,
+      "step": 13299
+    },
+    {
+      "epoch": 0.2128,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.935645161290323e-05,
+      "loss": 0.2089,
+      "step": 13300
+    },
+    {
+      "epoch": 0.212816,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.935483870967743e-05,
+      "loss": 0.1711,
+      "step": 13301
+    },
+    {
+      "epoch": 0.212832,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.935322580645162e-05,
+      "loss": 0.1865,
+      "step": 13302
+    },
+    {
+      "epoch": 0.212848,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.935161290322582e-05,
+      "loss": 0.162,
+      "step": 13303
+    },
+    {
+      "epoch": 0.212864,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.935e-05,
+      "loss": 0.1846,
+      "step": 13304
+    },
+    {
+      "epoch": 0.21288,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.93483870967742e-05,
+      "loss": 0.1527,
+      "step": 13305
+    },
+    {
+      "epoch": 0.212896,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.934677419354839e-05,
+      "loss": 0.1516,
+      "step": 13306
+    },
+    {
+      "epoch": 0.212912,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.934516129032259e-05,
+      "loss": 0.191,
+      "step": 13307
+    },
+    {
+      "epoch": 0.212928,
+      "grad_norm": 1.359375,
+      "learning_rate": 7.934354838709678e-05,
+      "loss": 0.1671,
+      "step": 13308
+    },
+    {
+      "epoch": 0.212944,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.934193548387096e-05,
+      "loss": 0.1799,
+      "step": 13309
+    },
+    {
+      "epoch": 0.21296,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.934032258064516e-05,
+      "loss": 0.1702,
+      "step": 13310
+    },
+    {
+      "epoch": 0.212976,
+      "grad_norm": 0.455078125,
+      "learning_rate": 7.933870967741935e-05,
+      "loss": 0.157,
+      "step": 13311
+    },
+    {
+      "epoch": 0.212992,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.933709677419355e-05,
+      "loss": 0.1781,
+      "step": 13312
+    },
+    {
+      "epoch": 0.213008,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.933548387096775e-05,
+      "loss": 0.1649,
+      "step": 13313
+    },
+    {
+      "epoch": 0.213024,
+      "grad_norm": 0.75,
+      "learning_rate": 7.933387096774195e-05,
+      "loss": 0.1717,
+      "step": 13314
+    },
+    {
+      "epoch": 0.21304,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.933225806451613e-05,
+      "loss": 0.1846,
+      "step": 13315
+    },
+    {
+      "epoch": 0.213056,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.933064516129033e-05,
+      "loss": 0.166,
+      "step": 13316
+    },
+    {
+      "epoch": 0.213072,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.932903225806452e-05,
+      "loss": 0.1938,
+      "step": 13317
+    },
+    {
+      "epoch": 0.213088,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.932741935483872e-05,
+      "loss": 0.1393,
+      "step": 13318
+    },
+    {
+      "epoch": 0.213104,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.93258064516129e-05,
+      "loss": 0.1526,
+      "step": 13319
+    },
+    {
+      "epoch": 0.21312,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.93241935483871e-05,
+      "loss": 0.1745,
+      "step": 13320
+    },
+    {
+      "epoch": 0.213136,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.932258064516129e-05,
+      "loss": 0.1522,
+      "step": 13321
+    },
+    {
+      "epoch": 0.213152,
+      "grad_norm": 0.4453125,
+      "learning_rate": 7.932096774193549e-05,
+      "loss": 0.1356,
+      "step": 13322
+    },
+    {
+      "epoch": 0.213168,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.931935483870968e-05,
+      "loss": 0.1661,
+      "step": 13323
+    },
+    {
+      "epoch": 0.213184,
+      "grad_norm": 0.625,
+      "learning_rate": 7.931774193548387e-05,
+      "loss": 0.1496,
+      "step": 13324
+    },
+    {
+      "epoch": 0.2132,
+      "grad_norm": 0.625,
+      "learning_rate": 7.931612903225807e-05,
+      "loss": 0.1985,
+      "step": 13325
+    },
+    {
+      "epoch": 0.213216,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.931451612903226e-05,
+      "loss": 0.1506,
+      "step": 13326
+    },
+    {
+      "epoch": 0.213232,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.931290322580646e-05,
+      "loss": 0.1644,
+      "step": 13327
+    },
+    {
+      "epoch": 0.213248,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.931129032258065e-05,
+      "loss": 0.2306,
+      "step": 13328
+    },
+    {
+      "epoch": 0.213264,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.930967741935485e-05,
+      "loss": 0.1556,
+      "step": 13329
+    },
+    {
+      "epoch": 0.21328,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.930806451612903e-05,
+      "loss": 0.1765,
+      "step": 13330
+    },
+    {
+      "epoch": 0.213296,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.930645161290323e-05,
+      "loss": 0.1735,
+      "step": 13331
+    },
+    {
+      "epoch": 0.213312,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.930483870967742e-05,
+      "loss": 0.1692,
+      "step": 13332
+    },
+    {
+      "epoch": 0.213328,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.930322580645162e-05,
+      "loss": 0.148,
+      "step": 13333
+    },
+    {
+      "epoch": 0.213344,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.93016129032258e-05,
+      "loss": 0.1991,
+      "step": 13334
+    },
+    {
+      "epoch": 0.21336,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.93e-05,
+      "loss": 0.1538,
+      "step": 13335
+    },
+    {
+      "epoch": 0.213376,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.92983870967742e-05,
+      "loss": 0.18,
+      "step": 13336
+    },
+    {
+      "epoch": 0.213392,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.929677419354839e-05,
+      "loss": 0.171,
+      "step": 13337
+    },
+    {
+      "epoch": 0.213408,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.929516129032259e-05,
+      "loss": 0.1835,
+      "step": 13338
+    },
+    {
+      "epoch": 0.213424,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.929354838709679e-05,
+      "loss": 0.1761,
+      "step": 13339
+    },
+    {
+      "epoch": 0.21344,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.929193548387097e-05,
+      "loss": 0.2053,
+      "step": 13340
+    },
+    {
+      "epoch": 0.213456,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.929032258064516e-05,
+      "loss": 0.1833,
+      "step": 13341
+    },
+    {
+      "epoch": 0.213472,
+      "grad_norm": 0.478515625,
+      "learning_rate": 7.928870967741936e-05,
+      "loss": 0.1446,
+      "step": 13342
+    },
+    {
+      "epoch": 0.213488,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.928709677419355e-05,
+      "loss": 0.1774,
+      "step": 13343
+    },
+    {
+      "epoch": 0.213504,
+      "grad_norm": 1.3671875,
+      "learning_rate": 7.928548387096775e-05,
+      "loss": 0.1865,
+      "step": 13344
+    },
+    {
+      "epoch": 0.21352,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.928387096774193e-05,
+      "loss": 0.2035,
+      "step": 13345
+    },
+    {
+      "epoch": 0.213536,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.928225806451613e-05,
+      "loss": 0.2082,
+      "step": 13346
+    },
+    {
+      "epoch": 0.213552,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.928064516129032e-05,
+      "loss": 0.1863,
+      "step": 13347
+    },
+    {
+      "epoch": 0.213568,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.927903225806452e-05,
+      "loss": 0.1592,
+      "step": 13348
+    },
+    {
+      "epoch": 0.213584,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.927741935483872e-05,
+      "loss": 0.1514,
+      "step": 13349
+    },
+    {
+      "epoch": 0.2136,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.927580645161292e-05,
+      "loss": 0.1354,
+      "step": 13350
+    },
+    {
+      "epoch": 0.213616,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.92741935483871e-05,
+      "loss": 0.1381,
+      "step": 13351
+    },
+    {
+      "epoch": 0.213632,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.92725806451613e-05,
+      "loss": 0.1987,
+      "step": 13352
+    },
+    {
+      "epoch": 0.213648,
+      "grad_norm": 1.4375,
+      "learning_rate": 7.927096774193549e-05,
+      "loss": 0.1847,
+      "step": 13353
+    },
+    {
+      "epoch": 0.213664,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.926935483870969e-05,
+      "loss": 0.1622,
+      "step": 13354
+    },
+    {
+      "epoch": 0.21368,
+      "grad_norm": 0.46875,
+      "learning_rate": 7.926774193548387e-05,
+      "loss": 0.1468,
+      "step": 13355
+    },
+    {
+      "epoch": 0.213696,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.926612903225806e-05,
+      "loss": 0.1705,
+      "step": 13356
+    },
+    {
+      "epoch": 0.213712,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.926451612903226e-05,
+      "loss": 0.1654,
+      "step": 13357
+    },
+    {
+      "epoch": 0.213728,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.926290322580645e-05,
+      "loss": 0.1719,
+      "step": 13358
+    },
+    {
+      "epoch": 0.213744,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.926129032258065e-05,
+      "loss": 0.2143,
+      "step": 13359
+    },
+    {
+      "epoch": 0.21376,
+      "grad_norm": 0.75,
+      "learning_rate": 7.925967741935484e-05,
+      "loss": 0.1515,
+      "step": 13360
+    },
+    {
+      "epoch": 0.213776,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.925806451612904e-05,
+      "loss": 0.1879,
+      "step": 13361
+    },
+    {
+      "epoch": 0.213792,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.925645161290323e-05,
+      "loss": 0.2108,
+      "step": 13362
+    },
+    {
+      "epoch": 0.213808,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.925483870967743e-05,
+      "loss": 0.1556,
+      "step": 13363
+    },
+    {
+      "epoch": 0.213824,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.925322580645162e-05,
+      "loss": 0.2035,
+      "step": 13364
+    },
+    {
+      "epoch": 0.21384,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.925161290322582e-05,
+      "loss": 0.1645,
+      "step": 13365
+    },
+    {
+      "epoch": 0.213856,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.925e-05,
+      "loss": 0.1953,
+      "step": 13366
+    },
+    {
+      "epoch": 0.213872,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.92483870967742e-05,
+      "loss": 0.1907,
+      "step": 13367
+    },
+    {
+      "epoch": 0.213888,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.924677419354839e-05,
+      "loss": 0.133,
+      "step": 13368
+    },
+    {
+      "epoch": 0.213904,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.924516129032259e-05,
+      "loss": 0.2055,
+      "step": 13369
+    },
+    {
+      "epoch": 0.21392,
+      "grad_norm": 1.2734375,
+      "learning_rate": 7.924354838709677e-05,
+      "loss": 0.2235,
+      "step": 13370
+    },
+    {
+      "epoch": 0.213936,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.924193548387096e-05,
+      "loss": 0.168,
+      "step": 13371
+    },
+    {
+      "epoch": 0.213952,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.924032258064516e-05,
+      "loss": 0.1434,
+      "step": 13372
+    },
+    {
+      "epoch": 0.213968,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.923870967741936e-05,
+      "loss": 0.1699,
+      "step": 13373
+    },
+    {
+      "epoch": 0.213984,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.923709677419356e-05,
+      "loss": 0.1471,
+      "step": 13374
+    },
+    {
+      "epoch": 0.214,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.923548387096774e-05,
+      "loss": 0.1703,
+      "step": 13375
+    },
+    {
+      "epoch": 0.214016,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.923387096774194e-05,
+      "loss": 0.1437,
+      "step": 13376
+    },
+    {
+      "epoch": 0.214032,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.923225806451613e-05,
+      "loss": 0.1828,
+      "step": 13377
+    },
+    {
+      "epoch": 0.214048,
+      "grad_norm": 1.0,
+      "learning_rate": 7.923064516129033e-05,
+      "loss": 0.1459,
+      "step": 13378
+    },
+    {
+      "epoch": 0.214064,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.922903225806452e-05,
+      "loss": 0.1617,
+      "step": 13379
+    },
+    {
+      "epoch": 0.21408,
+      "grad_norm": 1.265625,
+      "learning_rate": 7.922741935483872e-05,
+      "loss": 0.2078,
+      "step": 13380
+    },
+    {
+      "epoch": 0.214096,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.92258064516129e-05,
+      "loss": 0.1313,
+      "step": 13381
+    },
+    {
+      "epoch": 0.214112,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.92241935483871e-05,
+      "loss": 0.1697,
+      "step": 13382
+    },
+    {
+      "epoch": 0.214128,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.922258064516129e-05,
+      "loss": 0.1648,
+      "step": 13383
+    },
+    {
+      "epoch": 0.214144,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.922096774193549e-05,
+      "loss": 0.1723,
+      "step": 13384
+    },
+    {
+      "epoch": 0.21416,
+      "grad_norm": 0.40234375,
+      "learning_rate": 7.921935483870969e-05,
+      "loss": 0.1109,
+      "step": 13385
+    },
+    {
+      "epoch": 0.214176,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.921774193548389e-05,
+      "loss": 0.1526,
+      "step": 13386
+    },
+    {
+      "epoch": 0.214192,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.921612903225807e-05,
+      "loss": 0.1871,
+      "step": 13387
+    },
+    {
+      "epoch": 0.214208,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.921451612903226e-05,
+      "loss": 0.183,
+      "step": 13388
+    },
+    {
+      "epoch": 0.214224,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.921290322580646e-05,
+      "loss": 0.1616,
+      "step": 13389
+    },
+    {
+      "epoch": 0.21424,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.921129032258064e-05,
+      "loss": 0.1917,
+      "step": 13390
+    },
+    {
+      "epoch": 0.214256,
+      "grad_norm": 1.40625,
+      "learning_rate": 7.920967741935484e-05,
+      "loss": 0.2198,
+      "step": 13391
+    },
+    {
+      "epoch": 0.214272,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.920806451612903e-05,
+      "loss": 0.2201,
+      "step": 13392
+    },
+    {
+      "epoch": 0.214288,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.920645161290323e-05,
+      "loss": 0.1789,
+      "step": 13393
+    },
+    {
+      "epoch": 0.214304,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.920483870967742e-05,
+      "loss": 0.2104,
+      "step": 13394
+    },
+    {
+      "epoch": 0.21432,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.920322580645161e-05,
+      "loss": 0.1793,
+      "step": 13395
+    },
+    {
+      "epoch": 0.214336,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.920161290322581e-05,
+      "loss": 0.1796,
+      "step": 13396
+    },
+    {
+      "epoch": 0.214352,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.920000000000001e-05,
+      "loss": 0.1984,
+      "step": 13397
+    },
+    {
+      "epoch": 0.214368,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.91983870967742e-05,
+      "loss": 0.1923,
+      "step": 13398
+    },
+    {
+      "epoch": 0.214384,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.91967741935484e-05,
+      "loss": 0.1757,
+      "step": 13399
+    },
+    {
+      "epoch": 0.2144,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.919516129032259e-05,
+      "loss": 0.1935,
+      "step": 13400
+    },
+    {
+      "epoch": 0.214416,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.919354838709679e-05,
+      "loss": 0.1763,
+      "step": 13401
+    },
+    {
+      "epoch": 0.214432,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.919193548387097e-05,
+      "loss": 0.1468,
+      "step": 13402
+    },
+    {
+      "epoch": 0.214448,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.919032258064516e-05,
+      "loss": 0.1846,
+      "step": 13403
+    },
+    {
+      "epoch": 0.214464,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.918870967741936e-05,
+      "loss": 0.1671,
+      "step": 13404
+    },
+    {
+      "epoch": 0.21448,
+      "grad_norm": 1.3359375,
+      "learning_rate": 7.918709677419354e-05,
+      "loss": 0.2024,
+      "step": 13405
+    },
+    {
+      "epoch": 0.214496,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.918548387096774e-05,
+      "loss": 0.1717,
+      "step": 13406
+    },
+    {
+      "epoch": 0.214512,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.918387096774193e-05,
+      "loss": 0.1726,
+      "step": 13407
+    },
+    {
+      "epoch": 0.214528,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.918225806451613e-05,
+      "loss": 0.1553,
+      "step": 13408
+    },
+    {
+      "epoch": 0.214544,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.918064516129033e-05,
+      "loss": 0.1663,
+      "step": 13409
+    },
+    {
+      "epoch": 0.21456,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.917903225806453e-05,
+      "loss": 0.2375,
+      "step": 13410
+    },
+    {
+      "epoch": 0.214576,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.917741935483871e-05,
+      "loss": 0.1613,
+      "step": 13411
+    },
+    {
+      "epoch": 0.214592,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.917580645161291e-05,
+      "loss": 0.1584,
+      "step": 13412
+    },
+    {
+      "epoch": 0.214608,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.91741935483871e-05,
+      "loss": 0.2067,
+      "step": 13413
+    },
+    {
+      "epoch": 0.214624,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.91725806451613e-05,
+      "loss": 0.1726,
+      "step": 13414
+    },
+    {
+      "epoch": 0.21464,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.917096774193549e-05,
+      "loss": 0.1836,
+      "step": 13415
+    },
+    {
+      "epoch": 0.214656,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.916935483870969e-05,
+      "loss": 0.1452,
+      "step": 13416
+    },
+    {
+      "epoch": 0.214672,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.916774193548387e-05,
+      "loss": 0.1906,
+      "step": 13417
+    },
+    {
+      "epoch": 0.214688,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.916612903225806e-05,
+      "loss": 0.1646,
+      "step": 13418
+    },
+    {
+      "epoch": 0.214704,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.916451612903226e-05,
+      "loss": 0.1892,
+      "step": 13419
+    },
+    {
+      "epoch": 0.21472,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.916290322580646e-05,
+      "loss": 0.192,
+      "step": 13420
+    },
+    {
+      "epoch": 0.214736,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.916129032258066e-05,
+      "loss": 0.1795,
+      "step": 13421
+    },
+    {
+      "epoch": 0.214752,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.915967741935484e-05,
+      "loss": 0.1969,
+      "step": 13422
+    },
+    {
+      "epoch": 0.214768,
+      "grad_norm": 1.359375,
+      "learning_rate": 7.915806451612904e-05,
+      "loss": 0.1762,
+      "step": 13423
+    },
+    {
+      "epoch": 0.214784,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.915645161290323e-05,
+      "loss": 0.1838,
+      "step": 13424
+    },
+    {
+      "epoch": 0.2148,
+      "grad_norm": 1.3125,
+      "learning_rate": 7.915483870967743e-05,
+      "loss": 0.1827,
+      "step": 13425
+    },
+    {
+      "epoch": 0.214816,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.915322580645161e-05,
+      "loss": 0.1363,
+      "step": 13426
+    },
+    {
+      "epoch": 0.214832,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.915161290322581e-05,
+      "loss": 0.1517,
+      "step": 13427
+    },
+    {
+      "epoch": 0.214848,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.915e-05,
+      "loss": 0.1311,
+      "step": 13428
+    },
+    {
+      "epoch": 0.214864,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.91483870967742e-05,
+      "loss": 0.1415,
+      "step": 13429
+    },
+    {
+      "epoch": 0.21488,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.914677419354839e-05,
+      "loss": 0.189,
+      "step": 13430
+    },
+    {
+      "epoch": 0.214896,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.914516129032258e-05,
+      "loss": 0.1921,
+      "step": 13431
+    },
+    {
+      "epoch": 0.214912,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.914354838709677e-05,
+      "loss": 0.189,
+      "step": 13432
+    },
+    {
+      "epoch": 0.214928,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.914193548387097e-05,
+      "loss": 0.1609,
+      "step": 13433
+    },
+    {
+      "epoch": 0.214944,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.914032258064517e-05,
+      "loss": 0.1826,
+      "step": 13434
+    },
+    {
+      "epoch": 0.21496,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.913870967741936e-05,
+      "loss": 0.1412,
+      "step": 13435
+    },
+    {
+      "epoch": 0.214976,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.913709677419356e-05,
+      "loss": 0.1541,
+      "step": 13436
+    },
+    {
+      "epoch": 0.214992,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.913548387096774e-05,
+      "loss": 0.1481,
+      "step": 13437
+    },
+    {
+      "epoch": 0.215008,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.913387096774194e-05,
+      "loss": 0.1757,
+      "step": 13438
+    },
+    {
+      "epoch": 0.215024,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.913225806451613e-05,
+      "loss": 0.1703,
+      "step": 13439
+    },
+    {
+      "epoch": 0.21504,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.913064516129033e-05,
+      "loss": 0.2022,
+      "step": 13440
+    },
+    {
+      "epoch": 0.215056,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.912903225806451e-05,
+      "loss": 0.1644,
+      "step": 13441
+    },
+    {
+      "epoch": 0.215072,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.912741935483871e-05,
+      "loss": 0.1599,
+      "step": 13442
+    },
+    {
+      "epoch": 0.215088,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.91258064516129e-05,
+      "loss": 0.154,
+      "step": 13443
+    },
+    {
+      "epoch": 0.215104,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.91241935483871e-05,
+      "loss": 0.184,
+      "step": 13444
+    },
+    {
+      "epoch": 0.21512,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.91225806451613e-05,
+      "loss": 0.1361,
+      "step": 13445
+    },
+    {
+      "epoch": 0.215136,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.91209677419355e-05,
+      "loss": 0.1631,
+      "step": 13446
+    },
+    {
+      "epoch": 0.215152,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.911935483870968e-05,
+      "loss": 0.1773,
+      "step": 13447
+    },
+    {
+      "epoch": 0.215168,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.911774193548388e-05,
+      "loss": 0.1838,
+      "step": 13448
+    },
+    {
+      "epoch": 0.215184,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.911612903225807e-05,
+      "loss": 0.1839,
+      "step": 13449
+    },
+    {
+      "epoch": 0.2152,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.911451612903226e-05,
+      "loss": 0.1446,
+      "step": 13450
+    },
+    {
+      "epoch": 0.215216,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.911290322580646e-05,
+      "loss": 0.1753,
+      "step": 13451
+    },
+    {
+      "epoch": 0.215232,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.911129032258064e-05,
+      "loss": 0.1633,
+      "step": 13452
+    },
+    {
+      "epoch": 0.215248,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.910967741935484e-05,
+      "loss": 0.1559,
+      "step": 13453
+    },
+    {
+      "epoch": 0.215264,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.910806451612903e-05,
+      "loss": 0.1871,
+      "step": 13454
+    },
+    {
+      "epoch": 0.21528,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.910645161290323e-05,
+      "loss": 0.1482,
+      "step": 13455
+    },
+    {
+      "epoch": 0.215296,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.910483870967743e-05,
+      "loss": 0.1799,
+      "step": 13456
+    },
+    {
+      "epoch": 0.215312,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.910322580645163e-05,
+      "loss": 0.2161,
+      "step": 13457
+    },
+    {
+      "epoch": 0.215328,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.910161290322581e-05,
+      "loss": 0.1715,
+      "step": 13458
+    },
+    {
+      "epoch": 0.215344,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.910000000000001e-05,
+      "loss": 0.1672,
+      "step": 13459
+    },
+    {
+      "epoch": 0.21536,
+      "grad_norm": 0.75,
+      "learning_rate": 7.90983870967742e-05,
+      "loss": 0.1667,
+      "step": 13460
+    },
+    {
+      "epoch": 0.215376,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.90967741935484e-05,
+      "loss": 0.1624,
+      "step": 13461
+    },
+    {
+      "epoch": 0.215392,
+      "grad_norm": 1.2890625,
+      "learning_rate": 7.909516129032258e-05,
+      "loss": 0.2119,
+      "step": 13462
+    },
+    {
+      "epoch": 0.215408,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.909354838709678e-05,
+      "loss": 0.1525,
+      "step": 13463
+    },
+    {
+      "epoch": 0.215424,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.909193548387097e-05,
+      "loss": 0.1925,
+      "step": 13464
+    },
+    {
+      "epoch": 0.21544,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.909032258064516e-05,
+      "loss": 0.1621,
+      "step": 13465
+    },
+    {
+      "epoch": 0.215456,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.908870967741935e-05,
+      "loss": 0.1669,
+      "step": 13466
+    },
+    {
+      "epoch": 0.215472,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.908709677419354e-05,
+      "loss": 0.1437,
+      "step": 13467
+    },
+    {
+      "epoch": 0.215488,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.908548387096774e-05,
+      "loss": 0.1473,
+      "step": 13468
+    },
+    {
+      "epoch": 0.215504,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.908387096774194e-05,
+      "loss": 0.1855,
+      "step": 13469
+    },
+    {
+      "epoch": 0.21552,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.908225806451614e-05,
+      "loss": 0.1591,
+      "step": 13470
+    },
+    {
+      "epoch": 0.215536,
+      "grad_norm": 0.4375,
+      "learning_rate": 7.908064516129033e-05,
+      "loss": 0.1275,
+      "step": 13471
+    },
+    {
+      "epoch": 0.215552,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.907903225806453e-05,
+      "loss": 0.1804,
+      "step": 13472
+    },
+    {
+      "epoch": 0.215568,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.907741935483871e-05,
+      "loss": 0.1948,
+      "step": 13473
+    },
+    {
+      "epoch": 0.215584,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.907580645161291e-05,
+      "loss": 0.207,
+      "step": 13474
+    },
+    {
+      "epoch": 0.2156,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.90741935483871e-05,
+      "loss": 0.1592,
+      "step": 13475
+    },
+    {
+      "epoch": 0.215616,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.90725806451613e-05,
+      "loss": 0.1708,
+      "step": 13476
+    },
+    {
+      "epoch": 0.215632,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.907096774193548e-05,
+      "loss": 0.1578,
+      "step": 13477
+    },
+    {
+      "epoch": 0.215648,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.906935483870968e-05,
+      "loss": 0.1907,
+      "step": 13478
+    },
+    {
+      "epoch": 0.215664,
+      "grad_norm": 0.75,
+      "learning_rate": 7.906774193548387e-05,
+      "loss": 0.193,
+      "step": 13479
+    },
+    {
+      "epoch": 0.21568,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.906612903225807e-05,
+      "loss": 0.1801,
+      "step": 13480
+    },
+    {
+      "epoch": 0.215696,
+      "grad_norm": 1.25,
+      "learning_rate": 7.906451612903227e-05,
+      "loss": 0.2003,
+      "step": 13481
+    },
+    {
+      "epoch": 0.215712,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.906290322580645e-05,
+      "loss": 0.1598,
+      "step": 13482
+    },
+    {
+      "epoch": 0.215728,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.906129032258065e-05,
+      "loss": 0.1685,
+      "step": 13483
+    },
+    {
+      "epoch": 0.215744,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.905967741935484e-05,
+      "loss": 0.1359,
+      "step": 13484
+    },
+    {
+      "epoch": 0.21576,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.905806451612904e-05,
+      "loss": 0.1646,
+      "step": 13485
+    },
+    {
+      "epoch": 0.215776,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.905645161290323e-05,
+      "loss": 0.1792,
+      "step": 13486
+    },
+    {
+      "epoch": 0.215792,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.905483870967743e-05,
+      "loss": 0.1633,
+      "step": 13487
+    },
+    {
+      "epoch": 0.215808,
+      "grad_norm": 0.494140625,
+      "learning_rate": 7.905322580645161e-05,
+      "loss": 0.1705,
+      "step": 13488
+    },
+    {
+      "epoch": 0.215824,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.905161290322581e-05,
+      "loss": 0.1772,
+      "step": 13489
+    },
+    {
+      "epoch": 0.21584,
+      "grad_norm": 0.875,
+      "learning_rate": 7.905e-05,
+      "loss": 0.1664,
+      "step": 13490
+    },
+    {
+      "epoch": 0.215856,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.90483870967742e-05,
+      "loss": 0.1794,
+      "step": 13491
+    },
+    {
+      "epoch": 0.215872,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.90467741935484e-05,
+      "loss": 0.1702,
+      "step": 13492
+    },
+    {
+      "epoch": 0.215888,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.904516129032258e-05,
+      "loss": 0.1709,
+      "step": 13493
+    },
+    {
+      "epoch": 0.215904,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.904354838709678e-05,
+      "loss": 0.191,
+      "step": 13494
+    },
+    {
+      "epoch": 0.21592,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.904193548387098e-05,
+      "loss": 0.1619,
+      "step": 13495
+    },
+    {
+      "epoch": 0.215936,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.904032258064517e-05,
+      "loss": 0.1597,
+      "step": 13496
+    },
+    {
+      "epoch": 0.215952,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.903870967741935e-05,
+      "loss": 0.1657,
+      "step": 13497
+    },
+    {
+      "epoch": 0.215968,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.903709677419355e-05,
+      "loss": 0.1522,
+      "step": 13498
+    },
+    {
+      "epoch": 0.215984,
+      "grad_norm": 0.4921875,
+      "learning_rate": 7.903548387096774e-05,
+      "loss": 0.135,
+      "step": 13499
+    },
+    {
+      "epoch": 0.216,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.903387096774194e-05,
+      "loss": 0.1846,
+      "step": 13500
+    },
+    {
+      "epoch": 0.216016,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.903225806451613e-05,
+      "loss": 0.1546,
+      "step": 13501
+    },
+    {
+      "epoch": 0.216032,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.903064516129032e-05,
+      "loss": 0.205,
+      "step": 13502
+    },
+    {
+      "epoch": 0.216048,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.902903225806451e-05,
+      "loss": 0.1332,
+      "step": 13503
+    },
+    {
+      "epoch": 0.216064,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.902741935483871e-05,
+      "loss": 0.1697,
+      "step": 13504
+    },
+    {
+      "epoch": 0.21608,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.902580645161291e-05,
+      "loss": 0.1799,
+      "step": 13505
+    },
+    {
+      "epoch": 0.216096,
+      "grad_norm": 0.490234375,
+      "learning_rate": 7.902419354838711e-05,
+      "loss": 0.1403,
+      "step": 13506
+    },
+    {
+      "epoch": 0.216112,
+      "grad_norm": 0.75,
+      "learning_rate": 7.90225806451613e-05,
+      "loss": 0.178,
+      "step": 13507
+    },
+    {
+      "epoch": 0.216128,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.90209677419355e-05,
+      "loss": 0.1756,
+      "step": 13508
+    },
+    {
+      "epoch": 0.216144,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.901935483870968e-05,
+      "loss": 0.1455,
+      "step": 13509
+    },
+    {
+      "epoch": 0.21616,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.901774193548388e-05,
+      "loss": 0.1632,
+      "step": 13510
+    },
+    {
+      "epoch": 0.216176,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.901612903225807e-05,
+      "loss": 0.1564,
+      "step": 13511
+    },
+    {
+      "epoch": 0.216192,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.901451612903225e-05,
+      "loss": 0.174,
+      "step": 13512
+    },
+    {
+      "epoch": 0.216208,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.901290322580645e-05,
+      "loss": 0.2124,
+      "step": 13513
+    },
+    {
+      "epoch": 0.216224,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.901129032258064e-05,
+      "loss": 0.1519,
+      "step": 13514
+    },
+    {
+      "epoch": 0.21624,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.900967741935484e-05,
+      "loss": 0.1685,
+      "step": 13515
+    },
+    {
+      "epoch": 0.216256,
+      "grad_norm": 1.0,
+      "learning_rate": 7.900806451612904e-05,
+      "loss": 0.1791,
+      "step": 13516
+    },
+    {
+      "epoch": 0.216272,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.900645161290324e-05,
+      "loss": 0.1386,
+      "step": 13517
+    },
+    {
+      "epoch": 0.216288,
+      "grad_norm": 0.75,
+      "learning_rate": 7.900483870967742e-05,
+      "loss": 0.1236,
+      "step": 13518
+    },
+    {
+      "epoch": 0.216304,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.900322580645162e-05,
+      "loss": 0.1665,
+      "step": 13519
+    },
+    {
+      "epoch": 0.21632,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.900161290322581e-05,
+      "loss": 0.1708,
+      "step": 13520
+    },
+    {
+      "epoch": 0.216336,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.900000000000001e-05,
+      "loss": 0.1756,
+      "step": 13521
+    },
+    {
+      "epoch": 0.216352,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.89983870967742e-05,
+      "loss": 0.2184,
+      "step": 13522
+    },
+    {
+      "epoch": 0.216368,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.89967741935484e-05,
+      "loss": 0.166,
+      "step": 13523
+    },
+    {
+      "epoch": 0.216384,
+      "grad_norm": 1.8359375,
+      "learning_rate": 7.899516129032258e-05,
+      "loss": 0.1778,
+      "step": 13524
+    },
+    {
+      "epoch": 0.2164,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.899354838709678e-05,
+      "loss": 0.2077,
+      "step": 13525
+    },
+    {
+      "epoch": 0.216416,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.899193548387097e-05,
+      "loss": 0.1755,
+      "step": 13526
+    },
+    {
+      "epoch": 0.216432,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.899032258064515e-05,
+      "loss": 0.1871,
+      "step": 13527
+    },
+    {
+      "epoch": 0.216448,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.898870967741935e-05,
+      "loss": 0.1847,
+      "step": 13528
+    },
+    {
+      "epoch": 0.216464,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.898709677419355e-05,
+      "loss": 0.1599,
+      "step": 13529
+    },
+    {
+      "epoch": 0.21648,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.898548387096775e-05,
+      "loss": 0.1667,
+      "step": 13530
+    },
+    {
+      "epoch": 0.216496,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.898387096774194e-05,
+      "loss": 0.1431,
+      "step": 13531
+    },
+    {
+      "epoch": 0.216512,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.898225806451614e-05,
+      "loss": 0.2013,
+      "step": 13532
+    },
+    {
+      "epoch": 0.216528,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.898064516129032e-05,
+      "loss": 0.1718,
+      "step": 13533
+    },
+    {
+      "epoch": 0.216544,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.897903225806452e-05,
+      "loss": 0.1961,
+      "step": 13534
+    },
+    {
+      "epoch": 0.21656,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.897741935483871e-05,
+      "loss": 0.1843,
+      "step": 13535
+    },
+    {
+      "epoch": 0.216576,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.897580645161291e-05,
+      "loss": 0.2277,
+      "step": 13536
+    },
+    {
+      "epoch": 0.216592,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.89741935483871e-05,
+      "loss": 0.2019,
+      "step": 13537
+    },
+    {
+      "epoch": 0.216608,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.89725806451613e-05,
+      "loss": 0.1554,
+      "step": 13538
+    },
+    {
+      "epoch": 0.216624,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.897096774193548e-05,
+      "loss": 0.1797,
+      "step": 13539
+    },
+    {
+      "epoch": 0.21664,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.896935483870968e-05,
+      "loss": 0.1684,
+      "step": 13540
+    },
+    {
+      "epoch": 0.216656,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.896774193548388e-05,
+      "loss": 0.1724,
+      "step": 13541
+    },
+    {
+      "epoch": 0.216672,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.896612903225808e-05,
+      "loss": 0.1674,
+      "step": 13542
+    },
+    {
+      "epoch": 0.216688,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.896451612903227e-05,
+      "loss": 0.1247,
+      "step": 13543
+    },
+    {
+      "epoch": 0.216704,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.896290322580645e-05,
+      "loss": 0.1679,
+      "step": 13544
+    },
+    {
+      "epoch": 0.21672,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.896129032258065e-05,
+      "loss": 0.179,
+      "step": 13545
+    },
+    {
+      "epoch": 0.216736,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.895967741935484e-05,
+      "loss": 0.1168,
+      "step": 13546
+    },
+    {
+      "epoch": 0.216752,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.895806451612904e-05,
+      "loss": 0.2351,
+      "step": 13547
+    },
+    {
+      "epoch": 0.216768,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.895645161290322e-05,
+      "loss": 0.1572,
+      "step": 13548
+    },
+    {
+      "epoch": 0.216784,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.895483870967742e-05,
+      "loss": 0.1215,
+      "step": 13549
+    },
+    {
+      "epoch": 0.2168,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.895322580645161e-05,
+      "loss": 0.1938,
+      "step": 13550
+    },
+    {
+      "epoch": 0.216816,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.895161290322581e-05,
+      "loss": 0.1867,
+      "step": 13551
+    },
+    {
+      "epoch": 0.216832,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.895000000000001e-05,
+      "loss": 0.1755,
+      "step": 13552
+    },
+    {
+      "epoch": 0.216848,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.894838709677421e-05,
+      "loss": 0.1559,
+      "step": 13553
+    },
+    {
+      "epoch": 0.216864,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.89467741935484e-05,
+      "loss": 0.1805,
+      "step": 13554
+    },
+    {
+      "epoch": 0.21688,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.89451612903226e-05,
+      "loss": 0.1759,
+      "step": 13555
+    },
+    {
+      "epoch": 0.216896,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.894354838709678e-05,
+      "loss": 0.1697,
+      "step": 13556
+    },
+    {
+      "epoch": 0.216912,
+      "grad_norm": 1.0,
+      "learning_rate": 7.894193548387098e-05,
+      "loss": 0.1951,
+      "step": 13557
+    },
+    {
+      "epoch": 0.216928,
+      "grad_norm": 1.2421875,
+      "learning_rate": 7.894032258064517e-05,
+      "loss": 0.1896,
+      "step": 13558
+    },
+    {
+      "epoch": 0.216944,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.893870967741935e-05,
+      "loss": 0.1513,
+      "step": 13559
+    },
+    {
+      "epoch": 0.21696,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.893709677419355e-05,
+      "loss": 0.1723,
+      "step": 13560
+    },
+    {
+      "epoch": 0.216976,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.893548387096774e-05,
+      "loss": 0.1717,
+      "step": 13561
+    },
+    {
+      "epoch": 0.216992,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.893387096774194e-05,
+      "loss": 0.1535,
+      "step": 13562
+    },
+    {
+      "epoch": 0.217008,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.893225806451612e-05,
+      "loss": 0.1879,
+      "step": 13563
+    },
+    {
+      "epoch": 0.217024,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.893064516129032e-05,
+      "loss": 0.1545,
+      "step": 13564
+    },
+    {
+      "epoch": 0.21704,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.892903225806452e-05,
+      "loss": 0.175,
+      "step": 13565
+    },
+    {
+      "epoch": 0.217056,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.892741935483872e-05,
+      "loss": 0.1357,
+      "step": 13566
+    },
+    {
+      "epoch": 0.217072,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.892580645161291e-05,
+      "loss": 0.1626,
+      "step": 13567
+    },
+    {
+      "epoch": 0.217088,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.892419354838711e-05,
+      "loss": 0.1604,
+      "step": 13568
+    },
+    {
+      "epoch": 0.217104,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.89225806451613e-05,
+      "loss": 0.1651,
+      "step": 13569
+    },
+    {
+      "epoch": 0.21712,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.892096774193549e-05,
+      "loss": 0.1932,
+      "step": 13570
+    },
+    {
+      "epoch": 0.217136,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.891935483870968e-05,
+      "loss": 0.162,
+      "step": 13571
+    },
+    {
+      "epoch": 0.217152,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.891774193548388e-05,
+      "loss": 0.2167,
+      "step": 13572
+    },
+    {
+      "epoch": 0.217168,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.891612903225806e-05,
+      "loss": 0.1605,
+      "step": 13573
+    },
+    {
+      "epoch": 0.217184,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.891451612903225e-05,
+      "loss": 0.1678,
+      "step": 13574
+    },
+    {
+      "epoch": 0.2172,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.891290322580645e-05,
+      "loss": 0.1588,
+      "step": 13575
+    },
+    {
+      "epoch": 0.217216,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.891129032258065e-05,
+      "loss": 0.1642,
+      "step": 13576
+    },
+    {
+      "epoch": 0.217232,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.890967741935485e-05,
+      "loss": 0.1551,
+      "step": 13577
+    },
+    {
+      "epoch": 0.217248,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.890806451612904e-05,
+      "loss": 0.179,
+      "step": 13578
+    },
+    {
+      "epoch": 0.217264,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.890645161290324e-05,
+      "loss": 0.149,
+      "step": 13579
+    },
+    {
+      "epoch": 0.21728,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.890483870967742e-05,
+      "loss": 0.1652,
+      "step": 13580
+    },
+    {
+      "epoch": 0.217296,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.890322580645162e-05,
+      "loss": 0.1605,
+      "step": 13581
+    },
+    {
+      "epoch": 0.217312,
+      "grad_norm": 1.3828125,
+      "learning_rate": 7.890161290322581e-05,
+      "loss": 0.1861,
+      "step": 13582
+    },
+    {
+      "epoch": 0.217328,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.890000000000001e-05,
+      "loss": 0.1375,
+      "step": 13583
+    },
+    {
+      "epoch": 0.217344,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.889838709677419e-05,
+      "loss": 0.1714,
+      "step": 13584
+    },
+    {
+      "epoch": 0.21736,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.889677419354839e-05,
+      "loss": 0.2262,
+      "step": 13585
+    },
+    {
+      "epoch": 0.217376,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.889516129032258e-05,
+      "loss": 0.1539,
+      "step": 13586
+    },
+    {
+      "epoch": 0.217392,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.889354838709678e-05,
+      "loss": 0.1647,
+      "step": 13587
+    },
+    {
+      "epoch": 0.217408,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.889193548387098e-05,
+      "loss": 0.181,
+      "step": 13588
+    },
+    {
+      "epoch": 0.217424,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.889032258064516e-05,
+      "loss": 0.1807,
+      "step": 13589
+    },
+    {
+      "epoch": 0.21744,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.888870967741936e-05,
+      "loss": 0.1695,
+      "step": 13590
+    },
+    {
+      "epoch": 0.217456,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.888709677419355e-05,
+      "loss": 0.1814,
+      "step": 13591
+    },
+    {
+      "epoch": 0.217472,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.888548387096775e-05,
+      "loss": 0.1392,
+      "step": 13592
+    },
+    {
+      "epoch": 0.217488,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.888387096774194e-05,
+      "loss": 0.1754,
+      "step": 13593
+    },
+    {
+      "epoch": 0.217504,
+      "grad_norm": 0.75,
+      "learning_rate": 7.888225806451614e-05,
+      "loss": 0.1901,
+      "step": 13594
+    },
+    {
+      "epoch": 0.21752,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.888064516129032e-05,
+      "loss": 0.1388,
+      "step": 13595
+    },
+    {
+      "epoch": 0.217536,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.887903225806452e-05,
+      "loss": 0.1646,
+      "step": 13596
+    },
+    {
+      "epoch": 0.217552,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.887741935483871e-05,
+      "loss": 0.1521,
+      "step": 13597
+    },
+    {
+      "epoch": 0.217568,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.88758064516129e-05,
+      "loss": 0.1839,
+      "step": 13598
+    },
+    {
+      "epoch": 0.217584,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.887419354838709e-05,
+      "loss": 0.1545,
+      "step": 13599
+    },
+    {
+      "epoch": 0.2176,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.887258064516129e-05,
+      "loss": 0.1756,
+      "step": 13600
+    },
+    {
+      "epoch": 0.217616,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.887096774193549e-05,
+      "loss": 0.1853,
+      "step": 13601
+    },
+    {
+      "epoch": 0.217632,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.886935483870969e-05,
+      "loss": 0.1809,
+      "step": 13602
+    },
+    {
+      "epoch": 0.217648,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.886774193548388e-05,
+      "loss": 0.17,
+      "step": 13603
+    },
+    {
+      "epoch": 0.217664,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.886612903225808e-05,
+      "loss": 0.1897,
+      "step": 13604
+    },
+    {
+      "epoch": 0.21768,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.886451612903226e-05,
+      "loss": 0.1458,
+      "step": 13605
+    },
+    {
+      "epoch": 0.217696,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.886290322580645e-05,
+      "loss": 0.1591,
+      "step": 13606
+    },
+    {
+      "epoch": 0.217712,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.886129032258065e-05,
+      "loss": 0.1699,
+      "step": 13607
+    },
+    {
+      "epoch": 0.217728,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.885967741935484e-05,
+      "loss": 0.1598,
+      "step": 13608
+    },
+    {
+      "epoch": 0.217744,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.885806451612903e-05,
+      "loss": 0.1784,
+      "step": 13609
+    },
+    {
+      "epoch": 0.21776,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.885645161290322e-05,
+      "loss": 0.1758,
+      "step": 13610
+    },
+    {
+      "epoch": 0.217776,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.885483870967742e-05,
+      "loss": 0.1822,
+      "step": 13611
+    },
+    {
+      "epoch": 0.217792,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.885322580645162e-05,
+      "loss": 0.1455,
+      "step": 13612
+    },
+    {
+      "epoch": 0.217808,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.885161290322582e-05,
+      "loss": 0.1606,
+      "step": 13613
+    },
+    {
+      "epoch": 0.217824,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.885e-05,
+      "loss": 0.2216,
+      "step": 13614
+    },
+    {
+      "epoch": 0.21784,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.88483870967742e-05,
+      "loss": 0.1879,
+      "step": 13615
+    },
+    {
+      "epoch": 0.217856,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.884677419354839e-05,
+      "loss": 0.1491,
+      "step": 13616
+    },
+    {
+      "epoch": 0.217872,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.884516129032259e-05,
+      "loss": 0.1813,
+      "step": 13617
+    },
+    {
+      "epoch": 0.217888,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.884354838709678e-05,
+      "loss": 0.1589,
+      "step": 13618
+    },
+    {
+      "epoch": 0.217904,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.884193548387098e-05,
+      "loss": 0.1628,
+      "step": 13619
+    },
+    {
+      "epoch": 0.21792,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.884032258064516e-05,
+      "loss": 0.1601,
+      "step": 13620
+    },
+    {
+      "epoch": 0.217936,
+      "grad_norm": 1.890625,
+      "learning_rate": 7.883870967741935e-05,
+      "loss": 0.2114,
+      "step": 13621
+    },
+    {
+      "epoch": 0.217952,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.883709677419355e-05,
+      "loss": 0.1705,
+      "step": 13622
+    },
+    {
+      "epoch": 0.217968,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.883548387096773e-05,
+      "loss": 0.1907,
+      "step": 13623
+    },
+    {
+      "epoch": 0.217984,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.883387096774193e-05,
+      "loss": 0.2173,
+      "step": 13624
+    },
+    {
+      "epoch": 0.218,
+      "grad_norm": 1.578125,
+      "learning_rate": 7.883225806451613e-05,
+      "loss": 0.1704,
+      "step": 13625
+    },
+    {
+      "epoch": 0.218016,
+      "grad_norm": 1.3671875,
+      "learning_rate": 7.883064516129033e-05,
+      "loss": 0.188,
+      "step": 13626
+    },
+    {
+      "epoch": 0.218032,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.882903225806452e-05,
+      "loss": 0.1821,
+      "step": 13627
+    },
+    {
+      "epoch": 0.218048,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.882741935483872e-05,
+      "loss": 0.1562,
+      "step": 13628
+    },
+    {
+      "epoch": 0.218064,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.88258064516129e-05,
+      "loss": 0.1861,
+      "step": 13629
+    },
+    {
+      "epoch": 0.21808,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.88241935483871e-05,
+      "loss": 0.1871,
+      "step": 13630
+    },
+    {
+      "epoch": 0.218096,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.882258064516129e-05,
+      "loss": 0.1531,
+      "step": 13631
+    },
+    {
+      "epoch": 0.218112,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.882096774193549e-05,
+      "loss": 0.2013,
+      "step": 13632
+    },
+    {
+      "epoch": 0.218128,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.881935483870968e-05,
+      "loss": 0.1318,
+      "step": 13633
+    },
+    {
+      "epoch": 0.218144,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.881774193548388e-05,
+      "loss": 0.1966,
+      "step": 13634
+    },
+    {
+      "epoch": 0.21816,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.881612903225806e-05,
+      "loss": 0.1568,
+      "step": 13635
+    },
+    {
+      "epoch": 0.218176,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.881451612903226e-05,
+      "loss": 0.1505,
+      "step": 13636
+    },
+    {
+      "epoch": 0.218192,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.881290322580646e-05,
+      "loss": 0.1474,
+      "step": 13637
+    },
+    {
+      "epoch": 0.218208,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.881129032258065e-05,
+      "loss": 0.177,
+      "step": 13638
+    },
+    {
+      "epoch": 0.218224,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.880967741935485e-05,
+      "loss": 0.1812,
+      "step": 13639
+    },
+    {
+      "epoch": 0.21824,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.880806451612903e-05,
+      "loss": 0.1583,
+      "step": 13640
+    },
+    {
+      "epoch": 0.218256,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.880645161290323e-05,
+      "loss": 0.1442,
+      "step": 13641
+    },
+    {
+      "epoch": 0.218272,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.880483870967742e-05,
+      "loss": 0.1398,
+      "step": 13642
+    },
+    {
+      "epoch": 0.218288,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.880322580645162e-05,
+      "loss": 0.1673,
+      "step": 13643
+    },
+    {
+      "epoch": 0.218304,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.88016129032258e-05,
+      "loss": 0.1656,
+      "step": 13644
+    },
+    {
+      "epoch": 0.21832,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.88e-05,
+      "loss": 0.1505,
+      "step": 13645
+    },
+    {
+      "epoch": 0.218336,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.879838709677419e-05,
+      "loss": 0.1831,
+      "step": 13646
+    },
+    {
+      "epoch": 0.218352,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.879677419354839e-05,
+      "loss": 0.1754,
+      "step": 13647
+    },
+    {
+      "epoch": 0.218368,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.879516129032259e-05,
+      "loss": 0.1691,
+      "step": 13648
+    },
+    {
+      "epoch": 0.218384,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.879354838709679e-05,
+      "loss": 0.1787,
+      "step": 13649
+    },
+    {
+      "epoch": 0.2184,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.879193548387098e-05,
+      "loss": 0.1456,
+      "step": 13650
+    },
+    {
+      "epoch": 0.218416,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.879032258064518e-05,
+      "loss": 0.1496,
+      "step": 13651
+    },
+    {
+      "epoch": 0.218432,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.878870967741936e-05,
+      "loss": 0.1882,
+      "step": 13652
+    },
+    {
+      "epoch": 0.218448,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.878709677419355e-05,
+      "loss": 0.155,
+      "step": 13653
+    },
+    {
+      "epoch": 0.218464,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.878548387096775e-05,
+      "loss": 0.1361,
+      "step": 13654
+    },
+    {
+      "epoch": 0.21848,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.878387096774193e-05,
+      "loss": 0.1748,
+      "step": 13655
+    },
+    {
+      "epoch": 0.218496,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.878225806451613e-05,
+      "loss": 0.2071,
+      "step": 13656
+    },
+    {
+      "epoch": 0.218512,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.878064516129032e-05,
+      "loss": 0.1725,
+      "step": 13657
+    },
+    {
+      "epoch": 0.218528,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.877903225806452e-05,
+      "loss": 0.1486,
+      "step": 13658
+    },
+    {
+      "epoch": 0.218544,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.87774193548387e-05,
+      "loss": 0.137,
+      "step": 13659
+    },
+    {
+      "epoch": 0.21856,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.87758064516129e-05,
+      "loss": 0.1633,
+      "step": 13660
+    },
+    {
+      "epoch": 0.218576,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.87741935483871e-05,
+      "loss": 0.1929,
+      "step": 13661
+    },
+    {
+      "epoch": 0.218592,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.87725806451613e-05,
+      "loss": 0.1519,
+      "step": 13662
+    },
+    {
+      "epoch": 0.218608,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.877096774193549e-05,
+      "loss": 0.1822,
+      "step": 13663
+    },
+    {
+      "epoch": 0.218624,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.876935483870969e-05,
+      "loss": 0.1384,
+      "step": 13664
+    },
+    {
+      "epoch": 0.21864,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.876774193548388e-05,
+      "loss": 0.1932,
+      "step": 13665
+    },
+    {
+      "epoch": 0.218656,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.876612903225807e-05,
+      "loss": 0.1692,
+      "step": 13666
+    },
+    {
+      "epoch": 0.218672,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.876451612903226e-05,
+      "loss": 0.1401,
+      "step": 13667
+    },
+    {
+      "epoch": 0.218688,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.876290322580645e-05,
+      "loss": 0.1418,
+      "step": 13668
+    },
+    {
+      "epoch": 0.218704,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.876129032258065e-05,
+      "loss": 0.1477,
+      "step": 13669
+    },
+    {
+      "epoch": 0.21872,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.875967741935483e-05,
+      "loss": 0.2022,
+      "step": 13670
+    },
+    {
+      "epoch": 0.218736,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.875806451612903e-05,
+      "loss": 0.1657,
+      "step": 13671
+    },
+    {
+      "epoch": 0.218752,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.875645161290323e-05,
+      "loss": 0.1878,
+      "step": 13672
+    },
+    {
+      "epoch": 0.218768,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.875483870967743e-05,
+      "loss": 0.1733,
+      "step": 13673
+    },
+    {
+      "epoch": 0.218784,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.875322580645162e-05,
+      "loss": 0.1562,
+      "step": 13674
+    },
+    {
+      "epoch": 0.2188,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.875161290322582e-05,
+      "loss": 0.1802,
+      "step": 13675
+    },
+    {
+      "epoch": 0.218816,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.875e-05,
+      "loss": 0.1629,
+      "step": 13676
+    },
+    {
+      "epoch": 0.218832,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.87483870967742e-05,
+      "loss": 0.1529,
+      "step": 13677
+    },
+    {
+      "epoch": 0.218848,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.874677419354839e-05,
+      "loss": 0.1944,
+      "step": 13678
+    },
+    {
+      "epoch": 0.218864,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.874516129032259e-05,
+      "loss": 0.1648,
+      "step": 13679
+    },
+    {
+      "epoch": 0.21888,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.874354838709677e-05,
+      "loss": 0.1773,
+      "step": 13680
+    },
+    {
+      "epoch": 0.218896,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.874193548387097e-05,
+      "loss": 0.1681,
+      "step": 13681
+    },
+    {
+      "epoch": 0.218912,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.874032258064516e-05,
+      "loss": 0.1594,
+      "step": 13682
+    },
+    {
+      "epoch": 0.218928,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.873870967741936e-05,
+      "loss": 0.1859,
+      "step": 13683
+    },
+    {
+      "epoch": 0.218944,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.873709677419355e-05,
+      "loss": 0.1284,
+      "step": 13684
+    },
+    {
+      "epoch": 0.21896,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.873548387096775e-05,
+      "loss": 0.1894,
+      "step": 13685
+    },
+    {
+      "epoch": 0.218976,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.873387096774195e-05,
+      "loss": 0.1902,
+      "step": 13686
+    },
+    {
+      "epoch": 0.218992,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.873225806451613e-05,
+      "loss": 0.2145,
+      "step": 13687
+    },
+    {
+      "epoch": 0.219008,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.873064516129033e-05,
+      "loss": 0.1827,
+      "step": 13688
+    },
+    {
+      "epoch": 0.219024,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.872903225806452e-05,
+      "loss": 0.162,
+      "step": 13689
+    },
+    {
+      "epoch": 0.21904,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.872741935483872e-05,
+      "loss": 0.1953,
+      "step": 13690
+    },
+    {
+      "epoch": 0.219056,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.87258064516129e-05,
+      "loss": 0.1861,
+      "step": 13691
+    },
+    {
+      "epoch": 0.219072,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.87241935483871e-05,
+      "loss": 0.1354,
+      "step": 13692
+    },
+    {
+      "epoch": 0.219088,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.872258064516129e-05,
+      "loss": 0.19,
+      "step": 13693
+    },
+    {
+      "epoch": 0.219104,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.872096774193549e-05,
+      "loss": 0.1525,
+      "step": 13694
+    },
+    {
+      "epoch": 0.21912,
+      "grad_norm": 0.482421875,
+      "learning_rate": 7.871935483870967e-05,
+      "loss": 0.1542,
+      "step": 13695
+    },
+    {
+      "epoch": 0.219136,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.871774193548387e-05,
+      "loss": 0.1823,
+      "step": 13696
+    },
+    {
+      "epoch": 0.219152,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.871612903225807e-05,
+      "loss": 0.1559,
+      "step": 13697
+    },
+    {
+      "epoch": 0.219168,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.871451612903226e-05,
+      "loss": 0.199,
+      "step": 13698
+    },
+    {
+      "epoch": 0.219184,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.871290322580646e-05,
+      "loss": 0.2052,
+      "step": 13699
+    },
+    {
+      "epoch": 0.2192,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.871129032258065e-05,
+      "loss": 0.1784,
+      "step": 13700
+    },
+    {
+      "epoch": 0.219216,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.870967741935484e-05,
+      "loss": 0.1847,
+      "step": 13701
+    },
+    {
+      "epoch": 0.219232,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.870806451612903e-05,
+      "loss": 0.1671,
+      "step": 13702
+    },
+    {
+      "epoch": 0.219248,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.870645161290323e-05,
+      "loss": 0.2054,
+      "step": 13703
+    },
+    {
+      "epoch": 0.219264,
+      "grad_norm": 1.125,
+      "learning_rate": 7.870483870967742e-05,
+      "loss": 0.1544,
+      "step": 13704
+    },
+    {
+      "epoch": 0.21928,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.870322580645162e-05,
+      "loss": 0.1804,
+      "step": 13705
+    },
+    {
+      "epoch": 0.219296,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.87016129032258e-05,
+      "loss": 0.153,
+      "step": 13706
+    },
+    {
+      "epoch": 0.219312,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.87e-05,
+      "loss": 0.1576,
+      "step": 13707
+    },
+    {
+      "epoch": 0.219328,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.86983870967742e-05,
+      "loss": 0.1374,
+      "step": 13708
+    },
+    {
+      "epoch": 0.219344,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.86967741935484e-05,
+      "loss": 0.1546,
+      "step": 13709
+    },
+    {
+      "epoch": 0.21936,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.869516129032259e-05,
+      "loss": 0.1757,
+      "step": 13710
+    },
+    {
+      "epoch": 0.219376,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.869354838709679e-05,
+      "loss": 0.1554,
+      "step": 13711
+    },
+    {
+      "epoch": 0.219392,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.869193548387097e-05,
+      "loss": 0.1901,
+      "step": 13712
+    },
+    {
+      "epoch": 0.219408,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.869032258064517e-05,
+      "loss": 0.2091,
+      "step": 13713
+    },
+    {
+      "epoch": 0.219424,
+      "grad_norm": 1.125,
+      "learning_rate": 7.868870967741936e-05,
+      "loss": 0.1749,
+      "step": 13714
+    },
+    {
+      "epoch": 0.21944,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.868709677419354e-05,
+      "loss": 0.1783,
+      "step": 13715
+    },
+    {
+      "epoch": 0.219456,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.868548387096774e-05,
+      "loss": 0.1566,
+      "step": 13716
+    },
+    {
+      "epoch": 0.219472,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.868387096774193e-05,
+      "loss": 0.1313,
+      "step": 13717
+    },
+    {
+      "epoch": 0.219488,
+      "grad_norm": 1.125,
+      "learning_rate": 7.868225806451613e-05,
+      "loss": 0.1463,
+      "step": 13718
+    },
+    {
+      "epoch": 0.219504,
+      "grad_norm": 1.453125,
+      "learning_rate": 7.868064516129032e-05,
+      "loss": 0.2027,
+      "step": 13719
+    },
+    {
+      "epoch": 0.21952,
+      "grad_norm": 1.2421875,
+      "learning_rate": 7.867903225806452e-05,
+      "loss": 0.2334,
+      "step": 13720
+    },
+    {
+      "epoch": 0.219536,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.867741935483872e-05,
+      "loss": 0.1972,
+      "step": 13721
+    },
+    {
+      "epoch": 0.219552,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.867580645161292e-05,
+      "loss": 0.1622,
+      "step": 13722
+    },
+    {
+      "epoch": 0.219568,
+      "grad_norm": 1.8984375,
+      "learning_rate": 7.86741935483871e-05,
+      "loss": 0.1684,
+      "step": 13723
+    },
+    {
+      "epoch": 0.219584,
+      "grad_norm": 1.453125,
+      "learning_rate": 7.86725806451613e-05,
+      "loss": 0.2306,
+      "step": 13724
+    },
+    {
+      "epoch": 0.2196,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.867096774193549e-05,
+      "loss": 0.2031,
+      "step": 13725
+    },
+    {
+      "epoch": 0.219616,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.866935483870969e-05,
+      "loss": 0.1663,
+      "step": 13726
+    },
+    {
+      "epoch": 0.219632,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.866774193548387e-05,
+      "loss": 0.1856,
+      "step": 13727
+    },
+    {
+      "epoch": 0.219648,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.866612903225807e-05,
+      "loss": 0.1432,
+      "step": 13728
+    },
+    {
+      "epoch": 0.219664,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.866451612903226e-05,
+      "loss": 0.1754,
+      "step": 13729
+    },
+    {
+      "epoch": 0.21968,
+      "grad_norm": 1.2890625,
+      "learning_rate": 7.866290322580644e-05,
+      "loss": 0.1832,
+      "step": 13730
+    },
+    {
+      "epoch": 0.219696,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.866129032258064e-05,
+      "loss": 0.1429,
+      "step": 13731
+    },
+    {
+      "epoch": 0.219712,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.865967741935484e-05,
+      "loss": 0.1901,
+      "step": 13732
+    },
+    {
+      "epoch": 0.219728,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.865806451612904e-05,
+      "loss": 0.1636,
+      "step": 13733
+    },
+    {
+      "epoch": 0.219744,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.865645161290323e-05,
+      "loss": 0.2204,
+      "step": 13734
+    },
+    {
+      "epoch": 0.21976,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.865483870967743e-05,
+      "loss": 0.1507,
+      "step": 13735
+    },
+    {
+      "epoch": 0.219776,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.865322580645162e-05,
+      "loss": 0.172,
+      "step": 13736
+    },
+    {
+      "epoch": 0.219792,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.865161290322581e-05,
+      "loss": 0.1718,
+      "step": 13737
+    },
+    {
+      "epoch": 0.219808,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.865e-05,
+      "loss": 0.1476,
+      "step": 13738
+    },
+    {
+      "epoch": 0.219824,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.86483870967742e-05,
+      "loss": 0.1695,
+      "step": 13739
+    },
+    {
+      "epoch": 0.21984,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.864677419354839e-05,
+      "loss": 0.1952,
+      "step": 13740
+    },
+    {
+      "epoch": 0.219856,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.864516129032259e-05,
+      "loss": 0.1813,
+      "step": 13741
+    },
+    {
+      "epoch": 0.219872,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.864354838709677e-05,
+      "loss": 0.2103,
+      "step": 13742
+    },
+    {
+      "epoch": 0.219888,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.864193548387097e-05,
+      "loss": 0.1961,
+      "step": 13743
+    },
+    {
+      "epoch": 0.219904,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.864032258064517e-05,
+      "loss": 0.1574,
+      "step": 13744
+    },
+    {
+      "epoch": 0.21992,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.863870967741936e-05,
+      "loss": 0.1626,
+      "step": 13745
+    },
+    {
+      "epoch": 0.219936,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.863709677419356e-05,
+      "loss": 0.2175,
+      "step": 13746
+    },
+    {
+      "epoch": 0.219952,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.863548387096774e-05,
+      "loss": 0.1573,
+      "step": 13747
+    },
+    {
+      "epoch": 0.219968,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.863387096774194e-05,
+      "loss": 0.1538,
+      "step": 13748
+    },
+    {
+      "epoch": 0.219984,
+      "grad_norm": 1.4375,
+      "learning_rate": 7.863225806451613e-05,
+      "loss": 0.1681,
+      "step": 13749
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.863064516129033e-05,
+      "loss": 0.1682,
+      "step": 13750
+    },
+    {
+      "epoch": 0.220016,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.862903225806451e-05,
+      "loss": 0.1507,
+      "step": 13751
+    },
+    {
+      "epoch": 0.220032,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.862741935483871e-05,
+      "loss": 0.1529,
+      "step": 13752
+    },
+    {
+      "epoch": 0.220048,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.86258064516129e-05,
+      "loss": 0.1445,
+      "step": 13753
+    },
+    {
+      "epoch": 0.220064,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.86241935483871e-05,
+      "loss": 0.1752,
+      "step": 13754
+    },
+    {
+      "epoch": 0.22008,
+      "grad_norm": 1.2578125,
+      "learning_rate": 7.862258064516129e-05,
+      "loss": 0.1899,
+      "step": 13755
+    },
+    {
+      "epoch": 0.220096,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.862096774193549e-05,
+      "loss": 0.1961,
+      "step": 13756
+    },
+    {
+      "epoch": 0.220112,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.861935483870969e-05,
+      "loss": 0.1884,
+      "step": 13757
+    },
+    {
+      "epoch": 0.220128,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.861774193548389e-05,
+      "loss": 0.1504,
+      "step": 13758
+    },
+    {
+      "epoch": 0.220144,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.861612903225807e-05,
+      "loss": 0.1578,
+      "step": 13759
+    },
+    {
+      "epoch": 0.22016,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.861451612903227e-05,
+      "loss": 0.1716,
+      "step": 13760
+    },
+    {
+      "epoch": 0.220176,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.861290322580646e-05,
+      "loss": 0.1706,
+      "step": 13761
+    },
+    {
+      "epoch": 0.220192,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.861129032258064e-05,
+      "loss": 0.1658,
+      "step": 13762
+    },
+    {
+      "epoch": 0.220208,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.860967741935484e-05,
+      "loss": 0.1582,
+      "step": 13763
+    },
+    {
+      "epoch": 0.220224,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.860806451612903e-05,
+      "loss": 0.1226,
+      "step": 13764
+    },
+    {
+      "epoch": 0.22024,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.860645161290323e-05,
+      "loss": 0.1509,
+      "step": 13765
+    },
+    {
+      "epoch": 0.220256,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.860483870967741e-05,
+      "loss": 0.1637,
+      "step": 13766
+    },
+    {
+      "epoch": 0.220272,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.860322580645161e-05,
+      "loss": 0.1737,
+      "step": 13767
+    },
+    {
+      "epoch": 0.220288,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.860161290322581e-05,
+      "loss": 0.1982,
+      "step": 13768
+    },
+    {
+      "epoch": 0.220304,
+      "grad_norm": 1.0,
+      "learning_rate": 7.860000000000001e-05,
+      "loss": 0.2036,
+      "step": 13769
+    },
+    {
+      "epoch": 0.22032,
+      "grad_norm": 1.3671875,
+      "learning_rate": 7.85983870967742e-05,
+      "loss": 0.1809,
+      "step": 13770
+    },
+    {
+      "epoch": 0.220336,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.85967741935484e-05,
+      "loss": 0.1889,
+      "step": 13771
+    },
+    {
+      "epoch": 0.220352,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.859516129032258e-05,
+      "loss": 0.14,
+      "step": 13772
+    },
+    {
+      "epoch": 0.220368,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.859354838709678e-05,
+      "loss": 0.1416,
+      "step": 13773
+    },
+    {
+      "epoch": 0.220384,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.859193548387097e-05,
+      "loss": 0.1498,
+      "step": 13774
+    },
+    {
+      "epoch": 0.2204,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.859032258064517e-05,
+      "loss": 0.1659,
+      "step": 13775
+    },
+    {
+      "epoch": 0.220416,
+      "grad_norm": 1.2421875,
+      "learning_rate": 7.858870967741936e-05,
+      "loss": 0.2002,
+      "step": 13776
+    },
+    {
+      "epoch": 0.220432,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.858709677419354e-05,
+      "loss": 0.1873,
+      "step": 13777
+    },
+    {
+      "epoch": 0.220448,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.858548387096774e-05,
+      "loss": 0.1545,
+      "step": 13778
+    },
+    {
+      "epoch": 0.220464,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.858387096774193e-05,
+      "loss": 0.1713,
+      "step": 13779
+    },
+    {
+      "epoch": 0.22048,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.858225806451613e-05,
+      "loss": 0.1364,
+      "step": 13780
+    },
+    {
+      "epoch": 0.220496,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.858064516129033e-05,
+      "loss": 0.13,
+      "step": 13781
+    },
+    {
+      "epoch": 0.220512,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.857903225806453e-05,
+      "loss": 0.1524,
+      "step": 13782
+    },
+    {
+      "epoch": 0.220528,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.857741935483871e-05,
+      "loss": 0.1644,
+      "step": 13783
+    },
+    {
+      "epoch": 0.220544,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.857580645161291e-05,
+      "loss": 0.1497,
+      "step": 13784
+    },
+    {
+      "epoch": 0.22056,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.85741935483871e-05,
+      "loss": 0.1542,
+      "step": 13785
+    },
+    {
+      "epoch": 0.220576,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.85725806451613e-05,
+      "loss": 0.1601,
+      "step": 13786
+    },
+    {
+      "epoch": 0.220592,
+      "grad_norm": 1.421875,
+      "learning_rate": 7.857096774193548e-05,
+      "loss": 0.1304,
+      "step": 13787
+    },
+    {
+      "epoch": 0.220608,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.856935483870968e-05,
+      "loss": 0.1466,
+      "step": 13788
+    },
+    {
+      "epoch": 0.220624,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.856774193548387e-05,
+      "loss": 0.1917,
+      "step": 13789
+    },
+    {
+      "epoch": 0.22064,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.856612903225807e-05,
+      "loss": 0.144,
+      "step": 13790
+    },
+    {
+      "epoch": 0.220656,
+      "grad_norm": 0.875,
+      "learning_rate": 7.856451612903226e-05,
+      "loss": 0.173,
+      "step": 13791
+    },
+    {
+      "epoch": 0.220672,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.856290322580646e-05,
+      "loss": 0.2007,
+      "step": 13792
+    },
+    {
+      "epoch": 0.220688,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.856129032258066e-05,
+      "loss": 0.1532,
+      "step": 13793
+    },
+    {
+      "epoch": 0.220704,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.855967741935484e-05,
+      "loss": 0.2043,
+      "step": 13794
+    },
+    {
+      "epoch": 0.22072,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.855806451612904e-05,
+      "loss": 0.1998,
+      "step": 13795
+    },
+    {
+      "epoch": 0.220736,
+      "grad_norm": 0.427734375,
+      "learning_rate": 7.855645161290323e-05,
+      "loss": 0.1315,
+      "step": 13796
+    },
+    {
+      "epoch": 0.220752,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.855483870967743e-05,
+      "loss": 0.1658,
+      "step": 13797
+    },
+    {
+      "epoch": 0.220768,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.855322580645161e-05,
+      "loss": 0.1675,
+      "step": 13798
+    },
+    {
+      "epoch": 0.220784,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.855161290322581e-05,
+      "loss": 0.1634,
+      "step": 13799
+    },
+    {
+      "epoch": 0.2208,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.855e-05,
+      "loss": 0.1416,
+      "step": 13800
+    },
+    {
+      "epoch": 0.220816,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.85483870967742e-05,
+      "loss": 0.1501,
+      "step": 13801
+    },
+    {
+      "epoch": 0.220832,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.854677419354838e-05,
+      "loss": 0.1489,
+      "step": 13802
+    },
+    {
+      "epoch": 0.220848,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.854516129032258e-05,
+      "loss": 0.1358,
+      "step": 13803
+    },
+    {
+      "epoch": 0.220864,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.854354838709678e-05,
+      "loss": 0.1448,
+      "step": 13804
+    },
+    {
+      "epoch": 0.22088,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.854193548387098e-05,
+      "loss": 0.1943,
+      "step": 13805
+    },
+    {
+      "epoch": 0.220896,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.854032258064517e-05,
+      "loss": 0.1665,
+      "step": 13806
+    },
+    {
+      "epoch": 0.220912,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.853870967741936e-05,
+      "loss": 0.141,
+      "step": 13807
+    },
+    {
+      "epoch": 0.220928,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.853709677419355e-05,
+      "loss": 0.1886,
+      "step": 13808
+    },
+    {
+      "epoch": 0.220944,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.853548387096774e-05,
+      "loss": 0.1934,
+      "step": 13809
+    },
+    {
+      "epoch": 0.22096,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.853387096774194e-05,
+      "loss": 0.1589,
+      "step": 13810
+    },
+    {
+      "epoch": 0.220976,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.853225806451613e-05,
+      "loss": 0.1745,
+      "step": 13811
+    },
+    {
+      "epoch": 0.220992,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.853064516129033e-05,
+      "loss": 0.1616,
+      "step": 13812
+    },
+    {
+      "epoch": 0.221008,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.852903225806451e-05,
+      "loss": 0.1987,
+      "step": 13813
+    },
+    {
+      "epoch": 0.221024,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.852741935483871e-05,
+      "loss": 0.145,
+      "step": 13814
+    },
+    {
+      "epoch": 0.22104,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.85258064516129e-05,
+      "loss": 0.182,
+      "step": 13815
+    },
+    {
+      "epoch": 0.221056,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.85241935483871e-05,
+      "loss": 0.1285,
+      "step": 13816
+    },
+    {
+      "epoch": 0.221072,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.85225806451613e-05,
+      "loss": 0.2031,
+      "step": 13817
+    },
+    {
+      "epoch": 0.221088,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.85209677419355e-05,
+      "loss": 0.1414,
+      "step": 13818
+    },
+    {
+      "epoch": 0.221104,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.851935483870968e-05,
+      "loss": 0.2027,
+      "step": 13819
+    },
+    {
+      "epoch": 0.22112,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.851774193548388e-05,
+      "loss": 0.1866,
+      "step": 13820
+    },
+    {
+      "epoch": 0.221136,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.851612903225807e-05,
+      "loss": 0.1331,
+      "step": 13821
+    },
+    {
+      "epoch": 0.221152,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.851451612903227e-05,
+      "loss": 0.1885,
+      "step": 13822
+    },
+    {
+      "epoch": 0.221168,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.851290322580645e-05,
+      "loss": 0.1815,
+      "step": 13823
+    },
+    {
+      "epoch": 0.221184,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.851129032258064e-05,
+      "loss": 0.1629,
+      "step": 13824
+    },
+    {
+      "epoch": 0.2212,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.850967741935484e-05,
+      "loss": 0.2136,
+      "step": 13825
+    },
+    {
+      "epoch": 0.221216,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.850806451612903e-05,
+      "loss": 0.1854,
+      "step": 13826
+    },
+    {
+      "epoch": 0.221232,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.850645161290323e-05,
+      "loss": 0.1868,
+      "step": 13827
+    },
+    {
+      "epoch": 0.221248,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.850483870967743e-05,
+      "loss": 0.1582,
+      "step": 13828
+    },
+    {
+      "epoch": 0.221264,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.850322580645163e-05,
+      "loss": 0.1747,
+      "step": 13829
+    },
+    {
+      "epoch": 0.22128,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.850161290322581e-05,
+      "loss": 0.1885,
+      "step": 13830
+    },
+    {
+      "epoch": 0.221296,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.850000000000001e-05,
+      "loss": 0.1308,
+      "step": 13831
+    },
+    {
+      "epoch": 0.221312,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.84983870967742e-05,
+      "loss": 0.2112,
+      "step": 13832
+    },
+    {
+      "epoch": 0.221328,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.84967741935484e-05,
+      "loss": 0.2,
+      "step": 13833
+    },
+    {
+      "epoch": 0.221344,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.849516129032258e-05,
+      "loss": 0.1804,
+      "step": 13834
+    },
+    {
+      "epoch": 0.22136,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.849354838709678e-05,
+      "loss": 0.2094,
+      "step": 13835
+    },
+    {
+      "epoch": 0.221376,
+      "grad_norm": 0.478515625,
+      "learning_rate": 7.849193548387097e-05,
+      "loss": 0.1471,
+      "step": 13836
+    },
+    {
+      "epoch": 0.221392,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.849032258064517e-05,
+      "loss": 0.1596,
+      "step": 13837
+    },
+    {
+      "epoch": 0.221408,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.848870967741935e-05,
+      "loss": 0.1729,
+      "step": 13838
+    },
+    {
+      "epoch": 0.221424,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.848709677419355e-05,
+      "loss": 0.2077,
+      "step": 13839
+    },
+    {
+      "epoch": 0.22144,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.848548387096775e-05,
+      "loss": 0.1662,
+      "step": 13840
+    },
+    {
+      "epoch": 0.221456,
+      "grad_norm": 0.75,
+      "learning_rate": 7.848387096774194e-05,
+      "loss": 0.1915,
+      "step": 13841
+    },
+    {
+      "epoch": 0.221472,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.848225806451614e-05,
+      "loss": 0.1341,
+      "step": 13842
+    },
+    {
+      "epoch": 0.221488,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.848064516129032e-05,
+      "loss": 0.1901,
+      "step": 13843
+    },
+    {
+      "epoch": 0.221504,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.847903225806452e-05,
+      "loss": 0.161,
+      "step": 13844
+    },
+    {
+      "epoch": 0.22152,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.847741935483871e-05,
+      "loss": 0.1723,
+      "step": 13845
+    },
+    {
+      "epoch": 0.221536,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.847580645161291e-05,
+      "loss": 0.1431,
+      "step": 13846
+    },
+    {
+      "epoch": 0.221552,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.84741935483871e-05,
+      "loss": 0.1808,
+      "step": 13847
+    },
+    {
+      "epoch": 0.221568,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.84725806451613e-05,
+      "loss": 0.1762,
+      "step": 13848
+    },
+    {
+      "epoch": 0.221584,
+      "grad_norm": 1.4296875,
+      "learning_rate": 7.847096774193548e-05,
+      "loss": 0.1545,
+      "step": 13849
+    },
+    {
+      "epoch": 0.2216,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.846935483870968e-05,
+      "loss": 0.171,
+      "step": 13850
+    },
+    {
+      "epoch": 0.221616,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.846774193548387e-05,
+      "loss": 0.1597,
+      "step": 13851
+    },
+    {
+      "epoch": 0.221632,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.846612903225807e-05,
+      "loss": 0.174,
+      "step": 13852
+    },
+    {
+      "epoch": 0.221648,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.846451612903227e-05,
+      "loss": 0.1873,
+      "step": 13853
+    },
+    {
+      "epoch": 0.221664,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.846290322580645e-05,
+      "loss": 0.1576,
+      "step": 13854
+    },
+    {
+      "epoch": 0.22168,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.846129032258065e-05,
+      "loss": 0.1841,
+      "step": 13855
+    },
+    {
+      "epoch": 0.221696,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.845967741935484e-05,
+      "loss": 0.1929,
+      "step": 13856
+    },
+    {
+      "epoch": 0.221712,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.845806451612904e-05,
+      "loss": 0.1198,
+      "step": 13857
+    },
+    {
+      "epoch": 0.221728,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.845645161290322e-05,
+      "loss": 0.1765,
+      "step": 13858
+    },
+    {
+      "epoch": 0.221744,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.845483870967742e-05,
+      "loss": 0.172,
+      "step": 13859
+    },
+    {
+      "epoch": 0.22176,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.845322580645161e-05,
+      "loss": 0.1603,
+      "step": 13860
+    },
+    {
+      "epoch": 0.221776,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.845161290322581e-05,
+      "loss": 0.1336,
+      "step": 13861
+    },
+    {
+      "epoch": 0.221792,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.845e-05,
+      "loss": 0.198,
+      "step": 13862
+    },
+    {
+      "epoch": 0.221808,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.84483870967742e-05,
+      "loss": 0.1794,
+      "step": 13863
+    },
+    {
+      "epoch": 0.221824,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.84467741935484e-05,
+      "loss": 0.1504,
+      "step": 13864
+    },
+    {
+      "epoch": 0.22184,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.84451612903226e-05,
+      "loss": 0.1582,
+      "step": 13865
+    },
+    {
+      "epoch": 0.221856,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.844354838709678e-05,
+      "loss": 0.194,
+      "step": 13866
+    },
+    {
+      "epoch": 0.221872,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.844193548387098e-05,
+      "loss": 0.1748,
+      "step": 13867
+    },
+    {
+      "epoch": 0.221888,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.844032258064517e-05,
+      "loss": 0.1711,
+      "step": 13868
+    },
+    {
+      "epoch": 0.221904,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.843870967741935e-05,
+      "loss": 0.1193,
+      "step": 13869
+    },
+    {
+      "epoch": 0.22192,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.843709677419355e-05,
+      "loss": 0.2255,
+      "step": 13870
+    },
+    {
+      "epoch": 0.221936,
+      "grad_norm": 0.875,
+      "learning_rate": 7.843548387096774e-05,
+      "loss": 0.1691,
+      "step": 13871
+    },
+    {
+      "epoch": 0.221952,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.843387096774194e-05,
+      "loss": 0.1816,
+      "step": 13872
+    },
+    {
+      "epoch": 0.221968,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.843225806451612e-05,
+      "loss": 0.1358,
+      "step": 13873
+    },
+    {
+      "epoch": 0.221984,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.843064516129032e-05,
+      "loss": 0.1748,
+      "step": 13874
+    },
+    {
+      "epoch": 0.222,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.842903225806451e-05,
+      "loss": 0.1876,
+      "step": 13875
+    },
+    {
+      "epoch": 0.222016,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.842741935483871e-05,
+      "loss": 0.15,
+      "step": 13876
+    },
+    {
+      "epoch": 0.222032,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.842580645161291e-05,
+      "loss": 0.201,
+      "step": 13877
+    },
+    {
+      "epoch": 0.222048,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.842419354838711e-05,
+      "loss": 0.1734,
+      "step": 13878
+    },
+    {
+      "epoch": 0.222064,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.84225806451613e-05,
+      "loss": 0.1555,
+      "step": 13879
+    },
+    {
+      "epoch": 0.22208,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.84209677419355e-05,
+      "loss": 0.185,
+      "step": 13880
+    },
+    {
+      "epoch": 0.222096,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.841935483870968e-05,
+      "loss": 0.1918,
+      "step": 13881
+    },
+    {
+      "epoch": 0.222112,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.841774193548388e-05,
+      "loss": 0.1576,
+      "step": 13882
+    },
+    {
+      "epoch": 0.222128,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.841612903225807e-05,
+      "loss": 0.1649,
+      "step": 13883
+    },
+    {
+      "epoch": 0.222144,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.841451612903227e-05,
+      "loss": 0.1377,
+      "step": 13884
+    },
+    {
+      "epoch": 0.22216,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.841290322580645e-05,
+      "loss": 0.1809,
+      "step": 13885
+    },
+    {
+      "epoch": 0.222176,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.841129032258064e-05,
+      "loss": 0.2055,
+      "step": 13886
+    },
+    {
+      "epoch": 0.222192,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.840967741935484e-05,
+      "loss": 0.1543,
+      "step": 13887
+    },
+    {
+      "epoch": 0.222208,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.840806451612904e-05,
+      "loss": 0.1567,
+      "step": 13888
+    },
+    {
+      "epoch": 0.222224,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.840645161290324e-05,
+      "loss": 0.1627,
+      "step": 13889
+    },
+    {
+      "epoch": 0.22224,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.840483870967742e-05,
+      "loss": 0.1856,
+      "step": 13890
+    },
+    {
+      "epoch": 0.222256,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.840322580645162e-05,
+      "loss": 0.1729,
+      "step": 13891
+    },
+    {
+      "epoch": 0.222272,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.840161290322581e-05,
+      "loss": 0.189,
+      "step": 13892
+    },
+    {
+      "epoch": 0.222288,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.840000000000001e-05,
+      "loss": 0.1269,
+      "step": 13893
+    },
+    {
+      "epoch": 0.222304,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.83983870967742e-05,
+      "loss": 0.1548,
+      "step": 13894
+    },
+    {
+      "epoch": 0.22232,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.83967741935484e-05,
+      "loss": 0.2006,
+      "step": 13895
+    },
+    {
+      "epoch": 0.222336,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.839516129032258e-05,
+      "loss": 0.1537,
+      "step": 13896
+    },
+    {
+      "epoch": 0.222352,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.839354838709678e-05,
+      "loss": 0.1593,
+      "step": 13897
+    },
+    {
+      "epoch": 0.222368,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.839193548387097e-05,
+      "loss": 0.1595,
+      "step": 13898
+    },
+    {
+      "epoch": 0.222384,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.839032258064517e-05,
+      "loss": 0.1567,
+      "step": 13899
+    },
+    {
+      "epoch": 0.2224,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.838870967741937e-05,
+      "loss": 0.1924,
+      "step": 13900
+    },
+    {
+      "epoch": 0.222416,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.838709677419355e-05,
+      "loss": 0.1893,
+      "step": 13901
+    },
+    {
+      "epoch": 0.222432,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.838548387096775e-05,
+      "loss": 0.222,
+      "step": 13902
+    },
+    {
+      "epoch": 0.222448,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.838387096774194e-05,
+      "loss": 0.1665,
+      "step": 13903
+    },
+    {
+      "epoch": 0.222464,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.838225806451614e-05,
+      "loss": 0.1742,
+      "step": 13904
+    },
+    {
+      "epoch": 0.22248,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.838064516129032e-05,
+      "loss": 0.2009,
+      "step": 13905
+    },
+    {
+      "epoch": 0.222496,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.837903225806452e-05,
+      "loss": 0.1258,
+      "step": 13906
+    },
+    {
+      "epoch": 0.222512,
+      "grad_norm": 1.2421875,
+      "learning_rate": 7.837741935483871e-05,
+      "loss": 0.168,
+      "step": 13907
+    },
+    {
+      "epoch": 0.222528,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.837580645161291e-05,
+      "loss": 0.1838,
+      "step": 13908
+    },
+    {
+      "epoch": 0.222544,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.83741935483871e-05,
+      "loss": 0.1871,
+      "step": 13909
+    },
+    {
+      "epoch": 0.22256,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.83725806451613e-05,
+      "loss": 0.1643,
+      "step": 13910
+    },
+    {
+      "epoch": 0.222576,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.837096774193548e-05,
+      "loss": 0.159,
+      "step": 13911
+    },
+    {
+      "epoch": 0.222592,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.836935483870968e-05,
+      "loss": 0.1533,
+      "step": 13912
+    },
+    {
+      "epoch": 0.222608,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.836774193548388e-05,
+      "loss": 0.226,
+      "step": 13913
+    },
+    {
+      "epoch": 0.222624,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.836612903225808e-05,
+      "loss": 0.2264,
+      "step": 13914
+    },
+    {
+      "epoch": 0.22264,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.836451612903226e-05,
+      "loss": 0.189,
+      "step": 13915
+    },
+    {
+      "epoch": 0.222656,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.836290322580645e-05,
+      "loss": 0.1651,
+      "step": 13916
+    },
+    {
+      "epoch": 0.222672,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.836129032258065e-05,
+      "loss": 0.1513,
+      "step": 13917
+    },
+    {
+      "epoch": 0.222688,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.835967741935484e-05,
+      "loss": 0.1777,
+      "step": 13918
+    },
+    {
+      "epoch": 0.222704,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.835806451612904e-05,
+      "loss": 0.1591,
+      "step": 13919
+    },
+    {
+      "epoch": 0.22272,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.835645161290322e-05,
+      "loss": 0.1864,
+      "step": 13920
+    },
+    {
+      "epoch": 0.222736,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.835483870967742e-05,
+      "loss": 0.1739,
+      "step": 13921
+    },
+    {
+      "epoch": 0.222752,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.835322580645161e-05,
+      "loss": 0.2232,
+      "step": 13922
+    },
+    {
+      "epoch": 0.222768,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.835161290322581e-05,
+      "loss": 0.1643,
+      "step": 13923
+    },
+    {
+      "epoch": 0.222784,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.835000000000001e-05,
+      "loss": 0.2218,
+      "step": 13924
+    },
+    {
+      "epoch": 0.2228,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.83483870967742e-05,
+      "loss": 0.2278,
+      "step": 13925
+    },
+    {
+      "epoch": 0.222816,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.834677419354839e-05,
+      "loss": 0.2021,
+      "step": 13926
+    },
+    {
+      "epoch": 0.222832,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.834516129032259e-05,
+      "loss": 0.2093,
+      "step": 13927
+    },
+    {
+      "epoch": 0.222848,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.834354838709678e-05,
+      "loss": 0.1678,
+      "step": 13928
+    },
+    {
+      "epoch": 0.222864,
+      "grad_norm": 1.2734375,
+      "learning_rate": 7.834193548387098e-05,
+      "loss": 0.1766,
+      "step": 13929
+    },
+    {
+      "epoch": 0.22288,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.834032258064516e-05,
+      "loss": 0.1676,
+      "step": 13930
+    },
+    {
+      "epoch": 0.222896,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.833870967741936e-05,
+      "loss": 0.142,
+      "step": 13931
+    },
+    {
+      "epoch": 0.222912,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.833709677419355e-05,
+      "loss": 0.1583,
+      "step": 13932
+    },
+    {
+      "epoch": 0.222928,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.833548387096774e-05,
+      "loss": 0.1581,
+      "step": 13933
+    },
+    {
+      "epoch": 0.222944,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.833387096774194e-05,
+      "loss": 0.1878,
+      "step": 13934
+    },
+    {
+      "epoch": 0.22296,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.833225806451614e-05,
+      "loss": 0.1629,
+      "step": 13935
+    },
+    {
+      "epoch": 0.222976,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.833064516129032e-05,
+      "loss": 0.1574,
+      "step": 13936
+    },
+    {
+      "epoch": 0.222992,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.832903225806452e-05,
+      "loss": 0.1508,
+      "step": 13937
+    },
+    {
+      "epoch": 0.223008,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.832741935483872e-05,
+      "loss": 0.1426,
+      "step": 13938
+    },
+    {
+      "epoch": 0.223024,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.83258064516129e-05,
+      "loss": 0.1761,
+      "step": 13939
+    },
+    {
+      "epoch": 0.22304,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.83241935483871e-05,
+      "loss": 0.1337,
+      "step": 13940
+    },
+    {
+      "epoch": 0.223056,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.832258064516129e-05,
+      "loss": 0.2084,
+      "step": 13941
+    },
+    {
+      "epoch": 0.223072,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.832096774193549e-05,
+      "loss": 0.1724,
+      "step": 13942
+    },
+    {
+      "epoch": 0.223088,
+      "grad_norm": 1.328125,
+      "learning_rate": 7.831935483870968e-05,
+      "loss": 0.1487,
+      "step": 13943
+    },
+    {
+      "epoch": 0.223104,
+      "grad_norm": 0.625,
+      "learning_rate": 7.831774193548388e-05,
+      "loss": 0.1435,
+      "step": 13944
+    },
+    {
+      "epoch": 0.22312,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.831612903225806e-05,
+      "loss": 0.1932,
+      "step": 13945
+    },
+    {
+      "epoch": 0.223136,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.831451612903226e-05,
+      "loss": 0.1718,
+      "step": 13946
+    },
+    {
+      "epoch": 0.223152,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.831290322580645e-05,
+      "loss": 0.1781,
+      "step": 13947
+    },
+    {
+      "epoch": 0.223168,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.831129032258065e-05,
+      "loss": 0.1603,
+      "step": 13948
+    },
+    {
+      "epoch": 0.223184,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.830967741935485e-05,
+      "loss": 0.1589,
+      "step": 13949
+    },
+    {
+      "epoch": 0.2232,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.830806451612903e-05,
+      "loss": 0.2374,
+      "step": 13950
+    },
+    {
+      "epoch": 0.223216,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.830645161290323e-05,
+      "loss": 0.185,
+      "step": 13951
+    },
+    {
+      "epoch": 0.223232,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.830483870967742e-05,
+      "loss": 0.1773,
+      "step": 13952
+    },
+    {
+      "epoch": 0.223248,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.830322580645162e-05,
+      "loss": 0.1753,
+      "step": 13953
+    },
+    {
+      "epoch": 0.223264,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.83016129032258e-05,
+      "loss": 0.1733,
+      "step": 13954
+    },
+    {
+      "epoch": 0.22328,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.83e-05,
+      "loss": 0.1651,
+      "step": 13955
+    },
+    {
+      "epoch": 0.223296,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.829838709677419e-05,
+      "loss": 0.1787,
+      "step": 13956
+    },
+    {
+      "epoch": 0.223312,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.829677419354839e-05,
+      "loss": 0.1641,
+      "step": 13957
+    },
+    {
+      "epoch": 0.223328,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.829516129032258e-05,
+      "loss": 0.1391,
+      "step": 13958
+    },
+    {
+      "epoch": 0.223344,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.829354838709678e-05,
+      "loss": 0.1514,
+      "step": 13959
+    },
+    {
+      "epoch": 0.22336,
+      "grad_norm": 1.296875,
+      "learning_rate": 7.829193548387098e-05,
+      "loss": 0.1934,
+      "step": 13960
+    },
+    {
+      "epoch": 0.223376,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.829032258064518e-05,
+      "loss": 0.2015,
+      "step": 13961
+    },
+    {
+      "epoch": 0.223392,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.828870967741936e-05,
+      "loss": 0.1808,
+      "step": 13962
+    },
+    {
+      "epoch": 0.223408,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.828709677419355e-05,
+      "loss": 0.1393,
+      "step": 13963
+    },
+    {
+      "epoch": 0.223424,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.828548387096775e-05,
+      "loss": 0.1767,
+      "step": 13964
+    },
+    {
+      "epoch": 0.22344,
+      "grad_norm": 1.125,
+      "learning_rate": 7.828387096774193e-05,
+      "loss": 0.1671,
+      "step": 13965
+    },
+    {
+      "epoch": 0.223456,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.828225806451613e-05,
+      "loss": 0.1545,
+      "step": 13966
+    },
+    {
+      "epoch": 0.223472,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.828064516129032e-05,
+      "loss": 0.2106,
+      "step": 13967
+    },
+    {
+      "epoch": 0.223488,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.827903225806452e-05,
+      "loss": 0.163,
+      "step": 13968
+    },
+    {
+      "epoch": 0.223504,
+      "grad_norm": 1.75,
+      "learning_rate": 7.82774193548387e-05,
+      "loss": 0.1573,
+      "step": 13969
+    },
+    {
+      "epoch": 0.22352,
+      "grad_norm": 0.75,
+      "learning_rate": 7.82758064516129e-05,
+      "loss": 0.1483,
+      "step": 13970
+    },
+    {
+      "epoch": 0.223536,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.827419354838709e-05,
+      "loss": 0.1455,
+      "step": 13971
+    },
+    {
+      "epoch": 0.223552,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.827258064516129e-05,
+      "loss": 0.1765,
+      "step": 13972
+    },
+    {
+      "epoch": 0.223568,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.827096774193549e-05,
+      "loss": 0.1542,
+      "step": 13973
+    },
+    {
+      "epoch": 0.223584,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.826935483870969e-05,
+      "loss": 0.1813,
+      "step": 13974
+    },
+    {
+      "epoch": 0.2236,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.826774193548388e-05,
+      "loss": 0.163,
+      "step": 13975
+    },
+    {
+      "epoch": 0.223616,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.826612903225808e-05,
+      "loss": 0.2137,
+      "step": 13976
+    },
+    {
+      "epoch": 0.223632,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.826451612903226e-05,
+      "loss": 0.2287,
+      "step": 13977
+    },
+    {
+      "epoch": 0.223648,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.826290322580645e-05,
+      "loss": 0.1746,
+      "step": 13978
+    },
+    {
+      "epoch": 0.223664,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.826129032258065e-05,
+      "loss": 0.1698,
+      "step": 13979
+    },
+    {
+      "epoch": 0.22368,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.825967741935483e-05,
+      "loss": 0.1693,
+      "step": 13980
+    },
+    {
+      "epoch": 0.223696,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.825806451612903e-05,
+      "loss": 0.1966,
+      "step": 13981
+    },
+    {
+      "epoch": 0.223712,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.825645161290322e-05,
+      "loss": 0.1635,
+      "step": 13982
+    },
+    {
+      "epoch": 0.223728,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.825483870967742e-05,
+      "loss": 0.1868,
+      "step": 13983
+    },
+    {
+      "epoch": 0.223744,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.825322580645162e-05,
+      "loss": 0.1567,
+      "step": 13984
+    },
+    {
+      "epoch": 0.22376,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.825161290322582e-05,
+      "loss": 0.1699,
+      "step": 13985
+    },
+    {
+      "epoch": 0.223776,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.825e-05,
+      "loss": 0.1655,
+      "step": 13986
+    },
+    {
+      "epoch": 0.223792,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.82483870967742e-05,
+      "loss": 0.1942,
+      "step": 13987
+    },
+    {
+      "epoch": 0.223808,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.824677419354839e-05,
+      "loss": 0.1829,
+      "step": 13988
+    },
+    {
+      "epoch": 0.223824,
+      "grad_norm": 1.390625,
+      "learning_rate": 7.824516129032259e-05,
+      "loss": 0.1769,
+      "step": 13989
+    },
+    {
+      "epoch": 0.22384,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.824354838709678e-05,
+      "loss": 0.1744,
+      "step": 13990
+    },
+    {
+      "epoch": 0.223856,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.824193548387098e-05,
+      "loss": 0.1623,
+      "step": 13991
+    },
+    {
+      "epoch": 0.223872,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.824032258064516e-05,
+      "loss": 0.1887,
+      "step": 13992
+    },
+    {
+      "epoch": 0.223888,
+      "grad_norm": 1.125,
+      "learning_rate": 7.823870967741936e-05,
+      "loss": 0.1294,
+      "step": 13993
+    },
+    {
+      "epoch": 0.223904,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.823709677419355e-05,
+      "loss": 0.162,
+      "step": 13994
+    },
+    {
+      "epoch": 0.22392,
+      "grad_norm": 0.875,
+      "learning_rate": 7.823548387096775e-05,
+      "loss": 0.1498,
+      "step": 13995
+    },
+    {
+      "epoch": 0.223936,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.823387096774195e-05,
+      "loss": 0.1688,
+      "step": 13996
+    },
+    {
+      "epoch": 0.223952,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.823225806451613e-05,
+      "loss": 0.155,
+      "step": 13997
+    },
+    {
+      "epoch": 0.223968,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.823064516129033e-05,
+      "loss": 0.1933,
+      "step": 13998
+    },
+    {
+      "epoch": 0.223984,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.822903225806452e-05,
+      "loss": 0.1853,
+      "step": 13999
+    },
+    {
+      "epoch": 0.224,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.822741935483872e-05,
+      "loss": 0.1387,
+      "step": 14000
+    },
+    {
+      "epoch": 0.224016,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.82258064516129e-05,
+      "loss": 0.1736,
+      "step": 14001
+    },
+    {
+      "epoch": 0.224032,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.82241935483871e-05,
+      "loss": 0.198,
+      "step": 14002
+    },
+    {
+      "epoch": 0.224048,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.822258064516129e-05,
+      "loss": 0.1641,
+      "step": 14003
+    },
+    {
+      "epoch": 0.224064,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.822096774193549e-05,
+      "loss": 0.1551,
+      "step": 14004
+    },
+    {
+      "epoch": 0.22408,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.821935483870968e-05,
+      "loss": 0.1426,
+      "step": 14005
+    },
+    {
+      "epoch": 0.224096,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.821774193548388e-05,
+      "loss": 0.1415,
+      "step": 14006
+    },
+    {
+      "epoch": 0.224112,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.821612903225806e-05,
+      "loss": 0.1962,
+      "step": 14007
+    },
+    {
+      "epoch": 0.224128,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.821451612903226e-05,
+      "loss": 0.1439,
+      "step": 14008
+    },
+    {
+      "epoch": 0.224144,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.821290322580646e-05,
+      "loss": 0.2072,
+      "step": 14009
+    },
+    {
+      "epoch": 0.22416,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.821129032258065e-05,
+      "loss": 0.1742,
+      "step": 14010
+    },
+    {
+      "epoch": 0.224176,
+      "grad_norm": 1.765625,
+      "learning_rate": 7.820967741935485e-05,
+      "loss": 0.2172,
+      "step": 14011
+    },
+    {
+      "epoch": 0.224192,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.820806451612903e-05,
+      "loss": 0.1714,
+      "step": 14012
+    },
+    {
+      "epoch": 0.224208,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.820645161290323e-05,
+      "loss": 0.1929,
+      "step": 14013
+    },
+    {
+      "epoch": 0.224224,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.820483870967742e-05,
+      "loss": 0.1603,
+      "step": 14014
+    },
+    {
+      "epoch": 0.22424,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.820322580645162e-05,
+      "loss": 0.1753,
+      "step": 14015
+    },
+    {
+      "epoch": 0.224256,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.82016129032258e-05,
+      "loss": 0.1853,
+      "step": 14016
+    },
+    {
+      "epoch": 0.224272,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.82e-05,
+      "loss": 0.1863,
+      "step": 14017
+    },
+    {
+      "epoch": 0.224288,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.819838709677419e-05,
+      "loss": 0.1529,
+      "step": 14018
+    },
+    {
+      "epoch": 0.224304,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.819677419354839e-05,
+      "loss": 0.1495,
+      "step": 14019
+    },
+    {
+      "epoch": 0.22432,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.819516129032259e-05,
+      "loss": 0.1833,
+      "step": 14020
+    },
+    {
+      "epoch": 0.224336,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.819354838709679e-05,
+      "loss": 0.1779,
+      "step": 14021
+    },
+    {
+      "epoch": 0.224352,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.819193548387097e-05,
+      "loss": 0.1807,
+      "step": 14022
+    },
+    {
+      "epoch": 0.224368,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.819032258064517e-05,
+      "loss": 0.1421,
+      "step": 14023
+    },
+    {
+      "epoch": 0.224384,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.818870967741936e-05,
+      "loss": 0.1828,
+      "step": 14024
+    },
+    {
+      "epoch": 0.2244,
+      "grad_norm": 0.75,
+      "learning_rate": 7.818709677419355e-05,
+      "loss": 0.1885,
+      "step": 14025
+    },
+    {
+      "epoch": 0.224416,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.818548387096775e-05,
+      "loss": 0.1664,
+      "step": 14026
+    },
+    {
+      "epoch": 0.224432,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.818387096774193e-05,
+      "loss": 0.1521,
+      "step": 14027
+    },
+    {
+      "epoch": 0.224448,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.818225806451613e-05,
+      "loss": 0.2118,
+      "step": 14028
+    },
+    {
+      "epoch": 0.224464,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.818064516129032e-05,
+      "loss": 0.1807,
+      "step": 14029
+    },
+    {
+      "epoch": 0.22448,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.817903225806452e-05,
+      "loss": 0.1512,
+      "step": 14030
+    },
+    {
+      "epoch": 0.224496,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.81774193548387e-05,
+      "loss": 0.172,
+      "step": 14031
+    },
+    {
+      "epoch": 0.224512,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.81758064516129e-05,
+      "loss": 0.1426,
+      "step": 14032
+    },
+    {
+      "epoch": 0.224528,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.81741935483871e-05,
+      "loss": 0.1775,
+      "step": 14033
+    },
+    {
+      "epoch": 0.224544,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.81725806451613e-05,
+      "loss": 0.1846,
+      "step": 14034
+    },
+    {
+      "epoch": 0.22456,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.817096774193549e-05,
+      "loss": 0.1547,
+      "step": 14035
+    },
+    {
+      "epoch": 0.224576,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.816935483870969e-05,
+      "loss": 0.1517,
+      "step": 14036
+    },
+    {
+      "epoch": 0.224592,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.816774193548387e-05,
+      "loss": 0.1413,
+      "step": 14037
+    },
+    {
+      "epoch": 0.224608,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.816612903225807e-05,
+      "loss": 0.1774,
+      "step": 14038
+    },
+    {
+      "epoch": 0.224624,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.816451612903226e-05,
+      "loss": 0.1903,
+      "step": 14039
+    },
+    {
+      "epoch": 0.22464,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.816290322580646e-05,
+      "loss": 0.1739,
+      "step": 14040
+    },
+    {
+      "epoch": 0.224656,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.816129032258065e-05,
+      "loss": 0.2104,
+      "step": 14041
+    },
+    {
+      "epoch": 0.224672,
+      "grad_norm": 0.875,
+      "learning_rate": 7.815967741935483e-05,
+      "loss": 0.181,
+      "step": 14042
+    },
+    {
+      "epoch": 0.224688,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.815806451612903e-05,
+      "loss": 0.1758,
+      "step": 14043
+    },
+    {
+      "epoch": 0.224704,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.815645161290323e-05,
+      "loss": 0.162,
+      "step": 14044
+    },
+    {
+      "epoch": 0.22472,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.815483870967743e-05,
+      "loss": 0.1824,
+      "step": 14045
+    },
+    {
+      "epoch": 0.224736,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.815322580645162e-05,
+      "loss": 0.1684,
+      "step": 14046
+    },
+    {
+      "epoch": 0.224752,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.815161290322582e-05,
+      "loss": 0.2358,
+      "step": 14047
+    },
+    {
+      "epoch": 0.224768,
+      "grad_norm": 0.625,
+      "learning_rate": 7.815e-05,
+      "loss": 0.1611,
+      "step": 14048
+    },
+    {
+      "epoch": 0.224784,
+      "grad_norm": 1.453125,
+      "learning_rate": 7.81483870967742e-05,
+      "loss": 0.2033,
+      "step": 14049
+    },
+    {
+      "epoch": 0.2248,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.814677419354839e-05,
+      "loss": 0.1923,
+      "step": 14050
+    },
+    {
+      "epoch": 0.224816,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.814516129032259e-05,
+      "loss": 0.1972,
+      "step": 14051
+    },
+    {
+      "epoch": 0.224832,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.814354838709677e-05,
+      "loss": 0.1639,
+      "step": 14052
+    },
+    {
+      "epoch": 0.224848,
+      "grad_norm": 1.4296875,
+      "learning_rate": 7.814193548387097e-05,
+      "loss": 0.1846,
+      "step": 14053
+    },
+    {
+      "epoch": 0.224864,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.814032258064516e-05,
+      "loss": 0.1634,
+      "step": 14054
+    },
+    {
+      "epoch": 0.22488,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.813870967741936e-05,
+      "loss": 0.1816,
+      "step": 14055
+    },
+    {
+      "epoch": 0.224896,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.813709677419356e-05,
+      "loss": 0.181,
+      "step": 14056
+    },
+    {
+      "epoch": 0.224912,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.813548387096774e-05,
+      "loss": 0.1471,
+      "step": 14057
+    },
+    {
+      "epoch": 0.224928,
+      "grad_norm": 0.486328125,
+      "learning_rate": 7.813387096774194e-05,
+      "loss": 0.1558,
+      "step": 14058
+    },
+    {
+      "epoch": 0.224944,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.813225806451613e-05,
+      "loss": 0.1537,
+      "step": 14059
+    },
+    {
+      "epoch": 0.22496,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.813064516129033e-05,
+      "loss": 0.1667,
+      "step": 14060
+    },
+    {
+      "epoch": 0.224976,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.812903225806452e-05,
+      "loss": 0.1887,
+      "step": 14061
+    },
+    {
+      "epoch": 0.224992,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.812741935483872e-05,
+      "loss": 0.188,
+      "step": 14062
+    },
+    {
+      "epoch": 0.225008,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.81258064516129e-05,
+      "loss": 0.1686,
+      "step": 14063
+    },
+    {
+      "epoch": 0.225024,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.81241935483871e-05,
+      "loss": 0.1894,
+      "step": 14064
+    },
+    {
+      "epoch": 0.22504,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.812258064516129e-05,
+      "loss": 0.1771,
+      "step": 14065
+    },
+    {
+      "epoch": 0.225056,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.812096774193549e-05,
+      "loss": 0.1514,
+      "step": 14066
+    },
+    {
+      "epoch": 0.225072,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.811935483870967e-05,
+      "loss": 0.2518,
+      "step": 14067
+    },
+    {
+      "epoch": 0.225088,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.811774193548387e-05,
+      "loss": 0.1362,
+      "step": 14068
+    },
+    {
+      "epoch": 0.225104,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.811612903225807e-05,
+      "loss": 0.2107,
+      "step": 14069
+    },
+    {
+      "epoch": 0.22512,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.811451612903227e-05,
+      "loss": 0.2027,
+      "step": 14070
+    },
+    {
+      "epoch": 0.225136,
+      "grad_norm": 0.4921875,
+      "learning_rate": 7.811290322580646e-05,
+      "loss": 0.1782,
+      "step": 14071
+    },
+    {
+      "epoch": 0.225152,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.811129032258064e-05,
+      "loss": 0.1502,
+      "step": 14072
+    },
+    {
+      "epoch": 0.225168,
+      "grad_norm": 1.125,
+      "learning_rate": 7.810967741935484e-05,
+      "loss": 0.1811,
+      "step": 14073
+    },
+    {
+      "epoch": 0.225184,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.810806451612903e-05,
+      "loss": 0.1768,
+      "step": 14074
+    },
+    {
+      "epoch": 0.2252,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.810645161290323e-05,
+      "loss": 0.1572,
+      "step": 14075
+    },
+    {
+      "epoch": 0.225216,
+      "grad_norm": 1.3359375,
+      "learning_rate": 7.810483870967742e-05,
+      "loss": 0.1875,
+      "step": 14076
+    },
+    {
+      "epoch": 0.225232,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.810322580645162e-05,
+      "loss": 0.1754,
+      "step": 14077
+    },
+    {
+      "epoch": 0.225248,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.81016129032258e-05,
+      "loss": 0.125,
+      "step": 14078
+    },
+    {
+      "epoch": 0.225264,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.81e-05,
+      "loss": 0.2091,
+      "step": 14079
+    },
+    {
+      "epoch": 0.22528,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.80983870967742e-05,
+      "loss": 0.193,
+      "step": 14080
+    },
+    {
+      "epoch": 0.225296,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.80967741935484e-05,
+      "loss": 0.1764,
+      "step": 14081
+    },
+    {
+      "epoch": 0.225312,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.809516129032259e-05,
+      "loss": 0.1149,
+      "step": 14082
+    },
+    {
+      "epoch": 0.225328,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.809354838709679e-05,
+      "loss": 0.172,
+      "step": 14083
+    },
+    {
+      "epoch": 0.225344,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.809193548387097e-05,
+      "loss": 0.1698,
+      "step": 14084
+    },
+    {
+      "epoch": 0.22536,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.809032258064517e-05,
+      "loss": 0.137,
+      "step": 14085
+    },
+    {
+      "epoch": 0.225376,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.808870967741936e-05,
+      "loss": 0.1196,
+      "step": 14086
+    },
+    {
+      "epoch": 0.225392,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.808709677419354e-05,
+      "loss": 0.1589,
+      "step": 14087
+    },
+    {
+      "epoch": 0.225408,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.808548387096774e-05,
+      "loss": 0.1654,
+      "step": 14088
+    },
+    {
+      "epoch": 0.225424,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.808387096774193e-05,
+      "loss": 0.2089,
+      "step": 14089
+    },
+    {
+      "epoch": 0.22544,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.808225806451613e-05,
+      "loss": 0.1744,
+      "step": 14090
+    },
+    {
+      "epoch": 0.225456,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.808064516129033e-05,
+      "loss": 0.2314,
+      "step": 14091
+    },
+    {
+      "epoch": 0.225472,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.807903225806453e-05,
+      "loss": 0.1615,
+      "step": 14092
+    },
+    {
+      "epoch": 0.225488,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.807741935483871e-05,
+      "loss": 0.1409,
+      "step": 14093
+    },
+    {
+      "epoch": 0.225504,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.807580645161291e-05,
+      "loss": 0.1678,
+      "step": 14094
+    },
+    {
+      "epoch": 0.22552,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.80741935483871e-05,
+      "loss": 0.1588,
+      "step": 14095
+    },
+    {
+      "epoch": 0.225536,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.80725806451613e-05,
+      "loss": 0.1546,
+      "step": 14096
+    },
+    {
+      "epoch": 0.225552,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.807096774193549e-05,
+      "loss": 0.1694,
+      "step": 14097
+    },
+    {
+      "epoch": 0.225568,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.806935483870969e-05,
+      "loss": 0.1621,
+      "step": 14098
+    },
+    {
+      "epoch": 0.225584,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.806774193548387e-05,
+      "loss": 0.1782,
+      "step": 14099
+    },
+    {
+      "epoch": 0.2256,
+      "grad_norm": 0.625,
+      "learning_rate": 7.806612903225807e-05,
+      "loss": 0.2007,
+      "step": 14100
+    },
+    {
+      "epoch": 0.225616,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.806451612903226e-05,
+      "loss": 0.1504,
+      "step": 14101
+    },
+    {
+      "epoch": 0.225632,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.806290322580646e-05,
+      "loss": 0.156,
+      "step": 14102
+    },
+    {
+      "epoch": 0.225648,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.806129032258064e-05,
+      "loss": 0.1531,
+      "step": 14103
+    },
+    {
+      "epoch": 0.225664,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.805967741935484e-05,
+      "loss": 0.1698,
+      "step": 14104
+    },
+    {
+      "epoch": 0.22568,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.805806451612904e-05,
+      "loss": 0.1732,
+      "step": 14105
+    },
+    {
+      "epoch": 0.225696,
+      "grad_norm": 1.484375,
+      "learning_rate": 7.805645161290323e-05,
+      "loss": 0.1876,
+      "step": 14106
+    },
+    {
+      "epoch": 0.225712,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.805483870967743e-05,
+      "loss": 0.1403,
+      "step": 14107
+    },
+    {
+      "epoch": 0.225728,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.805322580645161e-05,
+      "loss": 0.1709,
+      "step": 14108
+    },
+    {
+      "epoch": 0.225744,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.805161290322581e-05,
+      "loss": 0.1798,
+      "step": 14109
+    },
+    {
+      "epoch": 0.22576,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.805e-05,
+      "loss": 0.1563,
+      "step": 14110
+    },
+    {
+      "epoch": 0.225776,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.80483870967742e-05,
+      "loss": 0.1422,
+      "step": 14111
+    },
+    {
+      "epoch": 0.225792,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.804677419354839e-05,
+      "loss": 0.1296,
+      "step": 14112
+    },
+    {
+      "epoch": 0.225808,
+      "grad_norm": 1.0,
+      "learning_rate": 7.804516129032259e-05,
+      "loss": 0.1943,
+      "step": 14113
+    },
+    {
+      "epoch": 0.225824,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.804354838709677e-05,
+      "loss": 0.1701,
+      "step": 14114
+    },
+    {
+      "epoch": 0.22584,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.804193548387097e-05,
+      "loss": 0.1475,
+      "step": 14115
+    },
+    {
+      "epoch": 0.225856,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.804032258064517e-05,
+      "loss": 0.1934,
+      "step": 14116
+    },
+    {
+      "epoch": 0.225872,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.803870967741937e-05,
+      "loss": 0.1786,
+      "step": 14117
+    },
+    {
+      "epoch": 0.225888,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.803709677419356e-05,
+      "loss": 0.1308,
+      "step": 14118
+    },
+    {
+      "epoch": 0.225904,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.803548387096774e-05,
+      "loss": 0.183,
+      "step": 14119
+    },
+    {
+      "epoch": 0.22592,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.803387096774194e-05,
+      "loss": 0.1636,
+      "step": 14120
+    },
+    {
+      "epoch": 0.225936,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.803225806451613e-05,
+      "loss": 0.1825,
+      "step": 14121
+    },
+    {
+      "epoch": 0.225952,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.803064516129033e-05,
+      "loss": 0.1632,
+      "step": 14122
+    },
+    {
+      "epoch": 0.225968,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.802903225806451e-05,
+      "loss": 0.1555,
+      "step": 14123
+    },
+    {
+      "epoch": 0.225984,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.802741935483871e-05,
+      "loss": 0.1841,
+      "step": 14124
+    },
+    {
+      "epoch": 0.226,
+      "grad_norm": 0.625,
+      "learning_rate": 7.80258064516129e-05,
+      "loss": 0.1921,
+      "step": 14125
+    },
+    {
+      "epoch": 0.226016,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.80241935483871e-05,
+      "loss": 0.1827,
+      "step": 14126
+    },
+    {
+      "epoch": 0.226032,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.802258064516129e-05,
+      "loss": 0.1368,
+      "step": 14127
+    },
+    {
+      "epoch": 0.226048,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.802096774193548e-05,
+      "loss": 0.1815,
+      "step": 14128
+    },
+    {
+      "epoch": 0.226064,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.801935483870968e-05,
+      "loss": 0.1768,
+      "step": 14129
+    },
+    {
+      "epoch": 0.22608,
+      "grad_norm": 1.875,
+      "learning_rate": 7.801774193548388e-05,
+      "loss": 0.1963,
+      "step": 14130
+    },
+    {
+      "epoch": 0.226096,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.801612903225807e-05,
+      "loss": 0.1925,
+      "step": 14131
+    },
+    {
+      "epoch": 0.226112,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.801451612903227e-05,
+      "loss": 0.1846,
+      "step": 14132
+    },
+    {
+      "epoch": 0.226128,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.801290322580646e-05,
+      "loss": 0.1677,
+      "step": 14133
+    },
+    {
+      "epoch": 0.226144,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.801129032258064e-05,
+      "loss": 0.1817,
+      "step": 14134
+    },
+    {
+      "epoch": 0.22616,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.800967741935484e-05,
+      "loss": 0.2052,
+      "step": 14135
+    },
+    {
+      "epoch": 0.226176,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.800806451612903e-05,
+      "loss": 0.1814,
+      "step": 14136
+    },
+    {
+      "epoch": 0.226192,
+      "grad_norm": 0.490234375,
+      "learning_rate": 7.800645161290323e-05,
+      "loss": 0.1489,
+      "step": 14137
+    },
+    {
+      "epoch": 0.226208,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.800483870967741e-05,
+      "loss": 0.1589,
+      "step": 14138
+    },
+    {
+      "epoch": 0.226224,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.800322580645161e-05,
+      "loss": 0.14,
+      "step": 14139
+    },
+    {
+      "epoch": 0.22624,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.800161290322581e-05,
+      "loss": 0.1954,
+      "step": 14140
+    },
+    {
+      "epoch": 0.226256,
+      "grad_norm": 1.3984375,
+      "learning_rate": 7.800000000000001e-05,
+      "loss": 0.2287,
+      "step": 14141
+    },
+    {
+      "epoch": 0.226272,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.79983870967742e-05,
+      "loss": 0.2037,
+      "step": 14142
+    },
+    {
+      "epoch": 0.226288,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.79967741935484e-05,
+      "loss": 0.196,
+      "step": 14143
+    },
+    {
+      "epoch": 0.226304,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.799516129032258e-05,
+      "loss": 0.1334,
+      "step": 14144
+    },
+    {
+      "epoch": 0.22632,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.799354838709678e-05,
+      "loss": 0.1501,
+      "step": 14145
+    },
+    {
+      "epoch": 0.226336,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.799193548387097e-05,
+      "loss": 0.1438,
+      "step": 14146
+    },
+    {
+      "epoch": 0.226352,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.799032258064517e-05,
+      "loss": 0.1639,
+      "step": 14147
+    },
+    {
+      "epoch": 0.226368,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.798870967741936e-05,
+      "loss": 0.1672,
+      "step": 14148
+    },
+    {
+      "epoch": 0.226384,
+      "grad_norm": 0.478515625,
+      "learning_rate": 7.798709677419355e-05,
+      "loss": 0.1393,
+      "step": 14149
+    },
+    {
+      "epoch": 0.2264,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.798548387096774e-05,
+      "loss": 0.2443,
+      "step": 14150
+    },
+    {
+      "epoch": 0.226416,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.798387096774194e-05,
+      "loss": 0.1792,
+      "step": 14151
+    },
+    {
+      "epoch": 0.226432,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.798225806451614e-05,
+      "loss": 0.167,
+      "step": 14152
+    },
+    {
+      "epoch": 0.226448,
+      "grad_norm": 1.3125,
+      "learning_rate": 7.798064516129033e-05,
+      "loss": 0.1637,
+      "step": 14153
+    },
+    {
+      "epoch": 0.226464,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.797903225806453e-05,
+      "loss": 0.1558,
+      "step": 14154
+    },
+    {
+      "epoch": 0.22648,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.797741935483871e-05,
+      "loss": 0.185,
+      "step": 14155
+    },
+    {
+      "epoch": 0.226496,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.797580645161291e-05,
+      "loss": 0.1943,
+      "step": 14156
+    },
+    {
+      "epoch": 0.226512,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.79741935483871e-05,
+      "loss": 0.1838,
+      "step": 14157
+    },
+    {
+      "epoch": 0.226528,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.79725806451613e-05,
+      "loss": 0.2042,
+      "step": 14158
+    },
+    {
+      "epoch": 0.226544,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.797096774193548e-05,
+      "loss": 0.1801,
+      "step": 14159
+    },
+    {
+      "epoch": 0.22656,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.796935483870968e-05,
+      "loss": 0.1559,
+      "step": 14160
+    },
+    {
+      "epoch": 0.226576,
+      "grad_norm": 0.47265625,
+      "learning_rate": 7.796774193548387e-05,
+      "loss": 0.1794,
+      "step": 14161
+    },
+    {
+      "epoch": 0.226592,
+      "grad_norm": 1.4375,
+      "learning_rate": 7.796612903225807e-05,
+      "loss": 0.2072,
+      "step": 14162
+    },
+    {
+      "epoch": 0.226608,
+      "grad_norm": 1.2890625,
+      "learning_rate": 7.796451612903225e-05,
+      "loss": 0.1433,
+      "step": 14163
+    },
+    {
+      "epoch": 0.226624,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.796290322580645e-05,
+      "loss": 0.1675,
+      "step": 14164
+    },
+    {
+      "epoch": 0.22664,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.796129032258065e-05,
+      "loss": 0.1679,
+      "step": 14165
+    },
+    {
+      "epoch": 0.226656,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.795967741935484e-05,
+      "loss": 0.167,
+      "step": 14166
+    },
+    {
+      "epoch": 0.226672,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.795806451612904e-05,
+      "loss": 0.2036,
+      "step": 14167
+    },
+    {
+      "epoch": 0.226688,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.795645161290323e-05,
+      "loss": 0.177,
+      "step": 14168
+    },
+    {
+      "epoch": 0.226704,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.795483870967743e-05,
+      "loss": 0.1797,
+      "step": 14169
+    },
+    {
+      "epoch": 0.22672,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.795322580645161e-05,
+      "loss": 0.1847,
+      "step": 14170
+    },
+    {
+      "epoch": 0.226736,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.795161290322581e-05,
+      "loss": 0.196,
+      "step": 14171
+    },
+    {
+      "epoch": 0.226752,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.795e-05,
+      "loss": 0.134,
+      "step": 14172
+    },
+    {
+      "epoch": 0.226768,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.79483870967742e-05,
+      "loss": 0.1554,
+      "step": 14173
+    },
+    {
+      "epoch": 0.226784,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.794677419354838e-05,
+      "loss": 0.1927,
+      "step": 14174
+    },
+    {
+      "epoch": 0.2268,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.794516129032258e-05,
+      "loss": 0.1272,
+      "step": 14175
+    },
+    {
+      "epoch": 0.226816,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.794354838709678e-05,
+      "loss": 0.1385,
+      "step": 14176
+    },
+    {
+      "epoch": 0.226832,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.794193548387098e-05,
+      "loss": 0.1995,
+      "step": 14177
+    },
+    {
+      "epoch": 0.226848,
+      "grad_norm": 1.4609375,
+      "learning_rate": 7.794032258064517e-05,
+      "loss": 0.1801,
+      "step": 14178
+    },
+    {
+      "epoch": 0.226864,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.793870967741937e-05,
+      "loss": 0.1713,
+      "step": 14179
+    },
+    {
+      "epoch": 0.22688,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.793709677419355e-05,
+      "loss": 0.165,
+      "step": 14180
+    },
+    {
+      "epoch": 0.226896,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.793548387096774e-05,
+      "loss": 0.1959,
+      "step": 14181
+    },
+    {
+      "epoch": 0.226912,
+      "grad_norm": 0.486328125,
+      "learning_rate": 7.793387096774194e-05,
+      "loss": 0.1562,
+      "step": 14182
+    },
+    {
+      "epoch": 0.226928,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.793225806451613e-05,
+      "loss": 0.147,
+      "step": 14183
+    },
+    {
+      "epoch": 0.226944,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.793064516129033e-05,
+      "loss": 0.1478,
+      "step": 14184
+    },
+    {
+      "epoch": 0.22696,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.792903225806451e-05,
+      "loss": 0.1662,
+      "step": 14185
+    },
+    {
+      "epoch": 0.226976,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.792741935483871e-05,
+      "loss": 0.1653,
+      "step": 14186
+    },
+    {
+      "epoch": 0.226992,
+      "grad_norm": 1.453125,
+      "learning_rate": 7.792580645161291e-05,
+      "loss": 0.212,
+      "step": 14187
+    },
+    {
+      "epoch": 0.227008,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.79241935483871e-05,
+      "loss": 0.1748,
+      "step": 14188
+    },
+    {
+      "epoch": 0.227024,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.79225806451613e-05,
+      "loss": 0.1598,
+      "step": 14189
+    },
+    {
+      "epoch": 0.22704,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.79209677419355e-05,
+      "loss": 0.2103,
+      "step": 14190
+    },
+    {
+      "epoch": 0.227056,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.791935483870968e-05,
+      "loss": 0.1944,
+      "step": 14191
+    },
+    {
+      "epoch": 0.227072,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.791774193548388e-05,
+      "loss": 0.1461,
+      "step": 14192
+    },
+    {
+      "epoch": 0.227088,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.791612903225807e-05,
+      "loss": 0.1801,
+      "step": 14193
+    },
+    {
+      "epoch": 0.227104,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.791451612903227e-05,
+      "loss": 0.178,
+      "step": 14194
+    },
+    {
+      "epoch": 0.22712,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.791290322580645e-05,
+      "loss": 0.167,
+      "step": 14195
+    },
+    {
+      "epoch": 0.227136,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.791129032258064e-05,
+      "loss": 0.1788,
+      "step": 14196
+    },
+    {
+      "epoch": 0.227152,
+      "grad_norm": 1.4375,
+      "learning_rate": 7.790967741935484e-05,
+      "loss": 0.1695,
+      "step": 14197
+    },
+    {
+      "epoch": 0.227168,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.790806451612903e-05,
+      "loss": 0.1925,
+      "step": 14198
+    },
+    {
+      "epoch": 0.227184,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.790645161290322e-05,
+      "loss": 0.1696,
+      "step": 14199
+    },
+    {
+      "epoch": 0.2272,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.790483870967742e-05,
+      "loss": 0.1828,
+      "step": 14200
+    },
+    {
+      "epoch": 0.227216,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.790322580645162e-05,
+      "loss": 0.1934,
+      "step": 14201
+    },
+    {
+      "epoch": 0.227232,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.790161290322581e-05,
+      "loss": 0.2074,
+      "step": 14202
+    },
+    {
+      "epoch": 0.227248,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.790000000000001e-05,
+      "loss": 0.1638,
+      "step": 14203
+    },
+    {
+      "epoch": 0.227264,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.78983870967742e-05,
+      "loss": 0.1548,
+      "step": 14204
+    },
+    {
+      "epoch": 0.22728,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.78967741935484e-05,
+      "loss": 0.1819,
+      "step": 14205
+    },
+    {
+      "epoch": 0.227296,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.789516129032258e-05,
+      "loss": 0.1784,
+      "step": 14206
+    },
+    {
+      "epoch": 0.227312,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.789354838709678e-05,
+      "loss": 0.1881,
+      "step": 14207
+    },
+    {
+      "epoch": 0.227328,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.789193548387097e-05,
+      "loss": 0.2096,
+      "step": 14208
+    },
+    {
+      "epoch": 0.227344,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.789032258064517e-05,
+      "loss": 0.1583,
+      "step": 14209
+    },
+    {
+      "epoch": 0.22736,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.788870967741935e-05,
+      "loss": 0.1227,
+      "step": 14210
+    },
+    {
+      "epoch": 0.227376,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.788709677419355e-05,
+      "loss": 0.158,
+      "step": 14211
+    },
+    {
+      "epoch": 0.227392,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.788548387096775e-05,
+      "loss": 0.177,
+      "step": 14212
+    },
+    {
+      "epoch": 0.227408,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.788387096774194e-05,
+      "loss": 0.1608,
+      "step": 14213
+    },
+    {
+      "epoch": 0.227424,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.788225806451614e-05,
+      "loss": 0.1939,
+      "step": 14214
+    },
+    {
+      "epoch": 0.22744,
+      "grad_norm": 0.466796875,
+      "learning_rate": 7.788064516129032e-05,
+      "loss": 0.1206,
+      "step": 14215
+    },
+    {
+      "epoch": 0.227456,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.787903225806452e-05,
+      "loss": 0.1507,
+      "step": 14216
+    },
+    {
+      "epoch": 0.227472,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.787741935483871e-05,
+      "loss": 0.1401,
+      "step": 14217
+    },
+    {
+      "epoch": 0.227488,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.787580645161291e-05,
+      "loss": 0.1943,
+      "step": 14218
+    },
+    {
+      "epoch": 0.227504,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.78741935483871e-05,
+      "loss": 0.2018,
+      "step": 14219
+    },
+    {
+      "epoch": 0.22752,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.78725806451613e-05,
+      "loss": 0.1585,
+      "step": 14220
+    },
+    {
+      "epoch": 0.227536,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.787096774193548e-05,
+      "loss": 0.162,
+      "step": 14221
+    },
+    {
+      "epoch": 0.227552,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.786935483870968e-05,
+      "loss": 0.1648,
+      "step": 14222
+    },
+    {
+      "epoch": 0.227568,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.786774193548387e-05,
+      "loss": 0.1985,
+      "step": 14223
+    },
+    {
+      "epoch": 0.227584,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.786612903225807e-05,
+      "loss": 0.1331,
+      "step": 14224
+    },
+    {
+      "epoch": 0.2276,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.786451612903227e-05,
+      "loss": 0.1903,
+      "step": 14225
+    },
+    {
+      "epoch": 0.227616,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.786290322580647e-05,
+      "loss": 0.13,
+      "step": 14226
+    },
+    {
+      "epoch": 0.227632,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.786129032258065e-05,
+      "loss": 0.1961,
+      "step": 14227
+    },
+    {
+      "epoch": 0.227648,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.785967741935484e-05,
+      "loss": 0.168,
+      "step": 14228
+    },
+    {
+      "epoch": 0.227664,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.785806451612904e-05,
+      "loss": 0.1508,
+      "step": 14229
+    },
+    {
+      "epoch": 0.22768,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.785645161290322e-05,
+      "loss": 0.1943,
+      "step": 14230
+    },
+    {
+      "epoch": 0.227696,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.785483870967742e-05,
+      "loss": 0.1109,
+      "step": 14231
+    },
+    {
+      "epoch": 0.227712,
+      "grad_norm": 1.5234375,
+      "learning_rate": 7.785322580645161e-05,
+      "loss": 0.1904,
+      "step": 14232
+    },
+    {
+      "epoch": 0.227728,
+      "grad_norm": 1.34375,
+      "learning_rate": 7.785161290322581e-05,
+      "loss": 0.1691,
+      "step": 14233
+    },
+    {
+      "epoch": 0.227744,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.785e-05,
+      "loss": 0.1588,
+      "step": 14234
+    },
+    {
+      "epoch": 0.22776,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.78483870967742e-05,
+      "loss": 0.1909,
+      "step": 14235
+    },
+    {
+      "epoch": 0.227776,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.78467741935484e-05,
+      "loss": 0.1666,
+      "step": 14236
+    },
+    {
+      "epoch": 0.227792,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.78451612903226e-05,
+      "loss": 0.1641,
+      "step": 14237
+    },
+    {
+      "epoch": 0.227808,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.784354838709678e-05,
+      "loss": 0.177,
+      "step": 14238
+    },
+    {
+      "epoch": 0.227824,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.784193548387098e-05,
+      "loss": 0.1651,
+      "step": 14239
+    },
+    {
+      "epoch": 0.22784,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.784032258064517e-05,
+      "loss": 0.1815,
+      "step": 14240
+    },
+    {
+      "epoch": 0.227856,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.783870967741937e-05,
+      "loss": 0.2222,
+      "step": 14241
+    },
+    {
+      "epoch": 0.227872,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.783709677419355e-05,
+      "loss": 0.1632,
+      "step": 14242
+    },
+    {
+      "epoch": 0.227888,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.783548387096774e-05,
+      "loss": 0.1863,
+      "step": 14243
+    },
+    {
+      "epoch": 0.227904,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.783387096774194e-05,
+      "loss": 0.1665,
+      "step": 14244
+    },
+    {
+      "epoch": 0.22792,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.783225806451612e-05,
+      "loss": 0.1739,
+      "step": 14245
+    },
+    {
+      "epoch": 0.227936,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.783064516129032e-05,
+      "loss": 0.1946,
+      "step": 14246
+    },
+    {
+      "epoch": 0.227952,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.782903225806452e-05,
+      "loss": 0.1457,
+      "step": 14247
+    },
+    {
+      "epoch": 0.227968,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.782741935483872e-05,
+      "loss": 0.1215,
+      "step": 14248
+    },
+    {
+      "epoch": 0.227984,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.782580645161291e-05,
+      "loss": 0.2078,
+      "step": 14249
+    },
+    {
+      "epoch": 0.228,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.782419354838711e-05,
+      "loss": 0.1617,
+      "step": 14250
+    },
+    {
+      "epoch": 0.228016,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.78225806451613e-05,
+      "loss": 0.1545,
+      "step": 14251
+    },
+    {
+      "epoch": 0.228032,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.78209677419355e-05,
+      "loss": 0.1953,
+      "step": 14252
+    },
+    {
+      "epoch": 0.228048,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.781935483870968e-05,
+      "loss": 0.218,
+      "step": 14253
+    },
+    {
+      "epoch": 0.228064,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.781774193548388e-05,
+      "loss": 0.1332,
+      "step": 14254
+    },
+    {
+      "epoch": 0.22808,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.781612903225807e-05,
+      "loss": 0.1861,
+      "step": 14255
+    },
+    {
+      "epoch": 0.228096,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.781451612903226e-05,
+      "loss": 0.1538,
+      "step": 14256
+    },
+    {
+      "epoch": 0.228112,
+      "grad_norm": 1.375,
+      "learning_rate": 7.781290322580645e-05,
+      "loss": 0.2155,
+      "step": 14257
+    },
+    {
+      "epoch": 0.228128,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.781129032258065e-05,
+      "loss": 0.1899,
+      "step": 14258
+    },
+    {
+      "epoch": 0.228144,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.780967741935484e-05,
+      "loss": 0.1959,
+      "step": 14259
+    },
+    {
+      "epoch": 0.22816,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.780806451612904e-05,
+      "loss": 0.2133,
+      "step": 14260
+    },
+    {
+      "epoch": 0.228176,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.780645161290324e-05,
+      "loss": 0.1557,
+      "step": 14261
+    },
+    {
+      "epoch": 0.228192,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.780483870967742e-05,
+      "loss": 0.1264,
+      "step": 14262
+    },
+    {
+      "epoch": 0.228208,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.780322580645162e-05,
+      "loss": 0.1738,
+      "step": 14263
+    },
+    {
+      "epoch": 0.228224,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.780161290322581e-05,
+      "loss": 0.15,
+      "step": 14264
+    },
+    {
+      "epoch": 0.22824,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.780000000000001e-05,
+      "loss": 0.184,
+      "step": 14265
+    },
+    {
+      "epoch": 0.228256,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.77983870967742e-05,
+      "loss": 0.1602,
+      "step": 14266
+    },
+    {
+      "epoch": 0.228272,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.779677419354839e-05,
+      "loss": 0.179,
+      "step": 14267
+    },
+    {
+      "epoch": 0.228288,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.779516129032258e-05,
+      "loss": 0.2058,
+      "step": 14268
+    },
+    {
+      "epoch": 0.228304,
+      "grad_norm": 1.0,
+      "learning_rate": 7.779354838709678e-05,
+      "loss": 0.2038,
+      "step": 14269
+    },
+    {
+      "epoch": 0.22832,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.779193548387096e-05,
+      "loss": 0.1962,
+      "step": 14270
+    },
+    {
+      "epoch": 0.228336,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.779032258064516e-05,
+      "loss": 0.1421,
+      "step": 14271
+    },
+    {
+      "epoch": 0.228352,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.778870967741936e-05,
+      "loss": 0.168,
+      "step": 14272
+    },
+    {
+      "epoch": 0.228368,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.778709677419356e-05,
+      "loss": 0.198,
+      "step": 14273
+    },
+    {
+      "epoch": 0.228384,
+      "grad_norm": 1.2109375,
+      "learning_rate": 7.778548387096775e-05,
+      "loss": 0.1581,
+      "step": 14274
+    },
+    {
+      "epoch": 0.2284,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.778387096774194e-05,
+      "loss": 0.1818,
+      "step": 14275
+    },
+    {
+      "epoch": 0.228416,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.778225806451614e-05,
+      "loss": 0.1402,
+      "step": 14276
+    },
+    {
+      "epoch": 0.228432,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.778064516129032e-05,
+      "loss": 0.1908,
+      "step": 14277
+    },
+    {
+      "epoch": 0.228448,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.777903225806452e-05,
+      "loss": 0.1599,
+      "step": 14278
+    },
+    {
+      "epoch": 0.228464,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.777741935483871e-05,
+      "loss": 0.1675,
+      "step": 14279
+    },
+    {
+      "epoch": 0.22848,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.777580645161291e-05,
+      "loss": 0.1694,
+      "step": 14280
+    },
+    {
+      "epoch": 0.228496,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.777419354838709e-05,
+      "loss": 0.1571,
+      "step": 14281
+    },
+    {
+      "epoch": 0.228512,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.777258064516129e-05,
+      "loss": 0.1476,
+      "step": 14282
+    },
+    {
+      "epoch": 0.228528,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.777096774193548e-05,
+      "loss": 0.1911,
+      "step": 14283
+    },
+    {
+      "epoch": 0.228544,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.776935483870968e-05,
+      "loss": 0.1792,
+      "step": 14284
+    },
+    {
+      "epoch": 0.22856,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.776774193548388e-05,
+      "loss": 0.1673,
+      "step": 14285
+    },
+    {
+      "epoch": 0.228576,
+      "grad_norm": 1.296875,
+      "learning_rate": 7.776612903225808e-05,
+      "loss": 0.1706,
+      "step": 14286
+    },
+    {
+      "epoch": 0.228592,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.776451612903226e-05,
+      "loss": 0.1514,
+      "step": 14287
+    },
+    {
+      "epoch": 0.228608,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.776290322580646e-05,
+      "loss": 0.166,
+      "step": 14288
+    },
+    {
+      "epoch": 0.228624,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.776129032258065e-05,
+      "loss": 0.1581,
+      "step": 14289
+    },
+    {
+      "epoch": 0.22864,
+      "grad_norm": 0.46875,
+      "learning_rate": 7.775967741935484e-05,
+      "loss": 0.1272,
+      "step": 14290
+    },
+    {
+      "epoch": 0.228656,
+      "grad_norm": 1.0,
+      "learning_rate": 7.775806451612904e-05,
+      "loss": 0.2007,
+      "step": 14291
+    },
+    {
+      "epoch": 0.228672,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.775645161290322e-05,
+      "loss": 0.1894,
+      "step": 14292
+    },
+    {
+      "epoch": 0.228688,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.775483870967742e-05,
+      "loss": 0.2174,
+      "step": 14293
+    },
+    {
+      "epoch": 0.228704,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.775322580645161e-05,
+      "loss": 0.1782,
+      "step": 14294
+    },
+    {
+      "epoch": 0.22872,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.77516129032258e-05,
+      "loss": 0.1879,
+      "step": 14295
+    },
+    {
+      "epoch": 0.228736,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.775e-05,
+      "loss": 0.2214,
+      "step": 14296
+    },
+    {
+      "epoch": 0.228752,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.77483870967742e-05,
+      "loss": 0.1539,
+      "step": 14297
+    },
+    {
+      "epoch": 0.228768,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.774677419354839e-05,
+      "loss": 0.1923,
+      "step": 14298
+    },
+    {
+      "epoch": 0.228784,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.774516129032259e-05,
+      "loss": 0.1761,
+      "step": 14299
+    },
+    {
+      "epoch": 0.2288,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.774354838709678e-05,
+      "loss": 0.1868,
+      "step": 14300
+    },
+    {
+      "epoch": 0.228816,
+      "grad_norm": 1.2578125,
+      "learning_rate": 7.774193548387098e-05,
+      "loss": 0.1763,
+      "step": 14301
+    },
+    {
+      "epoch": 0.228832,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.774032258064516e-05,
+      "loss": 0.2278,
+      "step": 14302
+    },
+    {
+      "epoch": 0.228848,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.773870967741936e-05,
+      "loss": 0.1853,
+      "step": 14303
+    },
+    {
+      "epoch": 0.228864,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.773709677419355e-05,
+      "loss": 0.1989,
+      "step": 14304
+    },
+    {
+      "epoch": 0.22888,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.773548387096773e-05,
+      "loss": 0.1609,
+      "step": 14305
+    },
+    {
+      "epoch": 0.228896,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.773387096774193e-05,
+      "loss": 0.1663,
+      "step": 14306
+    },
+    {
+      "epoch": 0.228912,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.773225806451613e-05,
+      "loss": 0.1975,
+      "step": 14307
+    },
+    {
+      "epoch": 0.228928,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.773064516129033e-05,
+      "loss": 0.1622,
+      "step": 14308
+    },
+    {
+      "epoch": 0.228944,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.772903225806452e-05,
+      "loss": 0.1735,
+      "step": 14309
+    },
+    {
+      "epoch": 0.22896,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.772741935483872e-05,
+      "loss": 0.1774,
+      "step": 14310
+    },
+    {
+      "epoch": 0.228976,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.77258064516129e-05,
+      "loss": 0.1784,
+      "step": 14311
+    },
+    {
+      "epoch": 0.228992,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.77241935483871e-05,
+      "loss": 0.1961,
+      "step": 14312
+    },
+    {
+      "epoch": 0.229008,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.772258064516129e-05,
+      "loss": 0.1728,
+      "step": 14313
+    },
+    {
+      "epoch": 0.229024,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.772096774193549e-05,
+      "loss": 0.1714,
+      "step": 14314
+    },
+    {
+      "epoch": 0.22904,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.771935483870968e-05,
+      "loss": 0.1673,
+      "step": 14315
+    },
+    {
+      "epoch": 0.229056,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.771774193548388e-05,
+      "loss": 0.213,
+      "step": 14316
+    },
+    {
+      "epoch": 0.229072,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.771612903225806e-05,
+      "loss": 0.1285,
+      "step": 14317
+    },
+    {
+      "epoch": 0.229088,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.771451612903226e-05,
+      "loss": 0.1843,
+      "step": 14318
+    },
+    {
+      "epoch": 0.229104,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.771290322580645e-05,
+      "loss": 0.1552,
+      "step": 14319
+    },
+    {
+      "epoch": 0.22912,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.771129032258065e-05,
+      "loss": 0.1553,
+      "step": 14320
+    },
+    {
+      "epoch": 0.229136,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.770967741935485e-05,
+      "loss": 0.2047,
+      "step": 14321
+    },
+    {
+      "epoch": 0.229152,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.770806451612903e-05,
+      "loss": 0.1507,
+      "step": 14322
+    },
+    {
+      "epoch": 0.229168,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.770645161290323e-05,
+      "loss": 0.1504,
+      "step": 14323
+    },
+    {
+      "epoch": 0.229184,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.770483870967742e-05,
+      "loss": 0.1374,
+      "step": 14324
+    },
+    {
+      "epoch": 0.2292,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.770322580645162e-05,
+      "loss": 0.15,
+      "step": 14325
+    },
+    {
+      "epoch": 0.229216,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.77016129032258e-05,
+      "loss": 0.1563,
+      "step": 14326
+    },
+    {
+      "epoch": 0.229232,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.77e-05,
+      "loss": 0.1755,
+      "step": 14327
+    },
+    {
+      "epoch": 0.229248,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.769838709677419e-05,
+      "loss": 0.1813,
+      "step": 14328
+    },
+    {
+      "epoch": 0.229264,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.769677419354839e-05,
+      "loss": 0.1657,
+      "step": 14329
+    },
+    {
+      "epoch": 0.22928,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.769516129032258e-05,
+      "loss": 0.1289,
+      "step": 14330
+    },
+    {
+      "epoch": 0.229296,
+      "grad_norm": 0.625,
+      "learning_rate": 7.769354838709678e-05,
+      "loss": 0.1549,
+      "step": 14331
+    },
+    {
+      "epoch": 0.229312,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.769193548387098e-05,
+      "loss": 0.1958,
+      "step": 14332
+    },
+    {
+      "epoch": 0.229328,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.769032258064518e-05,
+      "loss": 0.1826,
+      "step": 14333
+    },
+    {
+      "epoch": 0.229344,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.768870967741936e-05,
+      "loss": 0.186,
+      "step": 14334
+    },
+    {
+      "epoch": 0.22936,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.768709677419356e-05,
+      "loss": 0.1738,
+      "step": 14335
+    },
+    {
+      "epoch": 0.229376,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.768548387096775e-05,
+      "loss": 0.1638,
+      "step": 14336
+    },
+    {
+      "epoch": 0.229392,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.768387096774193e-05,
+      "loss": 0.1375,
+      "step": 14337
+    },
+    {
+      "epoch": 0.229408,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.768225806451613e-05,
+      "loss": 0.1776,
+      "step": 14338
+    },
+    {
+      "epoch": 0.229424,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.768064516129032e-05,
+      "loss": 0.1605,
+      "step": 14339
+    },
+    {
+      "epoch": 0.22944,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.767903225806452e-05,
+      "loss": 0.1627,
+      "step": 14340
+    },
+    {
+      "epoch": 0.229456,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.76774193548387e-05,
+      "loss": 0.1528,
+      "step": 14341
+    },
+    {
+      "epoch": 0.229472,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.76758064516129e-05,
+      "loss": 0.2002,
+      "step": 14342
+    },
+    {
+      "epoch": 0.229488,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.76741935483871e-05,
+      "loss": 0.1832,
+      "step": 14343
+    },
+    {
+      "epoch": 0.229504,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.76725806451613e-05,
+      "loss": 0.1954,
+      "step": 14344
+    },
+    {
+      "epoch": 0.22952,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.767096774193549e-05,
+      "loss": 0.1299,
+      "step": 14345
+    },
+    {
+      "epoch": 0.229536,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.766935483870969e-05,
+      "loss": 0.1758,
+      "step": 14346
+    },
+    {
+      "epoch": 0.229552,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.766774193548388e-05,
+      "loss": 0.1336,
+      "step": 14347
+    },
+    {
+      "epoch": 0.229568,
+      "grad_norm": 1.2890625,
+      "learning_rate": 7.766612903225808e-05,
+      "loss": 0.1787,
+      "step": 14348
+    },
+    {
+      "epoch": 0.229584,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.766451612903226e-05,
+      "loss": 0.1977,
+      "step": 14349
+    },
+    {
+      "epoch": 0.2296,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.766290322580646e-05,
+      "loss": 0.1566,
+      "step": 14350
+    },
+    {
+      "epoch": 0.229616,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.766129032258065e-05,
+      "loss": 0.1367,
+      "step": 14351
+    },
+    {
+      "epoch": 0.229632,
+      "grad_norm": 1.3359375,
+      "learning_rate": 7.765967741935483e-05,
+      "loss": 0.1629,
+      "step": 14352
+    },
+    {
+      "epoch": 0.229648,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.765806451612903e-05,
+      "loss": 0.1736,
+      "step": 14353
+    },
+    {
+      "epoch": 0.229664,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.765645161290322e-05,
+      "loss": 0.1745,
+      "step": 14354
+    },
+    {
+      "epoch": 0.22968,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.765483870967742e-05,
+      "loss": 0.2298,
+      "step": 14355
+    },
+    {
+      "epoch": 0.229696,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.765322580645162e-05,
+      "loss": 0.1674,
+      "step": 14356
+    },
+    {
+      "epoch": 0.229712,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.765161290322582e-05,
+      "loss": 0.1637,
+      "step": 14357
+    },
+    {
+      "epoch": 0.229728,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.765e-05,
+      "loss": 0.1667,
+      "step": 14358
+    },
+    {
+      "epoch": 0.229744,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.76483870967742e-05,
+      "loss": 0.1639,
+      "step": 14359
+    },
+    {
+      "epoch": 0.22976,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.764677419354839e-05,
+      "loss": 0.1573,
+      "step": 14360
+    },
+    {
+      "epoch": 0.229776,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.764516129032259e-05,
+      "loss": 0.1629,
+      "step": 14361
+    },
+    {
+      "epoch": 0.229792,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.764354838709678e-05,
+      "loss": 0.1499,
+      "step": 14362
+    },
+    {
+      "epoch": 0.229808,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.764193548387097e-05,
+      "loss": 0.1514,
+      "step": 14363
+    },
+    {
+      "epoch": 0.229824,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.764032258064516e-05,
+      "loss": 0.1659,
+      "step": 14364
+    },
+    {
+      "epoch": 0.22984,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.763870967741936e-05,
+      "loss": 0.1821,
+      "step": 14365
+    },
+    {
+      "epoch": 0.229856,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.763709677419355e-05,
+      "loss": 0.1526,
+      "step": 14366
+    },
+    {
+      "epoch": 0.229872,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.763548387096775e-05,
+      "loss": 0.1496,
+      "step": 14367
+    },
+    {
+      "epoch": 0.229888,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.763387096774195e-05,
+      "loss": 0.1824,
+      "step": 14368
+    },
+    {
+      "epoch": 0.229904,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.763225806451613e-05,
+      "loss": 0.1826,
+      "step": 14369
+    },
+    {
+      "epoch": 0.22992,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.763064516129033e-05,
+      "loss": 0.1379,
+      "step": 14370
+    },
+    {
+      "epoch": 0.229936,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.762903225806452e-05,
+      "loss": 0.1636,
+      "step": 14371
+    },
+    {
+      "epoch": 0.229952,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.762741935483872e-05,
+      "loss": 0.1744,
+      "step": 14372
+    },
+    {
+      "epoch": 0.229968,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.76258064516129e-05,
+      "loss": 0.2205,
+      "step": 14373
+    },
+    {
+      "epoch": 0.229984,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.76241935483871e-05,
+      "loss": 0.1489,
+      "step": 14374
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.762258064516129e-05,
+      "loss": 0.1665,
+      "step": 14375
+    },
+    {
+      "epoch": 0.230016,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.762096774193549e-05,
+      "loss": 0.1611,
+      "step": 14376
+    },
+    {
+      "epoch": 0.230032,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.761935483870967e-05,
+      "loss": 0.1894,
+      "step": 14377
+    },
+    {
+      "epoch": 0.230048,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.761774193548387e-05,
+      "loss": 0.1515,
+      "step": 14378
+    },
+    {
+      "epoch": 0.230064,
+      "grad_norm": 1.2578125,
+      "learning_rate": 7.761612903225806e-05,
+      "loss": 0.2571,
+      "step": 14379
+    },
+    {
+      "epoch": 0.23008,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.761451612903226e-05,
+      "loss": 0.1659,
+      "step": 14380
+    },
+    {
+      "epoch": 0.230096,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.761290322580646e-05,
+      "loss": 0.2038,
+      "step": 14381
+    },
+    {
+      "epoch": 0.230112,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.761129032258066e-05,
+      "loss": 0.1567,
+      "step": 14382
+    },
+    {
+      "epoch": 0.230128,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.760967741935485e-05,
+      "loss": 0.1964,
+      "step": 14383
+    },
+    {
+      "epoch": 0.230144,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.760806451612903e-05,
+      "loss": 0.1674,
+      "step": 14384
+    },
+    {
+      "epoch": 0.23016,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.760645161290323e-05,
+      "loss": 0.1744,
+      "step": 14385
+    },
+    {
+      "epoch": 0.230176,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.760483870967742e-05,
+      "loss": 0.1824,
+      "step": 14386
+    },
+    {
+      "epoch": 0.230192,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.760322580645162e-05,
+      "loss": 0.1493,
+      "step": 14387
+    },
+    {
+      "epoch": 0.230208,
+      "grad_norm": 1.2421875,
+      "learning_rate": 7.76016129032258e-05,
+      "loss": 0.1837,
+      "step": 14388
+    },
+    {
+      "epoch": 0.230224,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.76e-05,
+      "loss": 0.1566,
+      "step": 14389
+    },
+    {
+      "epoch": 0.23024,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.759838709677419e-05,
+      "loss": 0.1464,
+      "step": 14390
+    },
+    {
+      "epoch": 0.230256,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.759677419354839e-05,
+      "loss": 0.1542,
+      "step": 14391
+    },
+    {
+      "epoch": 0.230272,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.759516129032259e-05,
+      "loss": 0.1953,
+      "step": 14392
+    },
+    {
+      "epoch": 0.230288,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.759354838709679e-05,
+      "loss": 0.1579,
+      "step": 14393
+    },
+    {
+      "epoch": 0.230304,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.759193548387097e-05,
+      "loss": 0.1781,
+      "step": 14394
+    },
+    {
+      "epoch": 0.23032,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.759032258064517e-05,
+      "loss": 0.1894,
+      "step": 14395
+    },
+    {
+      "epoch": 0.230336,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.758870967741936e-05,
+      "loss": 0.153,
+      "step": 14396
+    },
+    {
+      "epoch": 0.230352,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.758709677419356e-05,
+      "loss": 0.1984,
+      "step": 14397
+    },
+    {
+      "epoch": 0.230368,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.758548387096774e-05,
+      "loss": 0.1479,
+      "step": 14398
+    },
+    {
+      "epoch": 0.230384,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.758387096774193e-05,
+      "loss": 0.1515,
+      "step": 14399
+    },
+    {
+      "epoch": 0.2304,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.758225806451613e-05,
+      "loss": 0.1526,
+      "step": 14400
+    },
+    {
+      "epoch": 0.230416,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.758064516129032e-05,
+      "loss": 0.1755,
+      "step": 14401
+    },
+    {
+      "epoch": 0.230432,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.757903225806452e-05,
+      "loss": 0.178,
+      "step": 14402
+    },
+    {
+      "epoch": 0.230448,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.757741935483872e-05,
+      "loss": 0.1545,
+      "step": 14403
+    },
+    {
+      "epoch": 0.230464,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.757580645161292e-05,
+      "loss": 0.1746,
+      "step": 14404
+    },
+    {
+      "epoch": 0.23048,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.75741935483871e-05,
+      "loss": 0.1725,
+      "step": 14405
+    },
+    {
+      "epoch": 0.230496,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.75725806451613e-05,
+      "loss": 0.1845,
+      "step": 14406
+    },
+    {
+      "epoch": 0.230512,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.757096774193549e-05,
+      "loss": 0.2336,
+      "step": 14407
+    },
+    {
+      "epoch": 0.230528,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.756935483870969e-05,
+      "loss": 0.1655,
+      "step": 14408
+    },
+    {
+      "epoch": 0.230544,
+      "grad_norm": 0.625,
+      "learning_rate": 7.756774193548387e-05,
+      "loss": 0.1671,
+      "step": 14409
+    },
+    {
+      "epoch": 0.23056,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.756612903225807e-05,
+      "loss": 0.1615,
+      "step": 14410
+    },
+    {
+      "epoch": 0.230576,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.756451612903226e-05,
+      "loss": 0.1848,
+      "step": 14411
+    },
+    {
+      "epoch": 0.230592,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.756290322580646e-05,
+      "loss": 0.2085,
+      "step": 14412
+    },
+    {
+      "epoch": 0.230608,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.756129032258064e-05,
+      "loss": 0.1451,
+      "step": 14413
+    },
+    {
+      "epoch": 0.230624,
+      "grad_norm": 1.3359375,
+      "learning_rate": 7.755967741935483e-05,
+      "loss": 0.146,
+      "step": 14414
+    },
+    {
+      "epoch": 0.23064,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.755806451612903e-05,
+      "loss": 0.1391,
+      "step": 14415
+    },
+    {
+      "epoch": 0.230656,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.755645161290323e-05,
+      "loss": 0.1422,
+      "step": 14416
+    },
+    {
+      "epoch": 0.230672,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.755483870967743e-05,
+      "loss": 0.1551,
+      "step": 14417
+    },
+    {
+      "epoch": 0.230688,
+      "grad_norm": 1.265625,
+      "learning_rate": 7.755322580645162e-05,
+      "loss": 0.2078,
+      "step": 14418
+    },
+    {
+      "epoch": 0.230704,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.755161290322582e-05,
+      "loss": 0.1725,
+      "step": 14419
+    },
+    {
+      "epoch": 0.23072,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.755e-05,
+      "loss": 0.1655,
+      "step": 14420
+    },
+    {
+      "epoch": 0.230736,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.75483870967742e-05,
+      "loss": 0.1667,
+      "step": 14421
+    },
+    {
+      "epoch": 0.230752,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.754677419354839e-05,
+      "loss": 0.1839,
+      "step": 14422
+    },
+    {
+      "epoch": 0.230768,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.754516129032259e-05,
+      "loss": 0.1616,
+      "step": 14423
+    },
+    {
+      "epoch": 0.230784,
+      "grad_norm": 1.265625,
+      "learning_rate": 7.754354838709677e-05,
+      "loss": 0.1426,
+      "step": 14424
+    },
+    {
+      "epoch": 0.2308,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.754193548387097e-05,
+      "loss": 0.2025,
+      "step": 14425
+    },
+    {
+      "epoch": 0.230816,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.754032258064516e-05,
+      "loss": 0.178,
+      "step": 14426
+    },
+    {
+      "epoch": 0.230832,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.753870967741936e-05,
+      "loss": 0.2071,
+      "step": 14427
+    },
+    {
+      "epoch": 0.230848,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.753709677419356e-05,
+      "loss": 0.125,
+      "step": 14428
+    },
+    {
+      "epoch": 0.230864,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.753548387096776e-05,
+      "loss": 0.1853,
+      "step": 14429
+    },
+    {
+      "epoch": 0.23088,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.753387096774194e-05,
+      "loss": 0.1772,
+      "step": 14430
+    },
+    {
+      "epoch": 0.230896,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.753225806451613e-05,
+      "loss": 0.1919,
+      "step": 14431
+    },
+    {
+      "epoch": 0.230912,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.753064516129033e-05,
+      "loss": 0.1737,
+      "step": 14432
+    },
+    {
+      "epoch": 0.230928,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.752903225806452e-05,
+      "loss": 0.1985,
+      "step": 14433
+    },
+    {
+      "epoch": 0.230944,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.752741935483871e-05,
+      "loss": 0.1593,
+      "step": 14434
+    },
+    {
+      "epoch": 0.23096,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.75258064516129e-05,
+      "loss": 0.2084,
+      "step": 14435
+    },
+    {
+      "epoch": 0.230976,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.75241935483871e-05,
+      "loss": 0.1757,
+      "step": 14436
+    },
+    {
+      "epoch": 0.230992,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.752258064516129e-05,
+      "loss": 0.1492,
+      "step": 14437
+    },
+    {
+      "epoch": 0.231008,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.752096774193549e-05,
+      "loss": 0.1771,
+      "step": 14438
+    },
+    {
+      "epoch": 0.231024,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.751935483870969e-05,
+      "loss": 0.1808,
+      "step": 14439
+    },
+    {
+      "epoch": 0.23104,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.751774193548387e-05,
+      "loss": 0.1813,
+      "step": 14440
+    },
+    {
+      "epoch": 0.231056,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.751612903225807e-05,
+      "loss": 0.1994,
+      "step": 14441
+    },
+    {
+      "epoch": 0.231072,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.751451612903227e-05,
+      "loss": 0.1428,
+      "step": 14442
+    },
+    {
+      "epoch": 0.231088,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.751290322580646e-05,
+      "loss": 0.1227,
+      "step": 14443
+    },
+    {
+      "epoch": 0.231104,
+      "grad_norm": 0.625,
+      "learning_rate": 7.751129032258066e-05,
+      "loss": 0.1983,
+      "step": 14444
+    },
+    {
+      "epoch": 0.23112,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.750967741935484e-05,
+      "loss": 0.1265,
+      "step": 14445
+    },
+    {
+      "epoch": 0.231136,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.750806451612903e-05,
+      "loss": 0.1654,
+      "step": 14446
+    },
+    {
+      "epoch": 0.231152,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.750645161290323e-05,
+      "loss": 0.1845,
+      "step": 14447
+    },
+    {
+      "epoch": 0.231168,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.750483870967741e-05,
+      "loss": 0.1949,
+      "step": 14448
+    },
+    {
+      "epoch": 0.231184,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.750322580645161e-05,
+      "loss": 0.1503,
+      "step": 14449
+    },
+    {
+      "epoch": 0.2312,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.75016129032258e-05,
+      "loss": 0.1518,
+      "step": 14450
+    },
+    {
+      "epoch": 0.231216,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.75e-05,
+      "loss": 0.2024,
+      "step": 14451
+    },
+    {
+      "epoch": 0.231232,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.74983870967742e-05,
+      "loss": 0.1548,
+      "step": 14452
+    },
+    {
+      "epoch": 0.231248,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.74967741935484e-05,
+      "loss": 0.1597,
+      "step": 14453
+    },
+    {
+      "epoch": 0.231264,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.749516129032259e-05,
+      "loss": 0.1715,
+      "step": 14454
+    },
+    {
+      "epoch": 0.23128,
+      "grad_norm": 0.4921875,
+      "learning_rate": 7.749354838709678e-05,
+      "loss": 0.1572,
+      "step": 14455
+    },
+    {
+      "epoch": 0.231296,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.749193548387097e-05,
+      "loss": 0.1655,
+      "step": 14456
+    },
+    {
+      "epoch": 0.231312,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.749032258064517e-05,
+      "loss": 0.1692,
+      "step": 14457
+    },
+    {
+      "epoch": 0.231328,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.748870967741936e-05,
+      "loss": 0.1961,
+      "step": 14458
+    },
+    {
+      "epoch": 0.231344,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.748709677419356e-05,
+      "loss": 0.1531,
+      "step": 14459
+    },
+    {
+      "epoch": 0.23136,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.748548387096774e-05,
+      "loss": 0.1836,
+      "step": 14460
+    },
+    {
+      "epoch": 0.231376,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.748387096774193e-05,
+      "loss": 0.1415,
+      "step": 14461
+    },
+    {
+      "epoch": 0.231392,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.748225806451613e-05,
+      "loss": 0.1721,
+      "step": 14462
+    },
+    {
+      "epoch": 0.231408,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.748064516129033e-05,
+      "loss": 0.1948,
+      "step": 14463
+    },
+    {
+      "epoch": 0.231424,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.747903225806453e-05,
+      "loss": 0.1643,
+      "step": 14464
+    },
+    {
+      "epoch": 0.23144,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.747741935483871e-05,
+      "loss": 0.1712,
+      "step": 14465
+    },
+    {
+      "epoch": 0.231456,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.747580645161291e-05,
+      "loss": 0.2554,
+      "step": 14466
+    },
+    {
+      "epoch": 0.231472,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.74741935483871e-05,
+      "loss": 0.1651,
+      "step": 14467
+    },
+    {
+      "epoch": 0.231488,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.74725806451613e-05,
+      "loss": 0.1968,
+      "step": 14468
+    },
+    {
+      "epoch": 0.231504,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.747096774193548e-05,
+      "loss": 0.1687,
+      "step": 14469
+    },
+    {
+      "epoch": 0.23152,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.746935483870968e-05,
+      "loss": 0.1871,
+      "step": 14470
+    },
+    {
+      "epoch": 0.231536,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.746774193548387e-05,
+      "loss": 0.1535,
+      "step": 14471
+    },
+    {
+      "epoch": 0.231552,
+      "grad_norm": 0.875,
+      "learning_rate": 7.746612903225807e-05,
+      "loss": 0.1916,
+      "step": 14472
+    },
+    {
+      "epoch": 0.231568,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.746451612903226e-05,
+      "loss": 0.1911,
+      "step": 14473
+    },
+    {
+      "epoch": 0.231584,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.746290322580646e-05,
+      "loss": 0.182,
+      "step": 14474
+    },
+    {
+      "epoch": 0.2316,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.746129032258064e-05,
+      "loss": 0.1649,
+      "step": 14475
+    },
+    {
+      "epoch": 0.231616,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.745967741935484e-05,
+      "loss": 0.1774,
+      "step": 14476
+    },
+    {
+      "epoch": 0.231632,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.745806451612904e-05,
+      "loss": 0.1799,
+      "step": 14477
+    },
+    {
+      "epoch": 0.231648,
+      "grad_norm": 0.48046875,
+      "learning_rate": 7.745645161290323e-05,
+      "loss": 0.1462,
+      "step": 14478
+    },
+    {
+      "epoch": 0.231664,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.745483870967743e-05,
+      "loss": 0.1887,
+      "step": 14479
+    },
+    {
+      "epoch": 0.23168,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.745322580645161e-05,
+      "loss": 0.1543,
+      "step": 14480
+    },
+    {
+      "epoch": 0.231696,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.745161290322581e-05,
+      "loss": 0.1991,
+      "step": 14481
+    },
+    {
+      "epoch": 0.231712,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.745e-05,
+      "loss": 0.1688,
+      "step": 14482
+    },
+    {
+      "epoch": 0.231728,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.74483870967742e-05,
+      "loss": 0.2674,
+      "step": 14483
+    },
+    {
+      "epoch": 0.231744,
+      "grad_norm": 0.5078125,
+      "learning_rate": 7.744677419354838e-05,
+      "loss": 0.1581,
+      "step": 14484
+    },
+    {
+      "epoch": 0.23176,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.744516129032258e-05,
+      "loss": 0.2153,
+      "step": 14485
+    },
+    {
+      "epoch": 0.231776,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.744354838709677e-05,
+      "loss": 0.1777,
+      "step": 14486
+    },
+    {
+      "epoch": 0.231792,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.744193548387097e-05,
+      "loss": 0.2189,
+      "step": 14487
+    },
+    {
+      "epoch": 0.231808,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.744032258064517e-05,
+      "loss": 0.1745,
+      "step": 14488
+    },
+    {
+      "epoch": 0.231824,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.743870967741937e-05,
+      "loss": 0.1517,
+      "step": 14489
+    },
+    {
+      "epoch": 0.23184,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.743709677419356e-05,
+      "loss": 0.1654,
+      "step": 14490
+    },
+    {
+      "epoch": 0.231856,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.743548387096775e-05,
+      "loss": 0.1645,
+      "step": 14491
+    },
+    {
+      "epoch": 0.231872,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.743387096774194e-05,
+      "loss": 0.1802,
+      "step": 14492
+    },
+    {
+      "epoch": 0.231888,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.743225806451613e-05,
+      "loss": 0.1787,
+      "step": 14493
+    },
+    {
+      "epoch": 0.231904,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.743064516129033e-05,
+      "loss": 0.2303,
+      "step": 14494
+    },
+    {
+      "epoch": 0.23192,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.742903225806451e-05,
+      "loss": 0.1993,
+      "step": 14495
+    },
+    {
+      "epoch": 0.231936,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.742741935483871e-05,
+      "loss": 0.1883,
+      "step": 14496
+    },
+    {
+      "epoch": 0.231952,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.74258064516129e-05,
+      "loss": 0.19,
+      "step": 14497
+    },
+    {
+      "epoch": 0.231968,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.74241935483871e-05,
+      "loss": 0.1695,
+      "step": 14498
+    },
+    {
+      "epoch": 0.231984,
+      "grad_norm": 1.125,
+      "learning_rate": 7.74225806451613e-05,
+      "loss": 0.1933,
+      "step": 14499
+    },
+    {
+      "epoch": 0.232,
+      "grad_norm": 0.75,
+      "learning_rate": 7.74209677419355e-05,
+      "loss": 0.1703,
+      "step": 14500
+    },
+    {
+      "epoch": 0.232016,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.741935483870968e-05,
+      "loss": 0.1623,
+      "step": 14501
+    },
+    {
+      "epoch": 0.232032,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.741774193548388e-05,
+      "loss": 0.2099,
+      "step": 14502
+    },
+    {
+      "epoch": 0.232048,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.741612903225807e-05,
+      "loss": 0.1847,
+      "step": 14503
+    },
+    {
+      "epoch": 0.232064,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.741451612903227e-05,
+      "loss": 0.157,
+      "step": 14504
+    },
+    {
+      "epoch": 0.23208,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.741290322580645e-05,
+      "loss": 0.1759,
+      "step": 14505
+    },
+    {
+      "epoch": 0.232096,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.741129032258065e-05,
+      "loss": 0.1788,
+      "step": 14506
+    },
+    {
+      "epoch": 0.232112,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.740967741935484e-05,
+      "loss": 0.2113,
+      "step": 14507
+    },
+    {
+      "epoch": 0.232128,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.740806451612903e-05,
+      "loss": 0.1739,
+      "step": 14508
+    },
+    {
+      "epoch": 0.232144,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.740645161290323e-05,
+      "loss": 0.2099,
+      "step": 14509
+    },
+    {
+      "epoch": 0.23216,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.740483870967741e-05,
+      "loss": 0.1731,
+      "step": 14510
+    },
+    {
+      "epoch": 0.232176,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.740322580645161e-05,
+      "loss": 0.172,
+      "step": 14511
+    },
+    {
+      "epoch": 0.232192,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.740161290322581e-05,
+      "loss": 0.1707,
+      "step": 14512
+    },
+    {
+      "epoch": 0.232208,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.740000000000001e-05,
+      "loss": 0.147,
+      "step": 14513
+    },
+    {
+      "epoch": 0.232224,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.73983870967742e-05,
+      "loss": 0.1861,
+      "step": 14514
+    },
+    {
+      "epoch": 0.23224,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.73967741935484e-05,
+      "loss": 0.154,
+      "step": 14515
+    },
+    {
+      "epoch": 0.232256,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.739516129032258e-05,
+      "loss": 0.1795,
+      "step": 14516
+    },
+    {
+      "epoch": 0.232272,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.739354838709678e-05,
+      "loss": 0.1219,
+      "step": 14517
+    },
+    {
+      "epoch": 0.232288,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.739193548387097e-05,
+      "loss": 0.1978,
+      "step": 14518
+    },
+    {
+      "epoch": 0.232304,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.739032258064517e-05,
+      "loss": 0.157,
+      "step": 14519
+    },
+    {
+      "epoch": 0.23232,
+      "grad_norm": 0.46875,
+      "learning_rate": 7.738870967741935e-05,
+      "loss": 0.157,
+      "step": 14520
+    },
+    {
+      "epoch": 0.232336,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.738709677419355e-05,
+      "loss": 0.1395,
+      "step": 14521
+    },
+    {
+      "epoch": 0.232352,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.738548387096774e-05,
+      "loss": 0.1759,
+      "step": 14522
+    },
+    {
+      "epoch": 0.232368,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.738387096774194e-05,
+      "loss": 0.1716,
+      "step": 14523
+    },
+    {
+      "epoch": 0.232384,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.738225806451614e-05,
+      "loss": 0.1425,
+      "step": 14524
+    },
+    {
+      "epoch": 0.2324,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.738064516129033e-05,
+      "loss": 0.187,
+      "step": 14525
+    },
+    {
+      "epoch": 0.232416,
+      "grad_norm": 0.48828125,
+      "learning_rate": 7.737903225806452e-05,
+      "loss": 0.1543,
+      "step": 14526
+    },
+    {
+      "epoch": 0.232432,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.737741935483871e-05,
+      "loss": 0.1479,
+      "step": 14527
+    },
+    {
+      "epoch": 0.232448,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.737580645161291e-05,
+      "loss": 0.1621,
+      "step": 14528
+    },
+    {
+      "epoch": 0.232464,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.73741935483871e-05,
+      "loss": 0.1518,
+      "step": 14529
+    },
+    {
+      "epoch": 0.23248,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.73725806451613e-05,
+      "loss": 0.1883,
+      "step": 14530
+    },
+    {
+      "epoch": 0.232496,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.737096774193548e-05,
+      "loss": 0.1953,
+      "step": 14531
+    },
+    {
+      "epoch": 0.232512,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.736935483870968e-05,
+      "loss": 0.1689,
+      "step": 14532
+    },
+    {
+      "epoch": 0.232528,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.736774193548387e-05,
+      "loss": 0.1612,
+      "step": 14533
+    },
+    {
+      "epoch": 0.232544,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.736612903225807e-05,
+      "loss": 0.1472,
+      "step": 14534
+    },
+    {
+      "epoch": 0.23256,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.736451612903225e-05,
+      "loss": 0.1515,
+      "step": 14535
+    },
+    {
+      "epoch": 0.232576,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.736290322580645e-05,
+      "loss": 0.1305,
+      "step": 14536
+    },
+    {
+      "epoch": 0.232592,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.736129032258065e-05,
+      "loss": 0.1584,
+      "step": 14537
+    },
+    {
+      "epoch": 0.232608,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.735967741935485e-05,
+      "loss": 0.15,
+      "step": 14538
+    },
+    {
+      "epoch": 0.232624,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.735806451612904e-05,
+      "loss": 0.1547,
+      "step": 14539
+    },
+    {
+      "epoch": 0.23264,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.735645161290322e-05,
+      "loss": 0.1404,
+      "step": 14540
+    },
+    {
+      "epoch": 0.232656,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.735483870967742e-05,
+      "loss": 0.1914,
+      "step": 14541
+    },
+    {
+      "epoch": 0.232672,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.735322580645161e-05,
+      "loss": 0.1555,
+      "step": 14542
+    },
+    {
+      "epoch": 0.232688,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.735161290322581e-05,
+      "loss": 0.1751,
+      "step": 14543
+    },
+    {
+      "epoch": 0.232704,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.735e-05,
+      "loss": 0.1863,
+      "step": 14544
+    },
+    {
+      "epoch": 0.23272,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.73483870967742e-05,
+      "loss": 0.1908,
+      "step": 14545
+    },
+    {
+      "epoch": 0.232736,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.734677419354838e-05,
+      "loss": 0.178,
+      "step": 14546
+    },
+    {
+      "epoch": 0.232752,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.734516129032258e-05,
+      "loss": 0.1412,
+      "step": 14547
+    },
+    {
+      "epoch": 0.232768,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.734354838709678e-05,
+      "loss": 0.1686,
+      "step": 14548
+    },
+    {
+      "epoch": 0.232784,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.734193548387098e-05,
+      "loss": 0.1458,
+      "step": 14549
+    },
+    {
+      "epoch": 0.2328,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.734032258064517e-05,
+      "loss": 0.1467,
+      "step": 14550
+    },
+    {
+      "epoch": 0.232816,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.733870967741937e-05,
+      "loss": 0.2078,
+      "step": 14551
+    },
+    {
+      "epoch": 0.232832,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.733709677419355e-05,
+      "loss": 0.1771,
+      "step": 14552
+    },
+    {
+      "epoch": 0.232848,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.733548387096775e-05,
+      "loss": 0.188,
+      "step": 14553
+    },
+    {
+      "epoch": 0.232864,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.733387096774194e-05,
+      "loss": 0.1963,
+      "step": 14554
+    },
+    {
+      "epoch": 0.23288,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.733225806451612e-05,
+      "loss": 0.1905,
+      "step": 14555
+    },
+    {
+      "epoch": 0.232896,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.733064516129032e-05,
+      "loss": 0.1442,
+      "step": 14556
+    },
+    {
+      "epoch": 0.232912,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.732903225806451e-05,
+      "loss": 0.1807,
+      "step": 14557
+    },
+    {
+      "epoch": 0.232928,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.732741935483871e-05,
+      "loss": 0.1952,
+      "step": 14558
+    },
+    {
+      "epoch": 0.232944,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.732580645161291e-05,
+      "loss": 0.1831,
+      "step": 14559
+    },
+    {
+      "epoch": 0.23296,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.732419354838711e-05,
+      "loss": 0.1819,
+      "step": 14560
+    },
+    {
+      "epoch": 0.232976,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.73225806451613e-05,
+      "loss": 0.1848,
+      "step": 14561
+    },
+    {
+      "epoch": 0.232992,
+      "grad_norm": 0.625,
+      "learning_rate": 7.73209677419355e-05,
+      "loss": 0.1649,
+      "step": 14562
+    },
+    {
+      "epoch": 0.233008,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.731935483870968e-05,
+      "loss": 0.2448,
+      "step": 14563
+    },
+    {
+      "epoch": 0.233024,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.731774193548388e-05,
+      "loss": 0.197,
+      "step": 14564
+    },
+    {
+      "epoch": 0.23304,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.731612903225807e-05,
+      "loss": 0.175,
+      "step": 14565
+    },
+    {
+      "epoch": 0.233056,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.731451612903227e-05,
+      "loss": 0.2002,
+      "step": 14566
+    },
+    {
+      "epoch": 0.233072,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.731290322580645e-05,
+      "loss": 0.175,
+      "step": 14567
+    },
+    {
+      "epoch": 0.233088,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.731129032258065e-05,
+      "loss": 0.1962,
+      "step": 14568
+    },
+    {
+      "epoch": 0.233104,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.730967741935484e-05,
+      "loss": 0.2008,
+      "step": 14569
+    },
+    {
+      "epoch": 0.23312,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.730806451612902e-05,
+      "loss": 0.1255,
+      "step": 14570
+    },
+    {
+      "epoch": 0.233136,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.730645161290322e-05,
+      "loss": 0.1599,
+      "step": 14571
+    },
+    {
+      "epoch": 0.233152,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.730483870967742e-05,
+      "loss": 0.2022,
+      "step": 14572
+    },
+    {
+      "epoch": 0.233168,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.730322580645162e-05,
+      "loss": 0.1465,
+      "step": 14573
+    },
+    {
+      "epoch": 0.233184,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.730161290322581e-05,
+      "loss": 0.1668,
+      "step": 14574
+    },
+    {
+      "epoch": 0.2332,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.730000000000001e-05,
+      "loss": 0.1568,
+      "step": 14575
+    },
+    {
+      "epoch": 0.233216,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.72983870967742e-05,
+      "loss": 0.1804,
+      "step": 14576
+    },
+    {
+      "epoch": 0.233232,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.72967741935484e-05,
+      "loss": 0.1706,
+      "step": 14577
+    },
+    {
+      "epoch": 0.233248,
+      "grad_norm": 1.0,
+      "learning_rate": 7.729516129032258e-05,
+      "loss": 0.1963,
+      "step": 14578
+    },
+    {
+      "epoch": 0.233264,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.729354838709678e-05,
+      "loss": 0.2005,
+      "step": 14579
+    },
+    {
+      "epoch": 0.23328,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.729193548387097e-05,
+      "loss": 0.1567,
+      "step": 14580
+    },
+    {
+      "epoch": 0.233296,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.729032258064517e-05,
+      "loss": 0.2156,
+      "step": 14581
+    },
+    {
+      "epoch": 0.233312,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.728870967741935e-05,
+      "loss": 0.1833,
+      "step": 14582
+    },
+    {
+      "epoch": 0.233328,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.728709677419355e-05,
+      "loss": 0.1325,
+      "step": 14583
+    },
+    {
+      "epoch": 0.233344,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.728548387096775e-05,
+      "loss": 0.1894,
+      "step": 14584
+    },
+    {
+      "epoch": 0.23336,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.728387096774194e-05,
+      "loss": 0.1633,
+      "step": 14585
+    },
+    {
+      "epoch": 0.233376,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.728225806451614e-05,
+      "loss": 0.1768,
+      "step": 14586
+    },
+    {
+      "epoch": 0.233392,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.728064516129032e-05,
+      "loss": 0.1712,
+      "step": 14587
+    },
+    {
+      "epoch": 0.233408,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.727903225806452e-05,
+      "loss": 0.1447,
+      "step": 14588
+    },
+    {
+      "epoch": 0.233424,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.727741935483871e-05,
+      "loss": 0.1785,
+      "step": 14589
+    },
+    {
+      "epoch": 0.23344,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.727580645161291e-05,
+      "loss": 0.1463,
+      "step": 14590
+    },
+    {
+      "epoch": 0.233456,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.72741935483871e-05,
+      "loss": 0.1514,
+      "step": 14591
+    },
+    {
+      "epoch": 0.233472,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.72725806451613e-05,
+      "loss": 0.1431,
+      "step": 14592
+    },
+    {
+      "epoch": 0.233488,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.727096774193548e-05,
+      "loss": 0.1827,
+      "step": 14593
+    },
+    {
+      "epoch": 0.233504,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.726935483870968e-05,
+      "loss": 0.1521,
+      "step": 14594
+    },
+    {
+      "epoch": 0.23352,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.726774193548388e-05,
+      "loss": 0.1825,
+      "step": 14595
+    },
+    {
+      "epoch": 0.233536,
+      "grad_norm": 1.2421875,
+      "learning_rate": 7.726612903225808e-05,
+      "loss": 0.1217,
+      "step": 14596
+    },
+    {
+      "epoch": 0.233552,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.726451612903227e-05,
+      "loss": 0.1756,
+      "step": 14597
+    },
+    {
+      "epoch": 0.233568,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.726290322580646e-05,
+      "loss": 0.149,
+      "step": 14598
+    },
+    {
+      "epoch": 0.233584,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.726129032258065e-05,
+      "loss": 0.1743,
+      "step": 14599
+    },
+    {
+      "epoch": 0.2336,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.725967741935485e-05,
+      "loss": 0.1392,
+      "step": 14600
+    },
+    {
+      "epoch": 0.233616,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.725806451612904e-05,
+      "loss": 0.1416,
+      "step": 14601
+    },
+    {
+      "epoch": 0.233632,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.725645161290322e-05,
+      "loss": 0.2279,
+      "step": 14602
+    },
+    {
+      "epoch": 0.233648,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.725483870967742e-05,
+      "loss": 0.1385,
+      "step": 14603
+    },
+    {
+      "epoch": 0.233664,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.725322580645161e-05,
+      "loss": 0.1738,
+      "step": 14604
+    },
+    {
+      "epoch": 0.23368,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.725161290322581e-05,
+      "loss": 0.1748,
+      "step": 14605
+    },
+    {
+      "epoch": 0.233696,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.725e-05,
+      "loss": 0.1617,
+      "step": 14606
+    },
+    {
+      "epoch": 0.233712,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.72483870967742e-05,
+      "loss": 0.2016,
+      "step": 14607
+    },
+    {
+      "epoch": 0.233728,
+      "grad_norm": 0.4921875,
+      "learning_rate": 7.724677419354839e-05,
+      "loss": 0.1475,
+      "step": 14608
+    },
+    {
+      "epoch": 0.233744,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.724516129032259e-05,
+      "loss": 0.1603,
+      "step": 14609
+    },
+    {
+      "epoch": 0.23376,
+      "grad_norm": 0.46484375,
+      "learning_rate": 7.724354838709678e-05,
+      "loss": 0.1538,
+      "step": 14610
+    },
+    {
+      "epoch": 0.233776,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.724193548387098e-05,
+      "loss": 0.1736,
+      "step": 14611
+    },
+    {
+      "epoch": 0.233792,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.724032258064516e-05,
+      "loss": 0.1441,
+      "step": 14612
+    },
+    {
+      "epoch": 0.233808,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.723870967741936e-05,
+      "loss": 0.152,
+      "step": 14613
+    },
+    {
+      "epoch": 0.233824,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.723709677419355e-05,
+      "loss": 0.1808,
+      "step": 14614
+    },
+    {
+      "epoch": 0.23384,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.723548387096775e-05,
+      "loss": 0.1889,
+      "step": 14615
+    },
+    {
+      "epoch": 0.233856,
+      "grad_norm": 1.2421875,
+      "learning_rate": 7.723387096774194e-05,
+      "loss": 0.1947,
+      "step": 14616
+    },
+    {
+      "epoch": 0.233872,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.723225806451612e-05,
+      "loss": 0.1778,
+      "step": 14617
+    },
+    {
+      "epoch": 0.233888,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.723064516129032e-05,
+      "loss": 0.1811,
+      "step": 14618
+    },
+    {
+      "epoch": 0.233904,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.722903225806452e-05,
+      "loss": 0.1591,
+      "step": 14619
+    },
+    {
+      "epoch": 0.23392,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.722741935483872e-05,
+      "loss": 0.1625,
+      "step": 14620
+    },
+    {
+      "epoch": 0.233936,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.722580645161291e-05,
+      "loss": 0.2065,
+      "step": 14621
+    },
+    {
+      "epoch": 0.233952,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.72241935483871e-05,
+      "loss": 0.1447,
+      "step": 14622
+    },
+    {
+      "epoch": 0.233968,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.722258064516129e-05,
+      "loss": 0.1697,
+      "step": 14623
+    },
+    {
+      "epoch": 0.233984,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.722096774193549e-05,
+      "loss": 0.1903,
+      "step": 14624
+    },
+    {
+      "epoch": 0.234,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.721935483870968e-05,
+      "loss": 0.2035,
+      "step": 14625
+    },
+    {
+      "epoch": 0.234016,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.721774193548388e-05,
+      "loss": 0.1608,
+      "step": 14626
+    },
+    {
+      "epoch": 0.234032,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.721612903225806e-05,
+      "loss": 0.172,
+      "step": 14627
+    },
+    {
+      "epoch": 0.234048,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.721451612903226e-05,
+      "loss": 0.1828,
+      "step": 14628
+    },
+    {
+      "epoch": 0.234064,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.721290322580645e-05,
+      "loss": 0.1308,
+      "step": 14629
+    },
+    {
+      "epoch": 0.23408,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.721129032258065e-05,
+      "loss": 0.1568,
+      "step": 14630
+    },
+    {
+      "epoch": 0.234096,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.720967741935484e-05,
+      "loss": 0.1594,
+      "step": 14631
+    },
+    {
+      "epoch": 0.234112,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.720806451612904e-05,
+      "loss": 0.1614,
+      "step": 14632
+    },
+    {
+      "epoch": 0.234128,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.720645161290323e-05,
+      "loss": 0.1897,
+      "step": 14633
+    },
+    {
+      "epoch": 0.234144,
+      "grad_norm": 1.296875,
+      "learning_rate": 7.720483870967742e-05,
+      "loss": 0.2202,
+      "step": 14634
+    },
+    {
+      "epoch": 0.23416,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.720322580645162e-05,
+      "loss": 0.1465,
+      "step": 14635
+    },
+    {
+      "epoch": 0.234176,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.72016129032258e-05,
+      "loss": 0.1508,
+      "step": 14636
+    },
+    {
+      "epoch": 0.234192,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.72e-05,
+      "loss": 0.1477,
+      "step": 14637
+    },
+    {
+      "epoch": 0.234208,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.719838709677419e-05,
+      "loss": 0.1708,
+      "step": 14638
+    },
+    {
+      "epoch": 0.234224,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.719677419354839e-05,
+      "loss": 0.1479,
+      "step": 14639
+    },
+    {
+      "epoch": 0.23424,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.719516129032258e-05,
+      "loss": 0.1528,
+      "step": 14640
+    },
+    {
+      "epoch": 0.234256,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.719354838709678e-05,
+      "loss": 0.1645,
+      "step": 14641
+    },
+    {
+      "epoch": 0.234272,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.719193548387096e-05,
+      "loss": 0.1589,
+      "step": 14642
+    },
+    {
+      "epoch": 0.234288,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.719032258064516e-05,
+      "loss": 0.1809,
+      "step": 14643
+    },
+    {
+      "epoch": 0.234304,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.718870967741936e-05,
+      "loss": 0.211,
+      "step": 14644
+    },
+    {
+      "epoch": 0.23432,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.718709677419356e-05,
+      "loss": 0.1461,
+      "step": 14645
+    },
+    {
+      "epoch": 0.234336,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.718548387096775e-05,
+      "loss": 0.1793,
+      "step": 14646
+    },
+    {
+      "epoch": 0.234352,
+      "grad_norm": 1.5390625,
+      "learning_rate": 7.718387096774195e-05,
+      "loss": 0.1654,
+      "step": 14647
+    },
+    {
+      "epoch": 0.234368,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.718225806451613e-05,
+      "loss": 0.169,
+      "step": 14648
+    },
+    {
+      "epoch": 0.234384,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.718064516129032e-05,
+      "loss": 0.2041,
+      "step": 14649
+    },
+    {
+      "epoch": 0.2344,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.717903225806452e-05,
+      "loss": 0.1824,
+      "step": 14650
+    },
+    {
+      "epoch": 0.234416,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.71774193548387e-05,
+      "loss": 0.162,
+      "step": 14651
+    },
+    {
+      "epoch": 0.234432,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.71758064516129e-05,
+      "loss": 0.1876,
+      "step": 14652
+    },
+    {
+      "epoch": 0.234448,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.717419354838709e-05,
+      "loss": 0.1779,
+      "step": 14653
+    },
+    {
+      "epoch": 0.234464,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.717258064516129e-05,
+      "loss": 0.1417,
+      "step": 14654
+    },
+    {
+      "epoch": 0.23448,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.717096774193549e-05,
+      "loss": 0.1771,
+      "step": 14655
+    },
+    {
+      "epoch": 0.234496,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.716935483870969e-05,
+      "loss": 0.2104,
+      "step": 14656
+    },
+    {
+      "epoch": 0.234512,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.716774193548388e-05,
+      "loss": 0.2187,
+      "step": 14657
+    },
+    {
+      "epoch": 0.234528,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.716612903225808e-05,
+      "loss": 0.1902,
+      "step": 14658
+    },
+    {
+      "epoch": 0.234544,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.716451612903226e-05,
+      "loss": 0.1766,
+      "step": 14659
+    },
+    {
+      "epoch": 0.23456,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.716290322580646e-05,
+      "loss": 0.1991,
+      "step": 14660
+    },
+    {
+      "epoch": 0.234576,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.716129032258065e-05,
+      "loss": 0.1907,
+      "step": 14661
+    },
+    {
+      "epoch": 0.234592,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.715967741935485e-05,
+      "loss": 0.1773,
+      "step": 14662
+    },
+    {
+      "epoch": 0.234608,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.715806451612903e-05,
+      "loss": 0.1509,
+      "step": 14663
+    },
+    {
+      "epoch": 0.234624,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.715645161290322e-05,
+      "loss": 0.1627,
+      "step": 14664
+    },
+    {
+      "epoch": 0.23464,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.715483870967742e-05,
+      "loss": 0.1772,
+      "step": 14665
+    },
+    {
+      "epoch": 0.234656,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.71532258064516e-05,
+      "loss": 0.1628,
+      "step": 14666
+    },
+    {
+      "epoch": 0.234672,
+      "grad_norm": 0.46875,
+      "learning_rate": 7.71516129032258e-05,
+      "loss": 0.1669,
+      "step": 14667
+    },
+    {
+      "epoch": 0.234688,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.715e-05,
+      "loss": 0.1793,
+      "step": 14668
+    },
+    {
+      "epoch": 0.234704,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.71483870967742e-05,
+      "loss": 0.17,
+      "step": 14669
+    },
+    {
+      "epoch": 0.23472,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.714677419354839e-05,
+      "loss": 0.1344,
+      "step": 14670
+    },
+    {
+      "epoch": 0.234736,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.714516129032259e-05,
+      "loss": 0.1913,
+      "step": 14671
+    },
+    {
+      "epoch": 0.234752,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.714354838709678e-05,
+      "loss": 0.1943,
+      "step": 14672
+    },
+    {
+      "epoch": 0.234768,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.714193548387098e-05,
+      "loss": 0.1736,
+      "step": 14673
+    },
+    {
+      "epoch": 0.234784,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.714032258064516e-05,
+      "loss": 0.1891,
+      "step": 14674
+    },
+    {
+      "epoch": 0.2348,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.713870967741936e-05,
+      "loss": 0.1728,
+      "step": 14675
+    },
+    {
+      "epoch": 0.234816,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.713709677419355e-05,
+      "loss": 0.1909,
+      "step": 14676
+    },
+    {
+      "epoch": 0.234832,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.713548387096775e-05,
+      "loss": 0.1348,
+      "step": 14677
+    },
+    {
+      "epoch": 0.234848,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.713387096774193e-05,
+      "loss": 0.1781,
+      "step": 14678
+    },
+    {
+      "epoch": 0.234864,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.713225806451613e-05,
+      "loss": 0.147,
+      "step": 14679
+    },
+    {
+      "epoch": 0.23488,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.713064516129033e-05,
+      "loss": 0.1912,
+      "step": 14680
+    },
+    {
+      "epoch": 0.234896,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.712903225806452e-05,
+      "loss": 0.1828,
+      "step": 14681
+    },
+    {
+      "epoch": 0.234912,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.712741935483872e-05,
+      "loss": 0.1836,
+      "step": 14682
+    },
+    {
+      "epoch": 0.234928,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.71258064516129e-05,
+      "loss": 0.1907,
+      "step": 14683
+    },
+    {
+      "epoch": 0.234944,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.71241935483871e-05,
+      "loss": 0.1763,
+      "step": 14684
+    },
+    {
+      "epoch": 0.23496,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.712258064516129e-05,
+      "loss": 0.2226,
+      "step": 14685
+    },
+    {
+      "epoch": 0.234976,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.712096774193549e-05,
+      "loss": 0.1258,
+      "step": 14686
+    },
+    {
+      "epoch": 0.234992,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.711935483870968e-05,
+      "loss": 0.1874,
+      "step": 14687
+    },
+    {
+      "epoch": 0.235008,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.711774193548388e-05,
+      "loss": 0.2055,
+      "step": 14688
+    },
+    {
+      "epoch": 0.235024,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.711612903225806e-05,
+      "loss": 0.1595,
+      "step": 14689
+    },
+    {
+      "epoch": 0.23504,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.711451612903226e-05,
+      "loss": 0.1446,
+      "step": 14690
+    },
+    {
+      "epoch": 0.235056,
+      "grad_norm": 1.0,
+      "learning_rate": 7.711290322580646e-05,
+      "loss": 0.1627,
+      "step": 14691
+    },
+    {
+      "epoch": 0.235072,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.711129032258065e-05,
+      "loss": 0.1803,
+      "step": 14692
+    },
+    {
+      "epoch": 0.235088,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.710967741935485e-05,
+      "loss": 0.1417,
+      "step": 14693
+    },
+    {
+      "epoch": 0.235104,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.710806451612903e-05,
+      "loss": 0.1533,
+      "step": 14694
+    },
+    {
+      "epoch": 0.23512,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.710645161290323e-05,
+      "loss": 0.1508,
+      "step": 14695
+    },
+    {
+      "epoch": 0.235136,
+      "grad_norm": 0.484375,
+      "learning_rate": 7.710483870967742e-05,
+      "loss": 0.1818,
+      "step": 14696
+    },
+    {
+      "epoch": 0.235152,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.710322580645162e-05,
+      "loss": 0.1595,
+      "step": 14697
+    },
+    {
+      "epoch": 0.235168,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.71016129032258e-05,
+      "loss": 0.1546,
+      "step": 14698
+    },
+    {
+      "epoch": 0.235184,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.71e-05,
+      "loss": 0.1695,
+      "step": 14699
+    },
+    {
+      "epoch": 0.2352,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.709838709677419e-05,
+      "loss": 0.2073,
+      "step": 14700
+    },
+    {
+      "epoch": 0.235216,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.709677419354839e-05,
+      "loss": 0.1876,
+      "step": 14701
+    },
+    {
+      "epoch": 0.235232,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.709516129032258e-05,
+      "loss": 0.164,
+      "step": 14702
+    },
+    {
+      "epoch": 0.235248,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.709354838709678e-05,
+      "loss": 0.1719,
+      "step": 14703
+    },
+    {
+      "epoch": 0.235264,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.709193548387097e-05,
+      "loss": 0.1635,
+      "step": 14704
+    },
+    {
+      "epoch": 0.23528,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.709032258064517e-05,
+      "loss": 0.1699,
+      "step": 14705
+    },
+    {
+      "epoch": 0.235296,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.708870967741936e-05,
+      "loss": 0.1756,
+      "step": 14706
+    },
+    {
+      "epoch": 0.235312,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.708709677419356e-05,
+      "loss": 0.1559,
+      "step": 14707
+    },
+    {
+      "epoch": 0.235328,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.708548387096775e-05,
+      "loss": 0.2052,
+      "step": 14708
+    },
+    {
+      "epoch": 0.235344,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.708387096774195e-05,
+      "loss": 0.1757,
+      "step": 14709
+    },
+    {
+      "epoch": 0.23536,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.708225806451613e-05,
+      "loss": 0.1693,
+      "step": 14710
+    },
+    {
+      "epoch": 0.235376,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.708064516129032e-05,
+      "loss": 0.1391,
+      "step": 14711
+    },
+    {
+      "epoch": 0.235392,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.707903225806452e-05,
+      "loss": 0.1721,
+      "step": 14712
+    },
+    {
+      "epoch": 0.235408,
+      "grad_norm": 1.296875,
+      "learning_rate": 7.70774193548387e-05,
+      "loss": 0.1735,
+      "step": 14713
+    },
+    {
+      "epoch": 0.235424,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.70758064516129e-05,
+      "loss": 0.1684,
+      "step": 14714
+    },
+    {
+      "epoch": 0.23544,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.70741935483871e-05,
+      "loss": 0.1554,
+      "step": 14715
+    },
+    {
+      "epoch": 0.235456,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.70725806451613e-05,
+      "loss": 0.2033,
+      "step": 14716
+    },
+    {
+      "epoch": 0.235472,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.707096774193549e-05,
+      "loss": 0.1515,
+      "step": 14717
+    },
+    {
+      "epoch": 0.235488,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.706935483870969e-05,
+      "loss": 0.1648,
+      "step": 14718
+    },
+    {
+      "epoch": 0.235504,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.706774193548387e-05,
+      "loss": 0.1385,
+      "step": 14719
+    },
+    {
+      "epoch": 0.23552,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.706612903225807e-05,
+      "loss": 0.1809,
+      "step": 14720
+    },
+    {
+      "epoch": 0.235536,
+      "grad_norm": 0.462890625,
+      "learning_rate": 7.706451612903226e-05,
+      "loss": 0.1535,
+      "step": 14721
+    },
+    {
+      "epoch": 0.235552,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.706290322580646e-05,
+      "loss": 0.1474,
+      "step": 14722
+    },
+    {
+      "epoch": 0.235568,
+      "grad_norm": 1.3828125,
+      "learning_rate": 7.706129032258065e-05,
+      "loss": 0.2057,
+      "step": 14723
+    },
+    {
+      "epoch": 0.235584,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.705967741935485e-05,
+      "loss": 0.1887,
+      "step": 14724
+    },
+    {
+      "epoch": 0.2356,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.705806451612903e-05,
+      "loss": 0.1889,
+      "step": 14725
+    },
+    {
+      "epoch": 0.235616,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.705645161290322e-05,
+      "loss": 0.1697,
+      "step": 14726
+    },
+    {
+      "epoch": 0.235632,
+      "grad_norm": 1.265625,
+      "learning_rate": 7.705483870967742e-05,
+      "loss": 0.2175,
+      "step": 14727
+    },
+    {
+      "epoch": 0.235648,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.705322580645162e-05,
+      "loss": 0.1848,
+      "step": 14728
+    },
+    {
+      "epoch": 0.235664,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.705161290322582e-05,
+      "loss": 0.1813,
+      "step": 14729
+    },
+    {
+      "epoch": 0.23568,
+      "grad_norm": 0.75,
+      "learning_rate": 7.705e-05,
+      "loss": 0.1894,
+      "step": 14730
+    },
+    {
+      "epoch": 0.235696,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.70483870967742e-05,
+      "loss": 0.1982,
+      "step": 14731
+    },
+    {
+      "epoch": 0.235712,
+      "grad_norm": 0.75,
+      "learning_rate": 7.704677419354839e-05,
+      "loss": 0.1871,
+      "step": 14732
+    },
+    {
+      "epoch": 0.235728,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.704516129032259e-05,
+      "loss": 0.1384,
+      "step": 14733
+    },
+    {
+      "epoch": 0.235744,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.704354838709677e-05,
+      "loss": 0.1712,
+      "step": 14734
+    },
+    {
+      "epoch": 0.23576,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.704193548387097e-05,
+      "loss": 0.1505,
+      "step": 14735
+    },
+    {
+      "epoch": 0.235776,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.704032258064516e-05,
+      "loss": 0.2143,
+      "step": 14736
+    },
+    {
+      "epoch": 0.235792,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.703870967741936e-05,
+      "loss": 0.1759,
+      "step": 14737
+    },
+    {
+      "epoch": 0.235808,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.703709677419355e-05,
+      "loss": 0.1659,
+      "step": 14738
+    },
+    {
+      "epoch": 0.235824,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.703548387096775e-05,
+      "loss": 0.1855,
+      "step": 14739
+    },
+    {
+      "epoch": 0.23584,
+      "grad_norm": 0.490234375,
+      "learning_rate": 7.703387096774194e-05,
+      "loss": 0.1665,
+      "step": 14740
+    },
+    {
+      "epoch": 0.235856,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.703225806451613e-05,
+      "loss": 0.1864,
+      "step": 14741
+    },
+    {
+      "epoch": 0.235872,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.703064516129033e-05,
+      "loss": 0.1848,
+      "step": 14742
+    },
+    {
+      "epoch": 0.235888,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.702903225806452e-05,
+      "loss": 0.1618,
+      "step": 14743
+    },
+    {
+      "epoch": 0.235904,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.702741935483872e-05,
+      "loss": 0.1868,
+      "step": 14744
+    },
+    {
+      "epoch": 0.23592,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.70258064516129e-05,
+      "loss": 0.176,
+      "step": 14745
+    },
+    {
+      "epoch": 0.235936,
+      "grad_norm": 1.2421875,
+      "learning_rate": 7.70241935483871e-05,
+      "loss": 0.1566,
+      "step": 14746
+    },
+    {
+      "epoch": 0.235952,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.702258064516129e-05,
+      "loss": 0.1675,
+      "step": 14747
+    },
+    {
+      "epoch": 0.235968,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.702096774193549e-05,
+      "loss": 0.2197,
+      "step": 14748
+    },
+    {
+      "epoch": 0.235984,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.701935483870967e-05,
+      "loss": 0.1695,
+      "step": 14749
+    },
+    {
+      "epoch": 0.236,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.701774193548387e-05,
+      "loss": 0.1549,
+      "step": 14750
+    },
+    {
+      "epoch": 0.236016,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.701612903225807e-05,
+      "loss": 0.1902,
+      "step": 14751
+    },
+    {
+      "epoch": 0.236032,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.701451612903227e-05,
+      "loss": 0.1828,
+      "step": 14752
+    },
+    {
+      "epoch": 0.236048,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.701290322580646e-05,
+      "loss": 0.1586,
+      "step": 14753
+    },
+    {
+      "epoch": 0.236064,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.701129032258066e-05,
+      "loss": 0.1627,
+      "step": 14754
+    },
+    {
+      "epoch": 0.23608,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.700967741935484e-05,
+      "loss": 0.1564,
+      "step": 14755
+    },
+    {
+      "epoch": 0.236096,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.700806451612904e-05,
+      "loss": 0.1705,
+      "step": 14756
+    },
+    {
+      "epoch": 0.236112,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.700645161290323e-05,
+      "loss": 0.1701,
+      "step": 14757
+    },
+    {
+      "epoch": 0.236128,
+      "grad_norm": 1.8671875,
+      "learning_rate": 7.700483870967742e-05,
+      "loss": 0.1633,
+      "step": 14758
+    },
+    {
+      "epoch": 0.236144,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.700322580645162e-05,
+      "loss": 0.1997,
+      "step": 14759
+    },
+    {
+      "epoch": 0.23616,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.70016129032258e-05,
+      "loss": 0.1816,
+      "step": 14760
+    },
+    {
+      "epoch": 0.236176,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.7e-05,
+      "loss": 0.1662,
+      "step": 14761
+    },
+    {
+      "epoch": 0.236192,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.699838709677419e-05,
+      "loss": 0.154,
+      "step": 14762
+    },
+    {
+      "epoch": 0.236208,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.699677419354839e-05,
+      "loss": 0.1564,
+      "step": 14763
+    },
+    {
+      "epoch": 0.236224,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.699516129032259e-05,
+      "loss": 0.1876,
+      "step": 14764
+    },
+    {
+      "epoch": 0.23624,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.699354838709679e-05,
+      "loss": 0.1602,
+      "step": 14765
+    },
+    {
+      "epoch": 0.236256,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.699193548387097e-05,
+      "loss": 0.1628,
+      "step": 14766
+    },
+    {
+      "epoch": 0.236272,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.699032258064517e-05,
+      "loss": 0.1659,
+      "step": 14767
+    },
+    {
+      "epoch": 0.236288,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.698870967741936e-05,
+      "loss": 0.1948,
+      "step": 14768
+    },
+    {
+      "epoch": 0.236304,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.698709677419356e-05,
+      "loss": 0.162,
+      "step": 14769
+    },
+    {
+      "epoch": 0.23632,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.698548387096774e-05,
+      "loss": 0.219,
+      "step": 14770
+    },
+    {
+      "epoch": 0.236336,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.698387096774194e-05,
+      "loss": 0.1707,
+      "step": 14771
+    },
+    {
+      "epoch": 0.236352,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.698225806451613e-05,
+      "loss": 0.1905,
+      "step": 14772
+    },
+    {
+      "epoch": 0.236368,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.698064516129032e-05,
+      "loss": 0.1869,
+      "step": 14773
+    },
+    {
+      "epoch": 0.236384,
+      "grad_norm": 0.75,
+      "learning_rate": 7.697903225806452e-05,
+      "loss": 0.1595,
+      "step": 14774
+    },
+    {
+      "epoch": 0.2364,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.697741935483871e-05,
+      "loss": 0.1765,
+      "step": 14775
+    },
+    {
+      "epoch": 0.236416,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.697580645161291e-05,
+      "loss": 0.1298,
+      "step": 14776
+    },
+    {
+      "epoch": 0.236432,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.69741935483871e-05,
+      "loss": 0.1925,
+      "step": 14777
+    },
+    {
+      "epoch": 0.236448,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.69725806451613e-05,
+      "loss": 0.1644,
+      "step": 14778
+    },
+    {
+      "epoch": 0.236464,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.697096774193549e-05,
+      "loss": 0.1728,
+      "step": 14779
+    },
+    {
+      "epoch": 0.23648,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.696935483870969e-05,
+      "loss": 0.1497,
+      "step": 14780
+    },
+    {
+      "epoch": 0.236496,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.696774193548387e-05,
+      "loss": 0.1492,
+      "step": 14781
+    },
+    {
+      "epoch": 0.236512,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.696612903225807e-05,
+      "loss": 0.1895,
+      "step": 14782
+    },
+    {
+      "epoch": 0.236528,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.696451612903226e-05,
+      "loss": 0.1605,
+      "step": 14783
+    },
+    {
+      "epoch": 0.236544,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.696290322580646e-05,
+      "loss": 0.1898,
+      "step": 14784
+    },
+    {
+      "epoch": 0.23656,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.696129032258064e-05,
+      "loss": 0.1498,
+      "step": 14785
+    },
+    {
+      "epoch": 0.236576,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.695967741935484e-05,
+      "loss": 0.1509,
+      "step": 14786
+    },
+    {
+      "epoch": 0.236592,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.695806451612903e-05,
+      "loss": 0.1988,
+      "step": 14787
+    },
+    {
+      "epoch": 0.236608,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.695645161290323e-05,
+      "loss": 0.1858,
+      "step": 14788
+    },
+    {
+      "epoch": 0.236624,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.695483870967743e-05,
+      "loss": 0.176,
+      "step": 14789
+    },
+    {
+      "epoch": 0.23664,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.695322580645161e-05,
+      "loss": 0.1528,
+      "step": 14790
+    },
+    {
+      "epoch": 0.236656,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.695161290322581e-05,
+      "loss": 0.1809,
+      "step": 14791
+    },
+    {
+      "epoch": 0.236672,
+      "grad_norm": 1.359375,
+      "learning_rate": 7.695e-05,
+      "loss": 0.1917,
+      "step": 14792
+    },
+    {
+      "epoch": 0.236688,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.69483870967742e-05,
+      "loss": 0.1333,
+      "step": 14793
+    },
+    {
+      "epoch": 0.236704,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.694677419354839e-05,
+      "loss": 0.181,
+      "step": 14794
+    },
+    {
+      "epoch": 0.23672,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.694516129032259e-05,
+      "loss": 0.1675,
+      "step": 14795
+    },
+    {
+      "epoch": 0.236736,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.694354838709677e-05,
+      "loss": 0.1735,
+      "step": 14796
+    },
+    {
+      "epoch": 0.236752,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.694193548387097e-05,
+      "loss": 0.1625,
+      "step": 14797
+    },
+    {
+      "epoch": 0.236768,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.694032258064516e-05,
+      "loss": 0.2483,
+      "step": 14798
+    },
+    {
+      "epoch": 0.236784,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.693870967741936e-05,
+      "loss": 0.1423,
+      "step": 14799
+    },
+    {
+      "epoch": 0.2368,
+      "grad_norm": 0.75,
+      "learning_rate": 7.693709677419356e-05,
+      "loss": 0.1822,
+      "step": 14800
+    },
+    {
+      "epoch": 0.236816,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.693548387096776e-05,
+      "loss": 0.1538,
+      "step": 14801
+    },
+    {
+      "epoch": 0.236832,
+      "grad_norm": 0.75,
+      "learning_rate": 7.693387096774194e-05,
+      "loss": 0.1928,
+      "step": 14802
+    },
+    {
+      "epoch": 0.236848,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.693225806451613e-05,
+      "loss": 0.1987,
+      "step": 14803
+    },
+    {
+      "epoch": 0.236864,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.693064516129033e-05,
+      "loss": 0.1815,
+      "step": 14804
+    },
+    {
+      "epoch": 0.23688,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.692903225806451e-05,
+      "loss": 0.1635,
+      "step": 14805
+    },
+    {
+      "epoch": 0.236896,
+      "grad_norm": 1.7421875,
+      "learning_rate": 7.692741935483871e-05,
+      "loss": 0.1837,
+      "step": 14806
+    },
+    {
+      "epoch": 0.236912,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.69258064516129e-05,
+      "loss": 0.1633,
+      "step": 14807
+    },
+    {
+      "epoch": 0.236928,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.69241935483871e-05,
+      "loss": 0.1931,
+      "step": 14808
+    },
+    {
+      "epoch": 0.236944,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.692258064516129e-05,
+      "loss": 0.1486,
+      "step": 14809
+    },
+    {
+      "epoch": 0.23696,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.692096774193549e-05,
+      "loss": 0.2121,
+      "step": 14810
+    },
+    {
+      "epoch": 0.236976,
+      "grad_norm": 0.75,
+      "learning_rate": 7.691935483870968e-05,
+      "loss": 0.1883,
+      "step": 14811
+    },
+    {
+      "epoch": 0.236992,
+      "grad_norm": 0.875,
+      "learning_rate": 7.691774193548388e-05,
+      "loss": 0.163,
+      "step": 14812
+    },
+    {
+      "epoch": 0.237008,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.691612903225807e-05,
+      "loss": 0.1864,
+      "step": 14813
+    },
+    {
+      "epoch": 0.237024,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.691451612903227e-05,
+      "loss": 0.1749,
+      "step": 14814
+    },
+    {
+      "epoch": 0.23704,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.691290322580646e-05,
+      "loss": 0.1479,
+      "step": 14815
+    },
+    {
+      "epoch": 0.237056,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.691129032258066e-05,
+      "loss": 0.2004,
+      "step": 14816
+    },
+    {
+      "epoch": 0.237072,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.690967741935484e-05,
+      "loss": 0.1919,
+      "step": 14817
+    },
+    {
+      "epoch": 0.237088,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.690806451612904e-05,
+      "loss": 0.2015,
+      "step": 14818
+    },
+    {
+      "epoch": 0.237104,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.690645161290323e-05,
+      "loss": 0.1721,
+      "step": 14819
+    },
+    {
+      "epoch": 0.23712,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.690483870967741e-05,
+      "loss": 0.1445,
+      "step": 14820
+    },
+    {
+      "epoch": 0.237136,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.690322580645161e-05,
+      "loss": 0.1612,
+      "step": 14821
+    },
+    {
+      "epoch": 0.237152,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.69016129032258e-05,
+      "loss": 0.2096,
+      "step": 14822
+    },
+    {
+      "epoch": 0.237168,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.69e-05,
+      "loss": 0.1314,
+      "step": 14823
+    },
+    {
+      "epoch": 0.237184,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.68983870967742e-05,
+      "loss": 0.1292,
+      "step": 14824
+    },
+    {
+      "epoch": 0.2372,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.68967741935484e-05,
+      "loss": 0.1682,
+      "step": 14825
+    },
+    {
+      "epoch": 0.237216,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.689516129032258e-05,
+      "loss": 0.1977,
+      "step": 14826
+    },
+    {
+      "epoch": 0.237232,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.689354838709678e-05,
+      "loss": 0.2006,
+      "step": 14827
+    },
+    {
+      "epoch": 0.237248,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.689193548387097e-05,
+      "loss": 0.2095,
+      "step": 14828
+    },
+    {
+      "epoch": 0.237264,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.689032258064517e-05,
+      "loss": 0.1672,
+      "step": 14829
+    },
+    {
+      "epoch": 0.23728,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.688870967741936e-05,
+      "loss": 0.2265,
+      "step": 14830
+    },
+    {
+      "epoch": 0.237296,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.688709677419356e-05,
+      "loss": 0.1714,
+      "step": 14831
+    },
+    {
+      "epoch": 0.237312,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.688548387096774e-05,
+      "loss": 0.1558,
+      "step": 14832
+    },
+    {
+      "epoch": 0.237328,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.688387096774194e-05,
+      "loss": 0.1621,
+      "step": 14833
+    },
+    {
+      "epoch": 0.237344,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.688225806451613e-05,
+      "loss": 0.143,
+      "step": 14834
+    },
+    {
+      "epoch": 0.23736,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.688064516129033e-05,
+      "loss": 0.216,
+      "step": 14835
+    },
+    {
+      "epoch": 0.237376,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.687903225806453e-05,
+      "loss": 0.1668,
+      "step": 14836
+    },
+    {
+      "epoch": 0.237392,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.687741935483871e-05,
+      "loss": 0.1932,
+      "step": 14837
+    },
+    {
+      "epoch": 0.237408,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.687580645161291e-05,
+      "loss": 0.1196,
+      "step": 14838
+    },
+    {
+      "epoch": 0.237424,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.68741935483871e-05,
+      "loss": 0.2015,
+      "step": 14839
+    },
+    {
+      "epoch": 0.23744,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.68725806451613e-05,
+      "loss": 0.1805,
+      "step": 14840
+    },
+    {
+      "epoch": 0.237456,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.687096774193548e-05,
+      "loss": 0.1669,
+      "step": 14841
+    },
+    {
+      "epoch": 0.237472,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.686935483870968e-05,
+      "loss": 0.1539,
+      "step": 14842
+    },
+    {
+      "epoch": 0.237488,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.686774193548387e-05,
+      "loss": 0.147,
+      "step": 14843
+    },
+    {
+      "epoch": 0.237504,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.686612903225807e-05,
+      "loss": 0.152,
+      "step": 14844
+    },
+    {
+      "epoch": 0.23752,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.686451612903226e-05,
+      "loss": 0.1678,
+      "step": 14845
+    },
+    {
+      "epoch": 0.237536,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.686290322580645e-05,
+      "loss": 0.1557,
+      "step": 14846
+    },
+    {
+      "epoch": 0.237552,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.686129032258065e-05,
+      "loss": 0.1697,
+      "step": 14847
+    },
+    {
+      "epoch": 0.237568,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.685967741935485e-05,
+      "loss": 0.1683,
+      "step": 14848
+    },
+    {
+      "epoch": 0.237584,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.685806451612904e-05,
+      "loss": 0.1742,
+      "step": 14849
+    },
+    {
+      "epoch": 0.2376,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.685645161290323e-05,
+      "loss": 0.1357,
+      "step": 14850
+    },
+    {
+      "epoch": 0.237616,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.685483870967743e-05,
+      "loss": 0.16,
+      "step": 14851
+    },
+    {
+      "epoch": 0.237632,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.685322580645161e-05,
+      "loss": 0.1442,
+      "step": 14852
+    },
+    {
+      "epoch": 0.237648,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.685161290322581e-05,
+      "loss": 0.1637,
+      "step": 14853
+    },
+    {
+      "epoch": 0.237664,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.685e-05,
+      "loss": 0.1722,
+      "step": 14854
+    },
+    {
+      "epoch": 0.23768,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.68483870967742e-05,
+      "loss": 0.1708,
+      "step": 14855
+    },
+    {
+      "epoch": 0.237696,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.684677419354838e-05,
+      "loss": 0.1722,
+      "step": 14856
+    },
+    {
+      "epoch": 0.237712,
+      "grad_norm": 1.0,
+      "learning_rate": 7.684516129032258e-05,
+      "loss": 0.1976,
+      "step": 14857
+    },
+    {
+      "epoch": 0.237728,
+      "grad_norm": 1.4453125,
+      "learning_rate": 7.684354838709677e-05,
+      "loss": 0.1795,
+      "step": 14858
+    },
+    {
+      "epoch": 0.237744,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.684193548387097e-05,
+      "loss": 0.1888,
+      "step": 14859
+    },
+    {
+      "epoch": 0.23776,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.684032258064517e-05,
+      "loss": 0.1499,
+      "step": 14860
+    },
+    {
+      "epoch": 0.237776,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.683870967741937e-05,
+      "loss": 0.1765,
+      "step": 14861
+    },
+    {
+      "epoch": 0.237792,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.683709677419355e-05,
+      "loss": 0.155,
+      "step": 14862
+    },
+    {
+      "epoch": 0.237808,
+      "grad_norm": 1.2734375,
+      "learning_rate": 7.683548387096775e-05,
+      "loss": 0.2298,
+      "step": 14863
+    },
+    {
+      "epoch": 0.237824,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.683387096774194e-05,
+      "loss": 0.1694,
+      "step": 14864
+    },
+    {
+      "epoch": 0.23784,
+      "grad_norm": 1.0,
+      "learning_rate": 7.683225806451613e-05,
+      "loss": 0.1654,
+      "step": 14865
+    },
+    {
+      "epoch": 0.237856,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.683064516129033e-05,
+      "loss": 0.2177,
+      "step": 14866
+    },
+    {
+      "epoch": 0.237872,
+      "grad_norm": 0.4453125,
+      "learning_rate": 7.682903225806451e-05,
+      "loss": 0.1297,
+      "step": 14867
+    },
+    {
+      "epoch": 0.237888,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.682741935483871e-05,
+      "loss": 0.2323,
+      "step": 14868
+    },
+    {
+      "epoch": 0.237904,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.68258064516129e-05,
+      "loss": 0.202,
+      "step": 14869
+    },
+    {
+      "epoch": 0.23792,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.68241935483871e-05,
+      "loss": 0.2022,
+      "step": 14870
+    },
+    {
+      "epoch": 0.237936,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.68225806451613e-05,
+      "loss": 0.2234,
+      "step": 14871
+    },
+    {
+      "epoch": 0.237952,
+      "grad_norm": 1.28125,
+      "learning_rate": 7.68209677419355e-05,
+      "loss": 0.1678,
+      "step": 14872
+    },
+    {
+      "epoch": 0.237968,
+      "grad_norm": 1.0,
+      "learning_rate": 7.681935483870968e-05,
+      "loss": 0.2161,
+      "step": 14873
+    },
+    {
+      "epoch": 0.237984,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.681774193548388e-05,
+      "loss": 0.186,
+      "step": 14874
+    },
+    {
+      "epoch": 0.238,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.681612903225807e-05,
+      "loss": 0.1406,
+      "step": 14875
+    },
+    {
+      "epoch": 0.238016,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.681451612903227e-05,
+      "loss": 0.1958,
+      "step": 14876
+    },
+    {
+      "epoch": 0.238032,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.681290322580645e-05,
+      "loss": 0.1776,
+      "step": 14877
+    },
+    {
+      "epoch": 0.238048,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.681129032258065e-05,
+      "loss": 0.1848,
+      "step": 14878
+    },
+    {
+      "epoch": 0.238064,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.680967741935484e-05,
+      "loss": 0.1438,
+      "step": 14879
+    },
+    {
+      "epoch": 0.23808,
+      "grad_norm": 1.328125,
+      "learning_rate": 7.680806451612904e-05,
+      "loss": 0.157,
+      "step": 14880
+    },
+    {
+      "epoch": 0.238096,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.680645161290323e-05,
+      "loss": 0.1724,
+      "step": 14881
+    },
+    {
+      "epoch": 0.238112,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.680483870967742e-05,
+      "loss": 0.1775,
+      "step": 14882
+    },
+    {
+      "epoch": 0.238128,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.680322580645161e-05,
+      "loss": 0.2174,
+      "step": 14883
+    },
+    {
+      "epoch": 0.238144,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.680161290322581e-05,
+      "loss": 0.1553,
+      "step": 14884
+    },
+    {
+      "epoch": 0.23816,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.680000000000001e-05,
+      "loss": 0.1651,
+      "step": 14885
+    },
+    {
+      "epoch": 0.238176,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.67983870967742e-05,
+      "loss": 0.2167,
+      "step": 14886
+    },
+    {
+      "epoch": 0.238192,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.67967741935484e-05,
+      "loss": 0.1669,
+      "step": 14887
+    },
+    {
+      "epoch": 0.238208,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.679516129032258e-05,
+      "loss": 0.2037,
+      "step": 14888
+    },
+    {
+      "epoch": 0.238224,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.679354838709678e-05,
+      "loss": 0.1729,
+      "step": 14889
+    },
+    {
+      "epoch": 0.23824,
+      "grad_norm": 1.2890625,
+      "learning_rate": 7.679193548387097e-05,
+      "loss": 0.1914,
+      "step": 14890
+    },
+    {
+      "epoch": 0.238256,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.679032258064517e-05,
+      "loss": 0.1706,
+      "step": 14891
+    },
+    {
+      "epoch": 0.238272,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.678870967741935e-05,
+      "loss": 0.1972,
+      "step": 14892
+    },
+    {
+      "epoch": 0.238288,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.678709677419355e-05,
+      "loss": 0.1646,
+      "step": 14893
+    },
+    {
+      "epoch": 0.238304,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.678548387096774e-05,
+      "loss": 0.1781,
+      "step": 14894
+    },
+    {
+      "epoch": 0.23832,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.678387096774194e-05,
+      "loss": 0.1567,
+      "step": 14895
+    },
+    {
+      "epoch": 0.238336,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.678225806451614e-05,
+      "loss": 0.1776,
+      "step": 14896
+    },
+    {
+      "epoch": 0.238352,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.678064516129032e-05,
+      "loss": 0.1928,
+      "step": 14897
+    },
+    {
+      "epoch": 0.238368,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.677903225806452e-05,
+      "loss": 0.2126,
+      "step": 14898
+    },
+    {
+      "epoch": 0.238384,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.677741935483871e-05,
+      "loss": 0.1798,
+      "step": 14899
+    },
+    {
+      "epoch": 0.2384,
+      "grad_norm": 1.2109375,
+      "learning_rate": 7.677580645161291e-05,
+      "loss": 0.2121,
+      "step": 14900
+    },
+    {
+      "epoch": 0.238416,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.67741935483871e-05,
+      "loss": 0.1772,
+      "step": 14901
+    },
+    {
+      "epoch": 0.238432,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.67725806451613e-05,
+      "loss": 0.169,
+      "step": 14902
+    },
+    {
+      "epoch": 0.238448,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.677096774193548e-05,
+      "loss": 0.1873,
+      "step": 14903
+    },
+    {
+      "epoch": 0.238464,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.676935483870968e-05,
+      "loss": 0.1544,
+      "step": 14904
+    },
+    {
+      "epoch": 0.23848,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.676774193548387e-05,
+      "loss": 0.176,
+      "step": 14905
+    },
+    {
+      "epoch": 0.238496,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.676612903225807e-05,
+      "loss": 0.1357,
+      "step": 14906
+    },
+    {
+      "epoch": 0.238512,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.676451612903227e-05,
+      "loss": 0.1482,
+      "step": 14907
+    },
+    {
+      "epoch": 0.238528,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.676290322580647e-05,
+      "loss": 0.1732,
+      "step": 14908
+    },
+    {
+      "epoch": 0.238544,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.676129032258065e-05,
+      "loss": 0.1969,
+      "step": 14909
+    },
+    {
+      "epoch": 0.23856,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.675967741935485e-05,
+      "loss": 0.1558,
+      "step": 14910
+    },
+    {
+      "epoch": 0.238576,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.675806451612904e-05,
+      "loss": 0.2007,
+      "step": 14911
+    },
+    {
+      "epoch": 0.238592,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.675645161290322e-05,
+      "loss": 0.1671,
+      "step": 14912
+    },
+    {
+      "epoch": 0.238608,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.675483870967742e-05,
+      "loss": 0.2014,
+      "step": 14913
+    },
+    {
+      "epoch": 0.238624,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.675322580645161e-05,
+      "loss": 0.1643,
+      "step": 14914
+    },
+    {
+      "epoch": 0.23864,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.675161290322581e-05,
+      "loss": 0.1873,
+      "step": 14915
+    },
+    {
+      "epoch": 0.238656,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.675e-05,
+      "loss": 0.2037,
+      "step": 14916
+    },
+    {
+      "epoch": 0.238672,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.67483870967742e-05,
+      "loss": 0.1497,
+      "step": 14917
+    },
+    {
+      "epoch": 0.238688,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.674677419354838e-05,
+      "loss": 0.1612,
+      "step": 14918
+    },
+    {
+      "epoch": 0.238704,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.674516129032258e-05,
+      "loss": 0.1891,
+      "step": 14919
+    },
+    {
+      "epoch": 0.23872,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.674354838709678e-05,
+      "loss": 0.1884,
+      "step": 14920
+    },
+    {
+      "epoch": 0.238736,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.674193548387098e-05,
+      "loss": 0.1986,
+      "step": 14921
+    },
+    {
+      "epoch": 0.238752,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.674032258064517e-05,
+      "loss": 0.2226,
+      "step": 14922
+    },
+    {
+      "epoch": 0.238768,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.673870967741937e-05,
+      "loss": 0.1585,
+      "step": 14923
+    },
+    {
+      "epoch": 0.238784,
+      "grad_norm": 1.4765625,
+      "learning_rate": 7.673709677419355e-05,
+      "loss": 0.1919,
+      "step": 14924
+    },
+    {
+      "epoch": 0.2388,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.673548387096775e-05,
+      "loss": 0.173,
+      "step": 14925
+    },
+    {
+      "epoch": 0.238816,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.673387096774194e-05,
+      "loss": 0.1812,
+      "step": 14926
+    },
+    {
+      "epoch": 0.238832,
+      "grad_norm": 1.4921875,
+      "learning_rate": 7.673225806451614e-05,
+      "loss": 0.2004,
+      "step": 14927
+    },
+    {
+      "epoch": 0.238848,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.673064516129032e-05,
+      "loss": 0.146,
+      "step": 14928
+    },
+    {
+      "epoch": 0.238864,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.672903225806451e-05,
+      "loss": 0.1615,
+      "step": 14929
+    },
+    {
+      "epoch": 0.23888,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.672741935483871e-05,
+      "loss": 0.1781,
+      "step": 14930
+    },
+    {
+      "epoch": 0.238896,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.672580645161291e-05,
+      "loss": 0.1812,
+      "step": 14931
+    },
+    {
+      "epoch": 0.238912,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.672419354838711e-05,
+      "loss": 0.2127,
+      "step": 14932
+    },
+    {
+      "epoch": 0.238928,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.67225806451613e-05,
+      "loss": 0.1572,
+      "step": 14933
+    },
+    {
+      "epoch": 0.238944,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.67209677419355e-05,
+      "loss": 0.1738,
+      "step": 14934
+    },
+    {
+      "epoch": 0.23896,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.671935483870968e-05,
+      "loss": 0.1578,
+      "step": 14935
+    },
+    {
+      "epoch": 0.238976,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.671774193548388e-05,
+      "loss": 0.1529,
+      "step": 14936
+    },
+    {
+      "epoch": 0.238992,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.671612903225807e-05,
+      "loss": 0.178,
+      "step": 14937
+    },
+    {
+      "epoch": 0.239008,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.671451612903227e-05,
+      "loss": 0.1853,
+      "step": 14938
+    },
+    {
+      "epoch": 0.239024,
+      "grad_norm": 0.46484375,
+      "learning_rate": 7.671290322580645e-05,
+      "loss": 0.1618,
+      "step": 14939
+    },
+    {
+      "epoch": 0.23904,
+      "grad_norm": 1.125,
+      "learning_rate": 7.671129032258065e-05,
+      "loss": 0.1929,
+      "step": 14940
+    },
+    {
+      "epoch": 0.239056,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.670967741935484e-05,
+      "loss": 0.139,
+      "step": 14941
+    },
+    {
+      "epoch": 0.239072,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.670806451612904e-05,
+      "loss": 0.1697,
+      "step": 14942
+    },
+    {
+      "epoch": 0.239088,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.670645161290324e-05,
+      "loss": 0.1887,
+      "step": 14943
+    },
+    {
+      "epoch": 0.239104,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.670483870967742e-05,
+      "loss": 0.1466,
+      "step": 14944
+    },
+    {
+      "epoch": 0.23912,
+      "grad_norm": 0.75,
+      "learning_rate": 7.670322580645162e-05,
+      "loss": 0.1822,
+      "step": 14945
+    },
+    {
+      "epoch": 0.239136,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.670161290322581e-05,
+      "loss": 0.1403,
+      "step": 14946
+    },
+    {
+      "epoch": 0.239152,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.670000000000001e-05,
+      "loss": 0.146,
+      "step": 14947
+    },
+    {
+      "epoch": 0.239168,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.66983870967742e-05,
+      "loss": 0.1882,
+      "step": 14948
+    },
+    {
+      "epoch": 0.239184,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.66967741935484e-05,
+      "loss": 0.1526,
+      "step": 14949
+    },
+    {
+      "epoch": 0.2392,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.669516129032258e-05,
+      "loss": 0.1735,
+      "step": 14950
+    },
+    {
+      "epoch": 0.239216,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.669354838709678e-05,
+      "loss": 0.216,
+      "step": 14951
+    },
+    {
+      "epoch": 0.239232,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.669193548387097e-05,
+      "loss": 0.1812,
+      "step": 14952
+    },
+    {
+      "epoch": 0.239248,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.669032258064516e-05,
+      "loss": 0.1487,
+      "step": 14953
+    },
+    {
+      "epoch": 0.239264,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.668870967741935e-05,
+      "loss": 0.1599,
+      "step": 14954
+    },
+    {
+      "epoch": 0.23928,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.668709677419355e-05,
+      "loss": 0.1553,
+      "step": 14955
+    },
+    {
+      "epoch": 0.239296,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.668548387096775e-05,
+      "loss": 0.1615,
+      "step": 14956
+    },
+    {
+      "epoch": 0.239312,
+      "grad_norm": 1.359375,
+      "learning_rate": 7.668387096774195e-05,
+      "loss": 0.1918,
+      "step": 14957
+    },
+    {
+      "epoch": 0.239328,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.668225806451614e-05,
+      "loss": 0.1477,
+      "step": 14958
+    },
+    {
+      "epoch": 0.239344,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.668064516129032e-05,
+      "loss": 0.1637,
+      "step": 14959
+    },
+    {
+      "epoch": 0.23936,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.667903225806452e-05,
+      "loss": 0.1619,
+      "step": 14960
+    },
+    {
+      "epoch": 0.239376,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.667741935483871e-05,
+      "loss": 0.1738,
+      "step": 14961
+    },
+    {
+      "epoch": 0.239392,
+      "grad_norm": 1.28125,
+      "learning_rate": 7.667580645161291e-05,
+      "loss": 0.1565,
+      "step": 14962
+    },
+    {
+      "epoch": 0.239408,
+      "grad_norm": 1.125,
+      "learning_rate": 7.66741935483871e-05,
+      "loss": 0.2039,
+      "step": 14963
+    },
+    {
+      "epoch": 0.239424,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.667258064516129e-05,
+      "loss": 0.1609,
+      "step": 14964
+    },
+    {
+      "epoch": 0.23944,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.667096774193548e-05,
+      "loss": 0.2149,
+      "step": 14965
+    },
+    {
+      "epoch": 0.239456,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.666935483870968e-05,
+      "loss": 0.1521,
+      "step": 14966
+    },
+    {
+      "epoch": 0.239472,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.666774193548388e-05,
+      "loss": 0.1628,
+      "step": 14967
+    },
+    {
+      "epoch": 0.239488,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.666612903225808e-05,
+      "loss": 0.1513,
+      "step": 14968
+    },
+    {
+      "epoch": 0.239504,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.666451612903226e-05,
+      "loss": 0.164,
+      "step": 14969
+    },
+    {
+      "epoch": 0.23952,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.666290322580646e-05,
+      "loss": 0.1765,
+      "step": 14970
+    },
+    {
+      "epoch": 0.239536,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.666129032258065e-05,
+      "loss": 0.1915,
+      "step": 14971
+    },
+    {
+      "epoch": 0.239552,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.665967741935485e-05,
+      "loss": 0.1991,
+      "step": 14972
+    },
+    {
+      "epoch": 0.239568,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.665806451612904e-05,
+      "loss": 0.1925,
+      "step": 14973
+    },
+    {
+      "epoch": 0.239584,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.665645161290322e-05,
+      "loss": 0.1744,
+      "step": 14974
+    },
+    {
+      "epoch": 0.2396,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.665483870967742e-05,
+      "loss": 0.155,
+      "step": 14975
+    },
+    {
+      "epoch": 0.239616,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.665322580645161e-05,
+      "loss": 0.1716,
+      "step": 14976
+    },
+    {
+      "epoch": 0.239632,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.665161290322581e-05,
+      "loss": 0.1593,
+      "step": 14977
+    },
+    {
+      "epoch": 0.239648,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.664999999999999e-05,
+      "loss": 0.1324,
+      "step": 14978
+    },
+    {
+      "epoch": 0.239664,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.664838709677419e-05,
+      "loss": 0.1531,
+      "step": 14979
+    },
+    {
+      "epoch": 0.23968,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.664677419354839e-05,
+      "loss": 0.1478,
+      "step": 14980
+    },
+    {
+      "epoch": 0.239696,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.664516129032259e-05,
+      "loss": 0.1792,
+      "step": 14981
+    },
+    {
+      "epoch": 0.239712,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.664354838709678e-05,
+      "loss": 0.1825,
+      "step": 14982
+    },
+    {
+      "epoch": 0.239728,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.664193548387098e-05,
+      "loss": 0.1806,
+      "step": 14983
+    },
+    {
+      "epoch": 0.239744,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.664032258064516e-05,
+      "loss": 0.1725,
+      "step": 14984
+    },
+    {
+      "epoch": 0.23976,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.663870967741936e-05,
+      "loss": 0.1744,
+      "step": 14985
+    },
+    {
+      "epoch": 0.239776,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.663709677419355e-05,
+      "loss": 0.1954,
+      "step": 14986
+    },
+    {
+      "epoch": 0.239792,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.663548387096775e-05,
+      "loss": 0.1715,
+      "step": 14987
+    },
+    {
+      "epoch": 0.239808,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.663387096774193e-05,
+      "loss": 0.1839,
+      "step": 14988
+    },
+    {
+      "epoch": 0.239824,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.663225806451613e-05,
+      "loss": 0.1509,
+      "step": 14989
+    },
+    {
+      "epoch": 0.23984,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.663064516129032e-05,
+      "loss": 0.1505,
+      "step": 14990
+    },
+    {
+      "epoch": 0.239856,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.662903225806452e-05,
+      "loss": 0.1367,
+      "step": 14991
+    },
+    {
+      "epoch": 0.239872,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.662741935483872e-05,
+      "loss": 0.1439,
+      "step": 14992
+    },
+    {
+      "epoch": 0.239888,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.66258064516129e-05,
+      "loss": 0.1544,
+      "step": 14993
+    },
+    {
+      "epoch": 0.239904,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.66241935483871e-05,
+      "loss": 0.1539,
+      "step": 14994
+    },
+    {
+      "epoch": 0.23992,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.662258064516129e-05,
+      "loss": 0.1345,
+      "step": 14995
+    },
+    {
+      "epoch": 0.239936,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.662096774193549e-05,
+      "loss": 0.1817,
+      "step": 14996
+    },
+    {
+      "epoch": 0.239952,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.661935483870968e-05,
+      "loss": 0.1247,
+      "step": 14997
+    },
+    {
+      "epoch": 0.239968,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.661774193548388e-05,
+      "loss": 0.229,
+      "step": 14998
+    },
+    {
+      "epoch": 0.239984,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.661612903225806e-05,
+      "loss": 0.1663,
+      "step": 14999
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.661451612903226e-05,
+      "loss": 0.185,
+      "step": 15000
+    },
+    {
+      "epoch": 0.240016,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.661290322580645e-05,
+      "loss": 0.1525,
+      "step": 15001
+    },
+    {
+      "epoch": 0.240032,
+      "grad_norm": 0.443359375,
+      "learning_rate": 7.661129032258065e-05,
+      "loss": 0.1523,
+      "step": 15002
+    },
+    {
+      "epoch": 0.240048,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.660967741935485e-05,
+      "loss": 0.1954,
+      "step": 15003
+    },
+    {
+      "epoch": 0.240064,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.660806451612905e-05,
+      "loss": 0.1884,
+      "step": 15004
+    },
+    {
+      "epoch": 0.24008,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.660645161290323e-05,
+      "loss": 0.1325,
+      "step": 15005
+    },
+    {
+      "epoch": 0.240096,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.660483870967742e-05,
+      "loss": 0.167,
+      "step": 15006
+    },
+    {
+      "epoch": 0.240112,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.660322580645162e-05,
+      "loss": 0.1644,
+      "step": 15007
+    },
+    {
+      "epoch": 0.240128,
+      "grad_norm": 1.2578125,
+      "learning_rate": 7.66016129032258e-05,
+      "loss": 0.1656,
+      "step": 15008
+    },
+    {
+      "epoch": 0.240144,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.66e-05,
+      "loss": 0.2127,
+      "step": 15009
+    },
+    {
+      "epoch": 0.24016,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.659838709677419e-05,
+      "loss": 0.1825,
+      "step": 15010
+    },
+    {
+      "epoch": 0.240176,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.659677419354839e-05,
+      "loss": 0.1687,
+      "step": 15011
+    },
+    {
+      "epoch": 0.240192,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.659516129032258e-05,
+      "loss": 0.187,
+      "step": 15012
+    },
+    {
+      "epoch": 0.240208,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.659354838709678e-05,
+      "loss": 0.1435,
+      "step": 15013
+    },
+    {
+      "epoch": 0.240224,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.659193548387096e-05,
+      "loss": 0.167,
+      "step": 15014
+    },
+    {
+      "epoch": 0.24024,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.659032258064516e-05,
+      "loss": 0.2134,
+      "step": 15015
+    },
+    {
+      "epoch": 0.240256,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.658870967741936e-05,
+      "loss": 0.1664,
+      "step": 15016
+    },
+    {
+      "epoch": 0.240272,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.658709677419356e-05,
+      "loss": 0.1853,
+      "step": 15017
+    },
+    {
+      "epoch": 0.240288,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.658548387096775e-05,
+      "loss": 0.1552,
+      "step": 15018
+    },
+    {
+      "epoch": 0.240304,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.658387096774195e-05,
+      "loss": 0.1974,
+      "step": 15019
+    },
+    {
+      "epoch": 0.24032,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.658225806451613e-05,
+      "loss": 0.149,
+      "step": 15020
+    },
+    {
+      "epoch": 0.240336,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.658064516129032e-05,
+      "loss": 0.1757,
+      "step": 15021
+    },
+    {
+      "epoch": 0.240352,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.657903225806452e-05,
+      "loss": 0.1914,
+      "step": 15022
+    },
+    {
+      "epoch": 0.240368,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.65774193548387e-05,
+      "loss": 0.1602,
+      "step": 15023
+    },
+    {
+      "epoch": 0.240384,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.65758064516129e-05,
+      "loss": 0.1961,
+      "step": 15024
+    },
+    {
+      "epoch": 0.2404,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.657419354838709e-05,
+      "loss": 0.163,
+      "step": 15025
+    },
+    {
+      "epoch": 0.240416,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.657258064516129e-05,
+      "loss": 0.167,
+      "step": 15026
+    },
+    {
+      "epoch": 0.240432,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.657096774193549e-05,
+      "loss": 0.1846,
+      "step": 15027
+    },
+    {
+      "epoch": 0.240448,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.656935483870969e-05,
+      "loss": 0.1937,
+      "step": 15028
+    },
+    {
+      "epoch": 0.240464,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.656774193548388e-05,
+      "loss": 0.1669,
+      "step": 15029
+    },
+    {
+      "epoch": 0.24048,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.656612903225808e-05,
+      "loss": 0.1603,
+      "step": 15030
+    },
+    {
+      "epoch": 0.240496,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.656451612903226e-05,
+      "loss": 0.1717,
+      "step": 15031
+    },
+    {
+      "epoch": 0.240512,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.656290322580646e-05,
+      "loss": 0.1671,
+      "step": 15032
+    },
+    {
+      "epoch": 0.240528,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.656129032258065e-05,
+      "loss": 0.1469,
+      "step": 15033
+    },
+    {
+      "epoch": 0.240544,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.655967741935485e-05,
+      "loss": 0.1698,
+      "step": 15034
+    },
+    {
+      "epoch": 0.24056,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.655806451612903e-05,
+      "loss": 0.1924,
+      "step": 15035
+    },
+    {
+      "epoch": 0.240576,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.655645161290323e-05,
+      "loss": 0.1458,
+      "step": 15036
+    },
+    {
+      "epoch": 0.240592,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.655483870967742e-05,
+      "loss": 0.1717,
+      "step": 15037
+    },
+    {
+      "epoch": 0.240608,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.655322580645162e-05,
+      "loss": 0.1794,
+      "step": 15038
+    },
+    {
+      "epoch": 0.240624,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.65516129032258e-05,
+      "loss": 0.138,
+      "step": 15039
+    },
+    {
+      "epoch": 0.24064,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.655e-05,
+      "loss": 0.2053,
+      "step": 15040
+    },
+    {
+      "epoch": 0.240656,
+      "grad_norm": 1.5390625,
+      "learning_rate": 7.65483870967742e-05,
+      "loss": 0.1844,
+      "step": 15041
+    },
+    {
+      "epoch": 0.240672,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.654677419354839e-05,
+      "loss": 0.1691,
+      "step": 15042
+    },
+    {
+      "epoch": 0.240688,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.654516129032259e-05,
+      "loss": 0.1937,
+      "step": 15043
+    },
+    {
+      "epoch": 0.240704,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.654354838709678e-05,
+      "loss": 0.1604,
+      "step": 15044
+    },
+    {
+      "epoch": 0.24072,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.654193548387098e-05,
+      "loss": 0.1636,
+      "step": 15045
+    },
+    {
+      "epoch": 0.240736,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.654032258064516e-05,
+      "loss": 0.1434,
+      "step": 15046
+    },
+    {
+      "epoch": 0.240752,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.653870967741936e-05,
+      "loss": 0.1454,
+      "step": 15047
+    },
+    {
+      "epoch": 0.240768,
+      "grad_norm": 1.2734375,
+      "learning_rate": 7.653709677419355e-05,
+      "loss": 0.1561,
+      "step": 15048
+    },
+    {
+      "epoch": 0.240784,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.653548387096775e-05,
+      "loss": 0.2137,
+      "step": 15049
+    },
+    {
+      "epoch": 0.2408,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.653387096774193e-05,
+      "loss": 0.1937,
+      "step": 15050
+    },
+    {
+      "epoch": 0.240816,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.653225806451613e-05,
+      "loss": 0.1698,
+      "step": 15051
+    },
+    {
+      "epoch": 0.240832,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.653064516129033e-05,
+      "loss": 0.207,
+      "step": 15052
+    },
+    {
+      "epoch": 0.240848,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.652903225806452e-05,
+      "loss": 0.1888,
+      "step": 15053
+    },
+    {
+      "epoch": 0.240864,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.652741935483872e-05,
+      "loss": 0.1639,
+      "step": 15054
+    },
+    {
+      "epoch": 0.24088,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.65258064516129e-05,
+      "loss": 0.2036,
+      "step": 15055
+    },
+    {
+      "epoch": 0.240896,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.65241935483871e-05,
+      "loss": 0.1362,
+      "step": 15056
+    },
+    {
+      "epoch": 0.240912,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.652258064516129e-05,
+      "loss": 0.1667,
+      "step": 15057
+    },
+    {
+      "epoch": 0.240928,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.652096774193549e-05,
+      "loss": 0.1513,
+      "step": 15058
+    },
+    {
+      "epoch": 0.240944,
+      "grad_norm": 0.625,
+      "learning_rate": 7.651935483870967e-05,
+      "loss": 0.1864,
+      "step": 15059
+    },
+    {
+      "epoch": 0.24096,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.651774193548387e-05,
+      "loss": 0.1776,
+      "step": 15060
+    },
+    {
+      "epoch": 0.240976,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.651612903225806e-05,
+      "loss": 0.1747,
+      "step": 15061
+    },
+    {
+      "epoch": 0.240992,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.651451612903226e-05,
+      "loss": 0.1764,
+      "step": 15062
+    },
+    {
+      "epoch": 0.241008,
+      "grad_norm": 1.4140625,
+      "learning_rate": 7.651290322580646e-05,
+      "loss": 0.1397,
+      "step": 15063
+    },
+    {
+      "epoch": 0.241024,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.651129032258066e-05,
+      "loss": 0.1864,
+      "step": 15064
+    },
+    {
+      "epoch": 0.24104,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.650967741935485e-05,
+      "loss": 0.1752,
+      "step": 15065
+    },
+    {
+      "epoch": 0.241056,
+      "grad_norm": 1.5234375,
+      "learning_rate": 7.650806451612905e-05,
+      "loss": 0.1684,
+      "step": 15066
+    },
+    {
+      "epoch": 0.241072,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.650645161290323e-05,
+      "loss": 0.134,
+      "step": 15067
+    },
+    {
+      "epoch": 0.241088,
+      "grad_norm": 1.3984375,
+      "learning_rate": 7.650483870967742e-05,
+      "loss": 0.1371,
+      "step": 15068
+    },
+    {
+      "epoch": 0.241104,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.650322580645162e-05,
+      "loss": 0.1808,
+      "step": 15069
+    },
+    {
+      "epoch": 0.24112,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.65016129032258e-05,
+      "loss": 0.1602,
+      "step": 15070
+    },
+    {
+      "epoch": 0.241136,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.65e-05,
+      "loss": 0.1802,
+      "step": 15071
+    },
+    {
+      "epoch": 0.241152,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.649838709677419e-05,
+      "loss": 0.164,
+      "step": 15072
+    },
+    {
+      "epoch": 0.241168,
+      "grad_norm": 0.625,
+      "learning_rate": 7.649677419354839e-05,
+      "loss": 0.1306,
+      "step": 15073
+    },
+    {
+      "epoch": 0.241184,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.649516129032257e-05,
+      "loss": 0.1278,
+      "step": 15074
+    },
+    {
+      "epoch": 0.2412,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.649354838709677e-05,
+      "loss": 0.1701,
+      "step": 15075
+    },
+    {
+      "epoch": 0.241216,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.649193548387097e-05,
+      "loss": 0.1872,
+      "step": 15076
+    },
+    {
+      "epoch": 0.241232,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.649032258064517e-05,
+      "loss": 0.1507,
+      "step": 15077
+    },
+    {
+      "epoch": 0.241248,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.648870967741936e-05,
+      "loss": 0.158,
+      "step": 15078
+    },
+    {
+      "epoch": 0.241264,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.648709677419356e-05,
+      "loss": 0.1835,
+      "step": 15079
+    },
+    {
+      "epoch": 0.24128,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.648548387096775e-05,
+      "loss": 0.1918,
+      "step": 15080
+    },
+    {
+      "epoch": 0.241296,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.648387096774194e-05,
+      "loss": 0.1349,
+      "step": 15081
+    },
+    {
+      "epoch": 0.241312,
+      "grad_norm": 1.3515625,
+      "learning_rate": 7.648225806451613e-05,
+      "loss": 0.2478,
+      "step": 15082
+    },
+    {
+      "epoch": 0.241328,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.648064516129032e-05,
+      "loss": 0.1772,
+      "step": 15083
+    },
+    {
+      "epoch": 0.241344,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.647903225806452e-05,
+      "loss": 0.1722,
+      "step": 15084
+    },
+    {
+      "epoch": 0.24136,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.64774193548387e-05,
+      "loss": 0.2024,
+      "step": 15085
+    },
+    {
+      "epoch": 0.241376,
+      "grad_norm": 0.625,
+      "learning_rate": 7.64758064516129e-05,
+      "loss": 0.1516,
+      "step": 15086
+    },
+    {
+      "epoch": 0.241392,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.64741935483871e-05,
+      "loss": 0.1667,
+      "step": 15087
+    },
+    {
+      "epoch": 0.241408,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.64725806451613e-05,
+      "loss": 0.2136,
+      "step": 15088
+    },
+    {
+      "epoch": 0.241424,
+      "grad_norm": 1.3203125,
+      "learning_rate": 7.647096774193549e-05,
+      "loss": 0.2195,
+      "step": 15089
+    },
+    {
+      "epoch": 0.24144,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.646935483870969e-05,
+      "loss": 0.1604,
+      "step": 15090
+    },
+    {
+      "epoch": 0.241456,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.646774193548387e-05,
+      "loss": 0.175,
+      "step": 15091
+    },
+    {
+      "epoch": 0.241472,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.646612903225807e-05,
+      "loss": 0.1487,
+      "step": 15092
+    },
+    {
+      "epoch": 0.241488,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.646451612903226e-05,
+      "loss": 0.1865,
+      "step": 15093
+    },
+    {
+      "epoch": 0.241504,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.646290322580646e-05,
+      "loss": 0.1477,
+      "step": 15094
+    },
+    {
+      "epoch": 0.24152,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.646129032258064e-05,
+      "loss": 0.1763,
+      "step": 15095
+    },
+    {
+      "epoch": 0.241536,
+      "grad_norm": 1.3359375,
+      "learning_rate": 7.645967741935484e-05,
+      "loss": 0.1554,
+      "step": 15096
+    },
+    {
+      "epoch": 0.241552,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.645806451612903e-05,
+      "loss": 0.1843,
+      "step": 15097
+    },
+    {
+      "epoch": 0.241568,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.645645161290323e-05,
+      "loss": 0.1651,
+      "step": 15098
+    },
+    {
+      "epoch": 0.241584,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.645483870967743e-05,
+      "loss": 0.1693,
+      "step": 15099
+    },
+    {
+      "epoch": 0.2416,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.645322580645162e-05,
+      "loss": 0.1728,
+      "step": 15100
+    },
+    {
+      "epoch": 0.241616,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.645161290322582e-05,
+      "loss": 0.1608,
+      "step": 15101
+    },
+    {
+      "epoch": 0.241632,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.645e-05,
+      "loss": 0.1772,
+      "step": 15102
+    },
+    {
+      "epoch": 0.241648,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.64483870967742e-05,
+      "loss": 0.1933,
+      "step": 15103
+    },
+    {
+      "epoch": 0.241664,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.644677419354839e-05,
+      "loss": 0.1716,
+      "step": 15104
+    },
+    {
+      "epoch": 0.24168,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.644516129032259e-05,
+      "loss": 0.1917,
+      "step": 15105
+    },
+    {
+      "epoch": 0.241696,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.644354838709677e-05,
+      "loss": 0.1944,
+      "step": 15106
+    },
+    {
+      "epoch": 0.241712,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.644193548387097e-05,
+      "loss": 0.1894,
+      "step": 15107
+    },
+    {
+      "epoch": 0.241728,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.644032258064516e-05,
+      "loss": 0.1532,
+      "step": 15108
+    },
+    {
+      "epoch": 0.241744,
+      "grad_norm": 0.875,
+      "learning_rate": 7.643870967741936e-05,
+      "loss": 0.1784,
+      "step": 15109
+    },
+    {
+      "epoch": 0.24176,
+      "grad_norm": 1.125,
+      "learning_rate": 7.643709677419354e-05,
+      "loss": 0.1972,
+      "step": 15110
+    },
+    {
+      "epoch": 0.241776,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.643548387096774e-05,
+      "loss": 0.1653,
+      "step": 15111
+    },
+    {
+      "epoch": 0.241792,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.643387096774194e-05,
+      "loss": 0.1573,
+      "step": 15112
+    },
+    {
+      "epoch": 0.241808,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.643225806451614e-05,
+      "loss": 0.1764,
+      "step": 15113
+    },
+    {
+      "epoch": 0.241824,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.643064516129033e-05,
+      "loss": 0.1813,
+      "step": 15114
+    },
+    {
+      "epoch": 0.24184,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.642903225806452e-05,
+      "loss": 0.191,
+      "step": 15115
+    },
+    {
+      "epoch": 0.241856,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.642741935483872e-05,
+      "loss": 0.1592,
+      "step": 15116
+    },
+    {
+      "epoch": 0.241872,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.64258064516129e-05,
+      "loss": 0.1779,
+      "step": 15117
+    },
+    {
+      "epoch": 0.241888,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.64241935483871e-05,
+      "loss": 0.1977,
+      "step": 15118
+    },
+    {
+      "epoch": 0.241904,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.642258064516129e-05,
+      "loss": 0.1564,
+      "step": 15119
+    },
+    {
+      "epoch": 0.24192,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.642096774193549e-05,
+      "loss": 0.1704,
+      "step": 15120
+    },
+    {
+      "epoch": 0.241936,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.641935483870967e-05,
+      "loss": 0.1652,
+      "step": 15121
+    },
+    {
+      "epoch": 0.241952,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.641774193548387e-05,
+      "loss": 0.1695,
+      "step": 15122
+    },
+    {
+      "epoch": 0.241968,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.641612903225807e-05,
+      "loss": 0.1714,
+      "step": 15123
+    },
+    {
+      "epoch": 0.241984,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.641451612903227e-05,
+      "loss": 0.1564,
+      "step": 15124
+    },
+    {
+      "epoch": 0.242,
+      "grad_norm": 1.2109375,
+      "learning_rate": 7.641290322580646e-05,
+      "loss": 0.1795,
+      "step": 15125
+    },
+    {
+      "epoch": 0.242016,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.641129032258066e-05,
+      "loss": 0.1619,
+      "step": 15126
+    },
+    {
+      "epoch": 0.242032,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.640967741935484e-05,
+      "loss": 0.1494,
+      "step": 15127
+    },
+    {
+      "epoch": 0.242048,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.640806451612904e-05,
+      "loss": 0.188,
+      "step": 15128
+    },
+    {
+      "epoch": 0.242064,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.640645161290323e-05,
+      "loss": 0.1692,
+      "step": 15129
+    },
+    {
+      "epoch": 0.24208,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.640483870967742e-05,
+      "loss": 0.186,
+      "step": 15130
+    },
+    {
+      "epoch": 0.242096,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.640322580645161e-05,
+      "loss": 0.2448,
+      "step": 15131
+    },
+    {
+      "epoch": 0.242112,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.64016129032258e-05,
+      "loss": 0.183,
+      "step": 15132
+    },
+    {
+      "epoch": 0.242128,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.64e-05,
+      "loss": 0.2106,
+      "step": 15133
+    },
+    {
+      "epoch": 0.242144,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.63983870967742e-05,
+      "loss": 0.1551,
+      "step": 15134
+    },
+    {
+      "epoch": 0.24216,
+      "grad_norm": 1.7421875,
+      "learning_rate": 7.639677419354839e-05,
+      "loss": 0.2284,
+      "step": 15135
+    },
+    {
+      "epoch": 0.242176,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.639516129032259e-05,
+      "loss": 0.1397,
+      "step": 15136
+    },
+    {
+      "epoch": 0.242192,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.639354838709679e-05,
+      "loss": 0.1805,
+      "step": 15137
+    },
+    {
+      "epoch": 0.242208,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.639193548387097e-05,
+      "loss": 0.1486,
+      "step": 15138
+    },
+    {
+      "epoch": 0.242224,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.639032258064517e-05,
+      "loss": 0.176,
+      "step": 15139
+    },
+    {
+      "epoch": 0.24224,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.638870967741936e-05,
+      "loss": 0.1414,
+      "step": 15140
+    },
+    {
+      "epoch": 0.242256,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.638709677419356e-05,
+      "loss": 0.1605,
+      "step": 15141
+    },
+    {
+      "epoch": 0.242272,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.638548387096774e-05,
+      "loss": 0.1373,
+      "step": 15142
+    },
+    {
+      "epoch": 0.242288,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.638387096774194e-05,
+      "loss": 0.1653,
+      "step": 15143
+    },
+    {
+      "epoch": 0.242304,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.638225806451613e-05,
+      "loss": 0.1682,
+      "step": 15144
+    },
+    {
+      "epoch": 0.24232,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.638064516129033e-05,
+      "loss": 0.1614,
+      "step": 15145
+    },
+    {
+      "epoch": 0.242336,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.637903225806451e-05,
+      "loss": 0.1769,
+      "step": 15146
+    },
+    {
+      "epoch": 0.242352,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.637741935483871e-05,
+      "loss": 0.1424,
+      "step": 15147
+    },
+    {
+      "epoch": 0.242368,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.637580645161291e-05,
+      "loss": 0.1534,
+      "step": 15148
+    },
+    {
+      "epoch": 0.242384,
+      "grad_norm": 0.625,
+      "learning_rate": 7.63741935483871e-05,
+      "loss": 0.1581,
+      "step": 15149
+    },
+    {
+      "epoch": 0.2424,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.63725806451613e-05,
+      "loss": 0.1644,
+      "step": 15150
+    },
+    {
+      "epoch": 0.242416,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.637096774193549e-05,
+      "loss": 0.1342,
+      "step": 15151
+    },
+    {
+      "epoch": 0.242432,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.636935483870968e-05,
+      "loss": 0.1714,
+      "step": 15152
+    },
+    {
+      "epoch": 0.242448,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.636774193548387e-05,
+      "loss": 0.1558,
+      "step": 15153
+    },
+    {
+      "epoch": 0.242464,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.636612903225807e-05,
+      "loss": 0.1704,
+      "step": 15154
+    },
+    {
+      "epoch": 0.24248,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.636451612903226e-05,
+      "loss": 0.2171,
+      "step": 15155
+    },
+    {
+      "epoch": 0.242496,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.636290322580646e-05,
+      "loss": 0.1781,
+      "step": 15156
+    },
+    {
+      "epoch": 0.242512,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.636129032258064e-05,
+      "loss": 0.152,
+      "step": 15157
+    },
+    {
+      "epoch": 0.242528,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.635967741935484e-05,
+      "loss": 0.1465,
+      "step": 15158
+    },
+    {
+      "epoch": 0.242544,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.635806451612904e-05,
+      "loss": 0.162,
+      "step": 15159
+    },
+    {
+      "epoch": 0.24256,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.635645161290324e-05,
+      "loss": 0.1445,
+      "step": 15160
+    },
+    {
+      "epoch": 0.242576,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.635483870967743e-05,
+      "loss": 0.1529,
+      "step": 15161
+    },
+    {
+      "epoch": 0.242592,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.635322580645161e-05,
+      "loss": 0.1556,
+      "step": 15162
+    },
+    {
+      "epoch": 0.242608,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.635161290322581e-05,
+      "loss": 0.162,
+      "step": 15163
+    },
+    {
+      "epoch": 0.242624,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.635e-05,
+      "loss": 0.1688,
+      "step": 15164
+    },
+    {
+      "epoch": 0.24264,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.63483870967742e-05,
+      "loss": 0.1783,
+      "step": 15165
+    },
+    {
+      "epoch": 0.242656,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.634677419354838e-05,
+      "loss": 0.2124,
+      "step": 15166
+    },
+    {
+      "epoch": 0.242672,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.634516129032258e-05,
+      "loss": 0.176,
+      "step": 15167
+    },
+    {
+      "epoch": 0.242688,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.634354838709677e-05,
+      "loss": 0.173,
+      "step": 15168
+    },
+    {
+      "epoch": 0.242704,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.634193548387097e-05,
+      "loss": 0.1765,
+      "step": 15169
+    },
+    {
+      "epoch": 0.24272,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.634032258064516e-05,
+      "loss": 0.1855,
+      "step": 15170
+    },
+    {
+      "epoch": 0.242736,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.633870967741936e-05,
+      "loss": 0.1789,
+      "step": 15171
+    },
+    {
+      "epoch": 0.242752,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.633709677419356e-05,
+      "loss": 0.1434,
+      "step": 15172
+    },
+    {
+      "epoch": 0.242768,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.633548387096776e-05,
+      "loss": 0.1596,
+      "step": 15173
+    },
+    {
+      "epoch": 0.242784,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.633387096774194e-05,
+      "loss": 0.1603,
+      "step": 15174
+    },
+    {
+      "epoch": 0.2428,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.633225806451614e-05,
+      "loss": 0.1262,
+      "step": 15175
+    },
+    {
+      "epoch": 0.242816,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.633064516129033e-05,
+      "loss": 0.1694,
+      "step": 15176
+    },
+    {
+      "epoch": 0.242832,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.632903225806451e-05,
+      "loss": 0.1799,
+      "step": 15177
+    },
+    {
+      "epoch": 0.242848,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.632741935483871e-05,
+      "loss": 0.1892,
+      "step": 15178
+    },
+    {
+      "epoch": 0.242864,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.63258064516129e-05,
+      "loss": 0.1586,
+      "step": 15179
+    },
+    {
+      "epoch": 0.24288,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.63241935483871e-05,
+      "loss": 0.2005,
+      "step": 15180
+    },
+    {
+      "epoch": 0.242896,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.632258064516128e-05,
+      "loss": 0.1622,
+      "step": 15181
+    },
+    {
+      "epoch": 0.242912,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.632096774193548e-05,
+      "loss": 0.1397,
+      "step": 15182
+    },
+    {
+      "epoch": 0.242928,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.631935483870968e-05,
+      "loss": 0.1896,
+      "step": 15183
+    },
+    {
+      "epoch": 0.242944,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.631774193548388e-05,
+      "loss": 0.1895,
+      "step": 15184
+    },
+    {
+      "epoch": 0.24296,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.631612903225807e-05,
+      "loss": 0.2009,
+      "step": 15185
+    },
+    {
+      "epoch": 0.242976,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.631451612903227e-05,
+      "loss": 0.1761,
+      "step": 15186
+    },
+    {
+      "epoch": 0.242992,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.631290322580646e-05,
+      "loss": 0.1882,
+      "step": 15187
+    },
+    {
+      "epoch": 0.243008,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.631129032258065e-05,
+      "loss": 0.1769,
+      "step": 15188
+    },
+    {
+      "epoch": 0.243024,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.630967741935484e-05,
+      "loss": 0.143,
+      "step": 15189
+    },
+    {
+      "epoch": 0.24304,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.630806451612904e-05,
+      "loss": 0.1712,
+      "step": 15190
+    },
+    {
+      "epoch": 0.243056,
+      "grad_norm": 0.75,
+      "learning_rate": 7.630645161290323e-05,
+      "loss": 0.1665,
+      "step": 15191
+    },
+    {
+      "epoch": 0.243072,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.630483870967741e-05,
+      "loss": 0.1644,
+      "step": 15192
+    },
+    {
+      "epoch": 0.243088,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.630322580645161e-05,
+      "loss": 0.1591,
+      "step": 15193
+    },
+    {
+      "epoch": 0.243104,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.630161290322581e-05,
+      "loss": 0.1259,
+      "step": 15194
+    },
+    {
+      "epoch": 0.24312,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.630000000000001e-05,
+      "loss": 0.1872,
+      "step": 15195
+    },
+    {
+      "epoch": 0.243136,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.62983870967742e-05,
+      "loss": 0.1304,
+      "step": 15196
+    },
+    {
+      "epoch": 0.243152,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.62967741935484e-05,
+      "loss": 0.1688,
+      "step": 15197
+    },
+    {
+      "epoch": 0.243168,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.629516129032258e-05,
+      "loss": 0.1401,
+      "step": 15198
+    },
+    {
+      "epoch": 0.243184,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.629354838709678e-05,
+      "loss": 0.1799,
+      "step": 15199
+    },
+    {
+      "epoch": 0.2432,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.629193548387097e-05,
+      "loss": 0.1895,
+      "step": 15200
+    },
+    {
+      "epoch": 0.243216,
+      "grad_norm": 1.0,
+      "learning_rate": 7.629032258064517e-05,
+      "loss": 0.19,
+      "step": 15201
+    },
+    {
+      "epoch": 0.243232,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.628870967741935e-05,
+      "loss": 0.1873,
+      "step": 15202
+    },
+    {
+      "epoch": 0.243248,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.628709677419355e-05,
+      "loss": 0.1532,
+      "step": 15203
+    },
+    {
+      "epoch": 0.243264,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.628548387096774e-05,
+      "loss": 0.1279,
+      "step": 15204
+    },
+    {
+      "epoch": 0.24328,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.628387096774194e-05,
+      "loss": 0.1572,
+      "step": 15205
+    },
+    {
+      "epoch": 0.243296,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.628225806451613e-05,
+      "loss": 0.187,
+      "step": 15206
+    },
+    {
+      "epoch": 0.243312,
+      "grad_norm": 1.3515625,
+      "learning_rate": 7.628064516129033e-05,
+      "loss": 0.1781,
+      "step": 15207
+    },
+    {
+      "epoch": 0.243328,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.627903225806453e-05,
+      "loss": 0.1806,
+      "step": 15208
+    },
+    {
+      "epoch": 0.243344,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.627741935483871e-05,
+      "loss": 0.1712,
+      "step": 15209
+    },
+    {
+      "epoch": 0.24336,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.627580645161291e-05,
+      "loss": 0.1815,
+      "step": 15210
+    },
+    {
+      "epoch": 0.243376,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.62741935483871e-05,
+      "loss": 0.1779,
+      "step": 15211
+    },
+    {
+      "epoch": 0.243392,
+      "grad_norm": 0.5078125,
+      "learning_rate": 7.62725806451613e-05,
+      "loss": 0.151,
+      "step": 15212
+    },
+    {
+      "epoch": 0.243408,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.627096774193548e-05,
+      "loss": 0.1624,
+      "step": 15213
+    },
+    {
+      "epoch": 0.243424,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.626935483870968e-05,
+      "loss": 0.1825,
+      "step": 15214
+    },
+    {
+      "epoch": 0.24344,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.626774193548387e-05,
+      "loss": 0.2002,
+      "step": 15215
+    },
+    {
+      "epoch": 0.243456,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.626612903225807e-05,
+      "loss": 0.1764,
+      "step": 15216
+    },
+    {
+      "epoch": 0.243472,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.626451612903225e-05,
+      "loss": 0.1616,
+      "step": 15217
+    },
+    {
+      "epoch": 0.243488,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.626290322580645e-05,
+      "loss": 0.1611,
+      "step": 15218
+    },
+    {
+      "epoch": 0.243504,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.626129032258065e-05,
+      "loss": 0.1299,
+      "step": 15219
+    },
+    {
+      "epoch": 0.24352,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.625967741935485e-05,
+      "loss": 0.148,
+      "step": 15220
+    },
+    {
+      "epoch": 0.243536,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.625806451612904e-05,
+      "loss": 0.1733,
+      "step": 15221
+    },
+    {
+      "epoch": 0.243552,
+      "grad_norm": 0.625,
+      "learning_rate": 7.625645161290324e-05,
+      "loss": 0.1633,
+      "step": 15222
+    },
+    {
+      "epoch": 0.243568,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.625483870967742e-05,
+      "loss": 0.1588,
+      "step": 15223
+    },
+    {
+      "epoch": 0.243584,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.625322580645161e-05,
+      "loss": 0.164,
+      "step": 15224
+    },
+    {
+      "epoch": 0.2436,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.625161290322581e-05,
+      "loss": 0.1708,
+      "step": 15225
+    },
+    {
+      "epoch": 0.243616,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.625e-05,
+      "loss": 0.1608,
+      "step": 15226
+    },
+    {
+      "epoch": 0.243632,
+      "grad_norm": 1.4375,
+      "learning_rate": 7.62483870967742e-05,
+      "loss": 0.2141,
+      "step": 15227
+    },
+    {
+      "epoch": 0.243648,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.624677419354838e-05,
+      "loss": 0.1713,
+      "step": 15228
+    },
+    {
+      "epoch": 0.243664,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.624516129032258e-05,
+      "loss": 0.1957,
+      "step": 15229
+    },
+    {
+      "epoch": 0.24368,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.624354838709677e-05,
+      "loss": 0.1669,
+      "step": 15230
+    },
+    {
+      "epoch": 0.243696,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.624193548387097e-05,
+      "loss": 0.1741,
+      "step": 15231
+    },
+    {
+      "epoch": 0.243712,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.624032258064517e-05,
+      "loss": 0.1361,
+      "step": 15232
+    },
+    {
+      "epoch": 0.243728,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.623870967741937e-05,
+      "loss": 0.1952,
+      "step": 15233
+    },
+    {
+      "epoch": 0.243744,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.623709677419355e-05,
+      "loss": 0.1701,
+      "step": 15234
+    },
+    {
+      "epoch": 0.24376,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.623548387096775e-05,
+      "loss": 0.154,
+      "step": 15235
+    },
+    {
+      "epoch": 0.243776,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.623387096774194e-05,
+      "loss": 0.1701,
+      "step": 15236
+    },
+    {
+      "epoch": 0.243792,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.623225806451614e-05,
+      "loss": 0.2209,
+      "step": 15237
+    },
+    {
+      "epoch": 0.243808,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.623064516129032e-05,
+      "loss": 0.165,
+      "step": 15238
+    },
+    {
+      "epoch": 0.243824,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.622903225806451e-05,
+      "loss": 0.1606,
+      "step": 15239
+    },
+    {
+      "epoch": 0.24384,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.622741935483871e-05,
+      "loss": 0.1867,
+      "step": 15240
+    },
+    {
+      "epoch": 0.243856,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.62258064516129e-05,
+      "loss": 0.2108,
+      "step": 15241
+    },
+    {
+      "epoch": 0.243872,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.62241935483871e-05,
+      "loss": 0.1539,
+      "step": 15242
+    },
+    {
+      "epoch": 0.243888,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.62225806451613e-05,
+      "loss": 0.1897,
+      "step": 15243
+    },
+    {
+      "epoch": 0.243904,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.62209677419355e-05,
+      "loss": 0.1453,
+      "step": 15244
+    },
+    {
+      "epoch": 0.24392,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.621935483870968e-05,
+      "loss": 0.1576,
+      "step": 15245
+    },
+    {
+      "epoch": 0.243936,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.621774193548388e-05,
+      "loss": 0.1661,
+      "step": 15246
+    },
+    {
+      "epoch": 0.243952,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.621612903225807e-05,
+      "loss": 0.1838,
+      "step": 15247
+    },
+    {
+      "epoch": 0.243968,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.621451612903227e-05,
+      "loss": 0.2185,
+      "step": 15248
+    },
+    {
+      "epoch": 0.243984,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.621290322580645e-05,
+      "loss": 0.1899,
+      "step": 15249
+    },
+    {
+      "epoch": 0.244,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.621129032258065e-05,
+      "loss": 0.1664,
+      "step": 15250
+    },
+    {
+      "epoch": 0.244016,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.620967741935484e-05,
+      "loss": 0.1435,
+      "step": 15251
+    },
+    {
+      "epoch": 0.244032,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.620806451612904e-05,
+      "loss": 0.2158,
+      "step": 15252
+    },
+    {
+      "epoch": 0.244048,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.620645161290322e-05,
+      "loss": 0.1743,
+      "step": 15253
+    },
+    {
+      "epoch": 0.244064,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.620483870967742e-05,
+      "loss": 0.1745,
+      "step": 15254
+    },
+    {
+      "epoch": 0.24408,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.620322580645162e-05,
+      "loss": 0.1892,
+      "step": 15255
+    },
+    {
+      "epoch": 0.244096,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.620161290322581e-05,
+      "loss": 0.1224,
+      "step": 15256
+    },
+    {
+      "epoch": 0.244112,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.620000000000001e-05,
+      "loss": 0.1848,
+      "step": 15257
+    },
+    {
+      "epoch": 0.244128,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.61983870967742e-05,
+      "loss": 0.1736,
+      "step": 15258
+    },
+    {
+      "epoch": 0.244144,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.61967741935484e-05,
+      "loss": 0.1579,
+      "step": 15259
+    },
+    {
+      "epoch": 0.24416,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.619516129032258e-05,
+      "loss": 0.1319,
+      "step": 15260
+    },
+    {
+      "epoch": 0.244176,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.619354838709678e-05,
+      "loss": 0.1857,
+      "step": 15261
+    },
+    {
+      "epoch": 0.244192,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.619193548387097e-05,
+      "loss": 0.1715,
+      "step": 15262
+    },
+    {
+      "epoch": 0.244208,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.619032258064517e-05,
+      "loss": 0.1808,
+      "step": 15263
+    },
+    {
+      "epoch": 0.244224,
+      "grad_norm": 1.3046875,
+      "learning_rate": 7.618870967741935e-05,
+      "loss": 0.1938,
+      "step": 15264
+    },
+    {
+      "epoch": 0.24424,
+      "grad_norm": 1.5078125,
+      "learning_rate": 7.618709677419355e-05,
+      "loss": 0.1698,
+      "step": 15265
+    },
+    {
+      "epoch": 0.244256,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.618548387096774e-05,
+      "loss": 0.2052,
+      "step": 15266
+    },
+    {
+      "epoch": 0.244272,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.618387096774194e-05,
+      "loss": 0.1568,
+      "step": 15267
+    },
+    {
+      "epoch": 0.244288,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.618225806451614e-05,
+      "loss": 0.1619,
+      "step": 15268
+    },
+    {
+      "epoch": 0.244304,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.618064516129034e-05,
+      "loss": 0.2027,
+      "step": 15269
+    },
+    {
+      "epoch": 0.24432,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.617903225806452e-05,
+      "loss": 0.137,
+      "step": 15270
+    },
+    {
+      "epoch": 0.244336,
+      "grad_norm": 0.75,
+      "learning_rate": 7.617741935483871e-05,
+      "loss": 0.2011,
+      "step": 15271
+    },
+    {
+      "epoch": 0.244352,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.617580645161291e-05,
+      "loss": 0.1637,
+      "step": 15272
+    },
+    {
+      "epoch": 0.244368,
+      "grad_norm": 1.4140625,
+      "learning_rate": 7.61741935483871e-05,
+      "loss": 0.1895,
+      "step": 15273
+    },
+    {
+      "epoch": 0.244384,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.61725806451613e-05,
+      "loss": 0.1743,
+      "step": 15274
+    },
+    {
+      "epoch": 0.2444,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.617096774193548e-05,
+      "loss": 0.1295,
+      "step": 15275
+    },
+    {
+      "epoch": 0.244416,
+      "grad_norm": 1.8359375,
+      "learning_rate": 7.616935483870968e-05,
+      "loss": 0.17,
+      "step": 15276
+    },
+    {
+      "epoch": 0.244432,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.616774193548387e-05,
+      "loss": 0.1978,
+      "step": 15277
+    },
+    {
+      "epoch": 0.244448,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.616612903225807e-05,
+      "loss": 0.1391,
+      "step": 15278
+    },
+    {
+      "epoch": 0.244464,
+      "grad_norm": 1.25,
+      "learning_rate": 7.616451612903227e-05,
+      "loss": 0.2166,
+      "step": 15279
+    },
+    {
+      "epoch": 0.24448,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.616290322580647e-05,
+      "loss": 0.1749,
+      "step": 15280
+    },
+    {
+      "epoch": 0.244496,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.616129032258065e-05,
+      "loss": 0.1865,
+      "step": 15281
+    },
+    {
+      "epoch": 0.244512,
+      "grad_norm": 1.0,
+      "learning_rate": 7.615967741935485e-05,
+      "loss": 0.1575,
+      "step": 15282
+    },
+    {
+      "epoch": 0.244528,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.615806451612904e-05,
+      "loss": 0.1685,
+      "step": 15283
+    },
+    {
+      "epoch": 0.244544,
+      "grad_norm": 1.40625,
+      "learning_rate": 7.615645161290324e-05,
+      "loss": 0.1733,
+      "step": 15284
+    },
+    {
+      "epoch": 0.24456,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.615483870967742e-05,
+      "loss": 0.1575,
+      "step": 15285
+    },
+    {
+      "epoch": 0.244576,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.615322580645161e-05,
+      "loss": 0.1638,
+      "step": 15286
+    },
+    {
+      "epoch": 0.244592,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.615161290322581e-05,
+      "loss": 0.1573,
+      "step": 15287
+    },
+    {
+      "epoch": 0.244608,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.615e-05,
+      "loss": 0.2123,
+      "step": 15288
+    },
+    {
+      "epoch": 0.244624,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.61483870967742e-05,
+      "loss": 0.1725,
+      "step": 15289
+    },
+    {
+      "epoch": 0.24464,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.61467741935484e-05,
+      "loss": 0.1831,
+      "step": 15290
+    },
+    {
+      "epoch": 0.244656,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.614516129032258e-05,
+      "loss": 0.1765,
+      "step": 15291
+    },
+    {
+      "epoch": 0.244672,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.614354838709678e-05,
+      "loss": 0.1365,
+      "step": 15292
+    },
+    {
+      "epoch": 0.244688,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.614193548387098e-05,
+      "loss": 0.1225,
+      "step": 15293
+    },
+    {
+      "epoch": 0.244704,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.614032258064516e-05,
+      "loss": 0.1673,
+      "step": 15294
+    },
+    {
+      "epoch": 0.24472,
+      "grad_norm": 1.0,
+      "learning_rate": 7.613870967741936e-05,
+      "loss": 0.2106,
+      "step": 15295
+    },
+    {
+      "epoch": 0.244736,
+      "grad_norm": 1.3125,
+      "learning_rate": 7.613709677419355e-05,
+      "loss": 0.2084,
+      "step": 15296
+    },
+    {
+      "epoch": 0.244752,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.613548387096775e-05,
+      "loss": 0.2426,
+      "step": 15297
+    },
+    {
+      "epoch": 0.244768,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.613387096774194e-05,
+      "loss": 0.1566,
+      "step": 15298
+    },
+    {
+      "epoch": 0.244784,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.613225806451614e-05,
+      "loss": 0.1583,
+      "step": 15299
+    },
+    {
+      "epoch": 0.2448,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.613064516129032e-05,
+      "loss": 0.1644,
+      "step": 15300
+    },
+    {
+      "epoch": 0.244816,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.612903225806451e-05,
+      "loss": 0.1364,
+      "step": 15301
+    },
+    {
+      "epoch": 0.244832,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.612741935483871e-05,
+      "loss": 0.1467,
+      "step": 15302
+    },
+    {
+      "epoch": 0.244848,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.612580645161291e-05,
+      "loss": 0.2061,
+      "step": 15303
+    },
+    {
+      "epoch": 0.244864,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.612419354838711e-05,
+      "loss": 0.1732,
+      "step": 15304
+    },
+    {
+      "epoch": 0.24488,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.612258064516129e-05,
+      "loss": 0.1855,
+      "step": 15305
+    },
+    {
+      "epoch": 0.244896,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.612096774193549e-05,
+      "loss": 0.1613,
+      "step": 15306
+    },
+    {
+      "epoch": 0.244912,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.611935483870968e-05,
+      "loss": 0.1856,
+      "step": 15307
+    },
+    {
+      "epoch": 0.244928,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.611774193548388e-05,
+      "loss": 0.1785,
+      "step": 15308
+    },
+    {
+      "epoch": 0.244944,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.611612903225806e-05,
+      "loss": 0.2328,
+      "step": 15309
+    },
+    {
+      "epoch": 0.24496,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.611451612903226e-05,
+      "loss": 0.1805,
+      "step": 15310
+    },
+    {
+      "epoch": 0.244976,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.611290322580645e-05,
+      "loss": 0.1492,
+      "step": 15311
+    },
+    {
+      "epoch": 0.244992,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.611129032258065e-05,
+      "loss": 0.1665,
+      "step": 15312
+    },
+    {
+      "epoch": 0.245008,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.610967741935484e-05,
+      "loss": 0.1888,
+      "step": 15313
+    },
+    {
+      "epoch": 0.245024,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.610806451612904e-05,
+      "loss": 0.1846,
+      "step": 15314
+    },
+    {
+      "epoch": 0.24504,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.610645161290324e-05,
+      "loss": 0.1601,
+      "step": 15315
+    },
+    {
+      "epoch": 0.245056,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.610483870967743e-05,
+      "loss": 0.1635,
+      "step": 15316
+    },
+    {
+      "epoch": 0.245072,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.610322580645162e-05,
+      "loss": 0.1887,
+      "step": 15317
+    },
+    {
+      "epoch": 0.245088,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.610161290322581e-05,
+      "loss": 0.1721,
+      "step": 15318
+    },
+    {
+      "epoch": 0.245104,
+      "grad_norm": 0.455078125,
+      "learning_rate": 7.61e-05,
+      "loss": 0.1694,
+      "step": 15319
+    },
+    {
+      "epoch": 0.24512,
+      "grad_norm": 1.0,
+      "learning_rate": 7.609838709677419e-05,
+      "loss": 0.191,
+      "step": 15320
+    },
+    {
+      "epoch": 0.245136,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.609677419354839e-05,
+      "loss": 0.16,
+      "step": 15321
+    },
+    {
+      "epoch": 0.245152,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.609516129032258e-05,
+      "loss": 0.188,
+      "step": 15322
+    },
+    {
+      "epoch": 0.245168,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.609354838709678e-05,
+      "loss": 0.1673,
+      "step": 15323
+    },
+    {
+      "epoch": 0.245184,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.609193548387096e-05,
+      "loss": 0.1634,
+      "step": 15324
+    },
+    {
+      "epoch": 0.2452,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.609032258064516e-05,
+      "loss": 0.1631,
+      "step": 15325
+    },
+    {
+      "epoch": 0.245216,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.608870967741935e-05,
+      "loss": 0.1809,
+      "step": 15326
+    },
+    {
+      "epoch": 0.245232,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.608709677419355e-05,
+      "loss": 0.1285,
+      "step": 15327
+    },
+    {
+      "epoch": 0.245248,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.608548387096775e-05,
+      "loss": 0.158,
+      "step": 15328
+    },
+    {
+      "epoch": 0.245264,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.608387096774195e-05,
+      "loss": 0.1482,
+      "step": 15329
+    },
+    {
+      "epoch": 0.24528,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.608225806451613e-05,
+      "loss": 0.1561,
+      "step": 15330
+    },
+    {
+      "epoch": 0.245296,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.608064516129033e-05,
+      "loss": 0.1692,
+      "step": 15331
+    },
+    {
+      "epoch": 0.245312,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.607903225806452e-05,
+      "loss": 0.1424,
+      "step": 15332
+    },
+    {
+      "epoch": 0.245328,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.60774193548387e-05,
+      "loss": 0.1636,
+      "step": 15333
+    },
+    {
+      "epoch": 0.245344,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.60758064516129e-05,
+      "loss": 0.1671,
+      "step": 15334
+    },
+    {
+      "epoch": 0.24536,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.607419354838709e-05,
+      "loss": 0.1572,
+      "step": 15335
+    },
+    {
+      "epoch": 0.245376,
+      "grad_norm": 0.455078125,
+      "learning_rate": 7.607258064516129e-05,
+      "loss": 0.1382,
+      "step": 15336
+    },
+    {
+      "epoch": 0.245392,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.607096774193548e-05,
+      "loss": 0.1336,
+      "step": 15337
+    },
+    {
+      "epoch": 0.245408,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.606935483870968e-05,
+      "loss": 0.1912,
+      "step": 15338
+    },
+    {
+      "epoch": 0.245424,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.606774193548388e-05,
+      "loss": 0.1745,
+      "step": 15339
+    },
+    {
+      "epoch": 0.24544,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.606612903225808e-05,
+      "loss": 0.1924,
+      "step": 15340
+    },
+    {
+      "epoch": 0.245456,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.606451612903226e-05,
+      "loss": 0.201,
+      "step": 15341
+    },
+    {
+      "epoch": 0.245472,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.606290322580646e-05,
+      "loss": 0.1552,
+      "step": 15342
+    },
+    {
+      "epoch": 0.245488,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.606129032258065e-05,
+      "loss": 0.2033,
+      "step": 15343
+    },
+    {
+      "epoch": 0.245504,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.605967741935485e-05,
+      "loss": 0.1758,
+      "step": 15344
+    },
+    {
+      "epoch": 0.24552,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.605806451612903e-05,
+      "loss": 0.154,
+      "step": 15345
+    },
+    {
+      "epoch": 0.245536,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.605645161290323e-05,
+      "loss": 0.1474,
+      "step": 15346
+    },
+    {
+      "epoch": 0.245552,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.605483870967742e-05,
+      "loss": 0.1948,
+      "step": 15347
+    },
+    {
+      "epoch": 0.245568,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.60532258064516e-05,
+      "loss": 0.1545,
+      "step": 15348
+    },
+    {
+      "epoch": 0.245584,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.60516129032258e-05,
+      "loss": 0.1423,
+      "step": 15349
+    },
+    {
+      "epoch": 0.2456,
+      "grad_norm": 0.625,
+      "learning_rate": 7.605e-05,
+      "loss": 0.1192,
+      "step": 15350
+    },
+    {
+      "epoch": 0.245616,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.60483870967742e-05,
+      "loss": 0.1834,
+      "step": 15351
+    },
+    {
+      "epoch": 0.245632,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.604677419354839e-05,
+      "loss": 0.1639,
+      "step": 15352
+    },
+    {
+      "epoch": 0.245648,
+      "grad_norm": 1.7421875,
+      "learning_rate": 7.604516129032259e-05,
+      "loss": 0.1868,
+      "step": 15353
+    },
+    {
+      "epoch": 0.245664,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.604354838709678e-05,
+      "loss": 0.1596,
+      "step": 15354
+    },
+    {
+      "epoch": 0.24568,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.604193548387098e-05,
+      "loss": 0.1634,
+      "step": 15355
+    },
+    {
+      "epoch": 0.245696,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.604032258064516e-05,
+      "loss": 0.1865,
+      "step": 15356
+    },
+    {
+      "epoch": 0.245712,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.603870967741936e-05,
+      "loss": 0.1673,
+      "step": 15357
+    },
+    {
+      "epoch": 0.245728,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.603709677419355e-05,
+      "loss": 0.1948,
+      "step": 15358
+    },
+    {
+      "epoch": 0.245744,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.603548387096775e-05,
+      "loss": 0.1891,
+      "step": 15359
+    },
+    {
+      "epoch": 0.24576,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.603387096774193e-05,
+      "loss": 0.1653,
+      "step": 15360
+    },
+    {
+      "epoch": 0.245776,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.603225806451613e-05,
+      "loss": 0.1975,
+      "step": 15361
+    },
+    {
+      "epoch": 0.245792,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.603064516129032e-05,
+      "loss": 0.1376,
+      "step": 15362
+    },
+    {
+      "epoch": 0.245808,
+      "grad_norm": 1.2109375,
+      "learning_rate": 7.602903225806452e-05,
+      "loss": 0.1973,
+      "step": 15363
+    },
+    {
+      "epoch": 0.245824,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.602741935483872e-05,
+      "loss": 0.1355,
+      "step": 15364
+    },
+    {
+      "epoch": 0.24584,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.60258064516129e-05,
+      "loss": 0.1659,
+      "step": 15365
+    },
+    {
+      "epoch": 0.245856,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.60241935483871e-05,
+      "loss": 0.1784,
+      "step": 15366
+    },
+    {
+      "epoch": 0.245872,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.602258064516129e-05,
+      "loss": 0.1284,
+      "step": 15367
+    },
+    {
+      "epoch": 0.245888,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.602096774193549e-05,
+      "loss": 0.1783,
+      "step": 15368
+    },
+    {
+      "epoch": 0.245904,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.601935483870968e-05,
+      "loss": 0.1501,
+      "step": 15369
+    },
+    {
+      "epoch": 0.24592,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.601774193548388e-05,
+      "loss": 0.1572,
+      "step": 15370
+    },
+    {
+      "epoch": 0.245936,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.601612903225806e-05,
+      "loss": 0.1768,
+      "step": 15371
+    },
+    {
+      "epoch": 0.245952,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.601451612903226e-05,
+      "loss": 0.1487,
+      "step": 15372
+    },
+    {
+      "epoch": 0.245968,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.601290322580645e-05,
+      "loss": 0.1582,
+      "step": 15373
+    },
+    {
+      "epoch": 0.245984,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.601129032258065e-05,
+      "loss": 0.1629,
+      "step": 15374
+    },
+    {
+      "epoch": 0.246,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.600967741935485e-05,
+      "loss": 0.205,
+      "step": 15375
+    },
+    {
+      "epoch": 0.246016,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.600806451612905e-05,
+      "loss": 0.1647,
+      "step": 15376
+    },
+    {
+      "epoch": 0.246032,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.600645161290323e-05,
+      "loss": 0.1456,
+      "step": 15377
+    },
+    {
+      "epoch": 0.246048,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.600483870967743e-05,
+      "loss": 0.1671,
+      "step": 15378
+    },
+    {
+      "epoch": 0.246064,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.600322580645162e-05,
+      "loss": 0.1544,
+      "step": 15379
+    },
+    {
+      "epoch": 0.24608,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.60016129032258e-05,
+      "loss": 0.1662,
+      "step": 15380
+    },
+    {
+      "epoch": 0.246096,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.6e-05,
+      "loss": 0.1855,
+      "step": 15381
+    },
+    {
+      "epoch": 0.246112,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.599838709677419e-05,
+      "loss": 0.1717,
+      "step": 15382
+    },
+    {
+      "epoch": 0.246128,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.599677419354839e-05,
+      "loss": 0.1707,
+      "step": 15383
+    },
+    {
+      "epoch": 0.246144,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.599516129032258e-05,
+      "loss": 0.1895,
+      "step": 15384
+    },
+    {
+      "epoch": 0.24616,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.599354838709678e-05,
+      "loss": 0.2248,
+      "step": 15385
+    },
+    {
+      "epoch": 0.246176,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.599193548387098e-05,
+      "loss": 0.2218,
+      "step": 15386
+    },
+    {
+      "epoch": 0.246192,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.599032258064516e-05,
+      "loss": 0.1471,
+      "step": 15387
+    },
+    {
+      "epoch": 0.246208,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.598870967741936e-05,
+      "loss": 0.133,
+      "step": 15388
+    },
+    {
+      "epoch": 0.246224,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.598709677419356e-05,
+      "loss": 0.1381,
+      "step": 15389
+    },
+    {
+      "epoch": 0.24624,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.598548387096775e-05,
+      "loss": 0.1618,
+      "step": 15390
+    },
+    {
+      "epoch": 0.246256,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.598387096774195e-05,
+      "loss": 0.1622,
+      "step": 15391
+    },
+    {
+      "epoch": 0.246272,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.598225806451613e-05,
+      "loss": 0.1751,
+      "step": 15392
+    },
+    {
+      "epoch": 0.246288,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.598064516129033e-05,
+      "loss": 0.1571,
+      "step": 15393
+    },
+    {
+      "epoch": 0.246304,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.597903225806452e-05,
+      "loss": 0.1821,
+      "step": 15394
+    },
+    {
+      "epoch": 0.24632,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.59774193548387e-05,
+      "loss": 0.2013,
+      "step": 15395
+    },
+    {
+      "epoch": 0.246336,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.59758064516129e-05,
+      "loss": 0.1484,
+      "step": 15396
+    },
+    {
+      "epoch": 0.246352,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.597419354838709e-05,
+      "loss": 0.2059,
+      "step": 15397
+    },
+    {
+      "epoch": 0.246368,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.597258064516129e-05,
+      "loss": 0.165,
+      "step": 15398
+    },
+    {
+      "epoch": 0.246384,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.597096774193549e-05,
+      "loss": 0.1951,
+      "step": 15399
+    },
+    {
+      "epoch": 0.2464,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.596935483870969e-05,
+      "loss": 0.1651,
+      "step": 15400
+    },
+    {
+      "epoch": 0.246416,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.596774193548387e-05,
+      "loss": 0.1955,
+      "step": 15401
+    },
+    {
+      "epoch": 0.246432,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.596612903225807e-05,
+      "loss": 0.1725,
+      "step": 15402
+    },
+    {
+      "epoch": 0.246448,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.596451612903226e-05,
+      "loss": 0.1962,
+      "step": 15403
+    },
+    {
+      "epoch": 0.246464,
+      "grad_norm": 0.625,
+      "learning_rate": 7.596290322580646e-05,
+      "loss": 0.1739,
+      "step": 15404
+    },
+    {
+      "epoch": 0.24648,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.596129032258065e-05,
+      "loss": 0.2024,
+      "step": 15405
+    },
+    {
+      "epoch": 0.246496,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.595967741935485e-05,
+      "loss": 0.1967,
+      "step": 15406
+    },
+    {
+      "epoch": 0.246512,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.595806451612903e-05,
+      "loss": 0.1294,
+      "step": 15407
+    },
+    {
+      "epoch": 0.246528,
+      "grad_norm": 1.3203125,
+      "learning_rate": 7.595645161290323e-05,
+      "loss": 0.1555,
+      "step": 15408
+    },
+    {
+      "epoch": 0.246544,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.595483870967742e-05,
+      "loss": 0.1373,
+      "step": 15409
+    },
+    {
+      "epoch": 0.24656,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.595322580645162e-05,
+      "loss": 0.1843,
+      "step": 15410
+    },
+    {
+      "epoch": 0.246576,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.595161290322582e-05,
+      "loss": 0.2012,
+      "step": 15411
+    },
+    {
+      "epoch": 0.246592,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.595e-05,
+      "loss": 0.1639,
+      "step": 15412
+    },
+    {
+      "epoch": 0.246608,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.59483870967742e-05,
+      "loss": 0.1533,
+      "step": 15413
+    },
+    {
+      "epoch": 0.246624,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.594677419354839e-05,
+      "loss": 0.1819,
+      "step": 15414
+    },
+    {
+      "epoch": 0.24664,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.594516129032259e-05,
+      "loss": 0.1921,
+      "step": 15415
+    },
+    {
+      "epoch": 0.246656,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.594354838709677e-05,
+      "loss": 0.1434,
+      "step": 15416
+    },
+    {
+      "epoch": 0.246672,
+      "grad_norm": 1.25,
+      "learning_rate": 7.594193548387097e-05,
+      "loss": 0.1647,
+      "step": 15417
+    },
+    {
+      "epoch": 0.246688,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.594032258064516e-05,
+      "loss": 0.1647,
+      "step": 15418
+    },
+    {
+      "epoch": 0.246704,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.593870967741936e-05,
+      "loss": 0.1694,
+      "step": 15419
+    },
+    {
+      "epoch": 0.24672,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.593709677419355e-05,
+      "loss": 0.1356,
+      "step": 15420
+    },
+    {
+      "epoch": 0.246736,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.593548387096775e-05,
+      "loss": 0.1509,
+      "step": 15421
+    },
+    {
+      "epoch": 0.246752,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.593387096774193e-05,
+      "loss": 0.1771,
+      "step": 15422
+    },
+    {
+      "epoch": 0.246768,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.593225806451613e-05,
+      "loss": 0.1893,
+      "step": 15423
+    },
+    {
+      "epoch": 0.246784,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.593064516129033e-05,
+      "loss": 0.1391,
+      "step": 15424
+    },
+    {
+      "epoch": 0.2468,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.592903225806453e-05,
+      "loss": 0.1636,
+      "step": 15425
+    },
+    {
+      "epoch": 0.246816,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.592741935483872e-05,
+      "loss": 0.2029,
+      "step": 15426
+    },
+    {
+      "epoch": 0.246832,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.59258064516129e-05,
+      "loss": 0.1697,
+      "step": 15427
+    },
+    {
+      "epoch": 0.246848,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.59241935483871e-05,
+      "loss": 0.1497,
+      "step": 15428
+    },
+    {
+      "epoch": 0.246864,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.592258064516129e-05,
+      "loss": 0.1606,
+      "step": 15429
+    },
+    {
+      "epoch": 0.24688,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.592096774193549e-05,
+      "loss": 0.1971,
+      "step": 15430
+    },
+    {
+      "epoch": 0.246896,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.591935483870967e-05,
+      "loss": 0.129,
+      "step": 15431
+    },
+    {
+      "epoch": 0.246912,
+      "grad_norm": 0.625,
+      "learning_rate": 7.591774193548387e-05,
+      "loss": 0.173,
+      "step": 15432
+    },
+    {
+      "epoch": 0.246928,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.591612903225806e-05,
+      "loss": 0.1462,
+      "step": 15433
+    },
+    {
+      "epoch": 0.246944,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.591451612903226e-05,
+      "loss": 0.1588,
+      "step": 15434
+    },
+    {
+      "epoch": 0.24696,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.591290322580646e-05,
+      "loss": 0.1648,
+      "step": 15435
+    },
+    {
+      "epoch": 0.246976,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.591129032258066e-05,
+      "loss": 0.1646,
+      "step": 15436
+    },
+    {
+      "epoch": 0.246992,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.590967741935484e-05,
+      "loss": 0.2133,
+      "step": 15437
+    },
+    {
+      "epoch": 0.247008,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.590806451612904e-05,
+      "loss": 0.1698,
+      "step": 15438
+    },
+    {
+      "epoch": 0.247024,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.590645161290323e-05,
+      "loss": 0.2047,
+      "step": 15439
+    },
+    {
+      "epoch": 0.24704,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.590483870967743e-05,
+      "loss": 0.1967,
+      "step": 15440
+    },
+    {
+      "epoch": 0.247056,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.590322580645162e-05,
+      "loss": 0.2066,
+      "step": 15441
+    },
+    {
+      "epoch": 0.247072,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.59016129032258e-05,
+      "loss": 0.2106,
+      "step": 15442
+    },
+    {
+      "epoch": 0.247088,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.59e-05,
+      "loss": 0.1532,
+      "step": 15443
+    },
+    {
+      "epoch": 0.247104,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.589838709677419e-05,
+      "loss": 0.1474,
+      "step": 15444
+    },
+    {
+      "epoch": 0.24712,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.589677419354839e-05,
+      "loss": 0.1662,
+      "step": 15445
+    },
+    {
+      "epoch": 0.247136,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.589516129032259e-05,
+      "loss": 0.1709,
+      "step": 15446
+    },
+    {
+      "epoch": 0.247152,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.589354838709679e-05,
+      "loss": 0.1873,
+      "step": 15447
+    },
+    {
+      "epoch": 0.247168,
+      "grad_norm": 0.75,
+      "learning_rate": 7.589193548387097e-05,
+      "loss": 0.1466,
+      "step": 15448
+    },
+    {
+      "epoch": 0.247184,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.589032258064517e-05,
+      "loss": 0.1193,
+      "step": 15449
+    },
+    {
+      "epoch": 0.2472,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.588870967741936e-05,
+      "loss": 0.1778,
+      "step": 15450
+    },
+    {
+      "epoch": 0.247216,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.588709677419356e-05,
+      "loss": 0.1362,
+      "step": 15451
+    },
+    {
+      "epoch": 0.247232,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.588548387096774e-05,
+      "loss": 0.1619,
+      "step": 15452
+    },
+    {
+      "epoch": 0.247248,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.588387096774194e-05,
+      "loss": 0.195,
+      "step": 15453
+    },
+    {
+      "epoch": 0.247264,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.588225806451613e-05,
+      "loss": 0.1695,
+      "step": 15454
+    },
+    {
+      "epoch": 0.24728,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.588064516129033e-05,
+      "loss": 0.1611,
+      "step": 15455
+    },
+    {
+      "epoch": 0.247296,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.587903225806452e-05,
+      "loss": 0.1709,
+      "step": 15456
+    },
+    {
+      "epoch": 0.247312,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.58774193548387e-05,
+      "loss": 0.1891,
+      "step": 15457
+    },
+    {
+      "epoch": 0.247328,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.58758064516129e-05,
+      "loss": 0.2051,
+      "step": 15458
+    },
+    {
+      "epoch": 0.247344,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.58741935483871e-05,
+      "loss": 0.1903,
+      "step": 15459
+    },
+    {
+      "epoch": 0.24736,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.58725806451613e-05,
+      "loss": 0.1573,
+      "step": 15460
+    },
+    {
+      "epoch": 0.247376,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.587096774193549e-05,
+      "loss": 0.1747,
+      "step": 15461
+    },
+    {
+      "epoch": 0.247392,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.586935483870969e-05,
+      "loss": 0.184,
+      "step": 15462
+    },
+    {
+      "epoch": 0.247408,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.586774193548387e-05,
+      "loss": 0.146,
+      "step": 15463
+    },
+    {
+      "epoch": 0.247424,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.586612903225807e-05,
+      "loss": 0.1849,
+      "step": 15464
+    },
+    {
+      "epoch": 0.24744,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.586451612903226e-05,
+      "loss": 0.133,
+      "step": 15465
+    },
+    {
+      "epoch": 0.247456,
+      "grad_norm": 0.75,
+      "learning_rate": 7.586290322580646e-05,
+      "loss": 0.1359,
+      "step": 15466
+    },
+    {
+      "epoch": 0.247472,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.586129032258064e-05,
+      "loss": 0.1769,
+      "step": 15467
+    },
+    {
+      "epoch": 0.247488,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.585967741935484e-05,
+      "loss": 0.1627,
+      "step": 15468
+    },
+    {
+      "epoch": 0.247504,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.585806451612903e-05,
+      "loss": 0.1635,
+      "step": 15469
+    },
+    {
+      "epoch": 0.24752,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.585645161290323e-05,
+      "loss": 0.1788,
+      "step": 15470
+    },
+    {
+      "epoch": 0.247536,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.585483870967743e-05,
+      "loss": 0.1828,
+      "step": 15471
+    },
+    {
+      "epoch": 0.247552,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.585322580645161e-05,
+      "loss": 0.1339,
+      "step": 15472
+    },
+    {
+      "epoch": 0.247568,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.585161290322581e-05,
+      "loss": 0.18,
+      "step": 15473
+    },
+    {
+      "epoch": 0.247584,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.585e-05,
+      "loss": 0.1708,
+      "step": 15474
+    },
+    {
+      "epoch": 0.2476,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.58483870967742e-05,
+      "loss": 0.2017,
+      "step": 15475
+    },
+    {
+      "epoch": 0.247616,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.584677419354839e-05,
+      "loss": 0.1744,
+      "step": 15476
+    },
+    {
+      "epoch": 0.247632,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.584516129032259e-05,
+      "loss": 0.1675,
+      "step": 15477
+    },
+    {
+      "epoch": 0.247648,
+      "grad_norm": 1.3125,
+      "learning_rate": 7.584354838709677e-05,
+      "loss": 0.1275,
+      "step": 15478
+    },
+    {
+      "epoch": 0.247664,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.584193548387097e-05,
+      "loss": 0.1665,
+      "step": 15479
+    },
+    {
+      "epoch": 0.24768,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.584032258064516e-05,
+      "loss": 0.1872,
+      "step": 15480
+    },
+    {
+      "epoch": 0.247696,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.583870967741936e-05,
+      "loss": 0.1734,
+      "step": 15481
+    },
+    {
+      "epoch": 0.247712,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.583709677419354e-05,
+      "loss": 0.1386,
+      "step": 15482
+    },
+    {
+      "epoch": 0.247728,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.583548387096774e-05,
+      "loss": 0.1715,
+      "step": 15483
+    },
+    {
+      "epoch": 0.247744,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.583387096774194e-05,
+      "loss": 0.1575,
+      "step": 15484
+    },
+    {
+      "epoch": 0.24776,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.583225806451614e-05,
+      "loss": 0.1695,
+      "step": 15485
+    },
+    {
+      "epoch": 0.247776,
+      "grad_norm": 1.515625,
+      "learning_rate": 7.583064516129033e-05,
+      "loss": 0.1985,
+      "step": 15486
+    },
+    {
+      "epoch": 0.247792,
+      "grad_norm": 0.875,
+      "learning_rate": 7.582903225806453e-05,
+      "loss": 0.2003,
+      "step": 15487
+    },
+    {
+      "epoch": 0.247808,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.582741935483871e-05,
+      "loss": 0.1464,
+      "step": 15488
+    },
+    {
+      "epoch": 0.247824,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.58258064516129e-05,
+      "loss": 0.1334,
+      "step": 15489
+    },
+    {
+      "epoch": 0.24784,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.58241935483871e-05,
+      "loss": 0.1565,
+      "step": 15490
+    },
+    {
+      "epoch": 0.247856,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.582258064516129e-05,
+      "loss": 0.184,
+      "step": 15491
+    },
+    {
+      "epoch": 0.247872,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.582096774193549e-05,
+      "loss": 0.139,
+      "step": 15492
+    },
+    {
+      "epoch": 0.247888,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.581935483870967e-05,
+      "loss": 0.1694,
+      "step": 15493
+    },
+    {
+      "epoch": 0.247904,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.581774193548387e-05,
+      "loss": 0.1793,
+      "step": 15494
+    },
+    {
+      "epoch": 0.24792,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.581612903225807e-05,
+      "loss": 0.1585,
+      "step": 15495
+    },
+    {
+      "epoch": 0.247936,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.581451612903227e-05,
+      "loss": 0.1715,
+      "step": 15496
+    },
+    {
+      "epoch": 0.247952,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.581290322580646e-05,
+      "loss": 0.1168,
+      "step": 15497
+    },
+    {
+      "epoch": 0.247968,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.581129032258066e-05,
+      "loss": 0.1509,
+      "step": 15498
+    },
+    {
+      "epoch": 0.247984,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.580967741935484e-05,
+      "loss": 0.1574,
+      "step": 15499
+    },
+    {
+      "epoch": 0.248,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.580806451612904e-05,
+      "loss": 0.1513,
+      "step": 15500
+    },
+    {
+      "epoch": 0.248016,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.580645161290323e-05,
+      "loss": 0.1766,
+      "step": 15501
+    },
+    {
+      "epoch": 0.248032,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.580483870967743e-05,
+      "loss": 0.1923,
+      "step": 15502
+    },
+    {
+      "epoch": 0.248048,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.580322580645161e-05,
+      "loss": 0.1471,
+      "step": 15503
+    },
+    {
+      "epoch": 0.248064,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.58016129032258e-05,
+      "loss": 0.1352,
+      "step": 15504
+    },
+    {
+      "epoch": 0.24808,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.58e-05,
+      "loss": 0.1639,
+      "step": 15505
+    },
+    {
+      "epoch": 0.248096,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.57983870967742e-05,
+      "loss": 0.178,
+      "step": 15506
+    },
+    {
+      "epoch": 0.248112,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.57967741935484e-05,
+      "loss": 0.1569,
+      "step": 15507
+    },
+    {
+      "epoch": 0.248128,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.579516129032258e-05,
+      "loss": 0.1608,
+      "step": 15508
+    },
+    {
+      "epoch": 0.248144,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.579354838709678e-05,
+      "loss": 0.1719,
+      "step": 15509
+    },
+    {
+      "epoch": 0.24816,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.579193548387097e-05,
+      "loss": 0.1764,
+      "step": 15510
+    },
+    {
+      "epoch": 0.248176,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.579032258064517e-05,
+      "loss": 0.1377,
+      "step": 15511
+    },
+    {
+      "epoch": 0.248192,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.578870967741936e-05,
+      "loss": 0.1934,
+      "step": 15512
+    },
+    {
+      "epoch": 0.248208,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.578709677419356e-05,
+      "loss": 0.1647,
+      "step": 15513
+    },
+    {
+      "epoch": 0.248224,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.578548387096774e-05,
+      "loss": 0.1418,
+      "step": 15514
+    },
+    {
+      "epoch": 0.24824,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.578387096774194e-05,
+      "loss": 0.1672,
+      "step": 15515
+    },
+    {
+      "epoch": 0.248256,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.578225806451613e-05,
+      "loss": 0.1936,
+      "step": 15516
+    },
+    {
+      "epoch": 0.248272,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.578064516129033e-05,
+      "loss": 0.1426,
+      "step": 15517
+    },
+    {
+      "epoch": 0.248288,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.577903225806451e-05,
+      "loss": 0.1363,
+      "step": 15518
+    },
+    {
+      "epoch": 0.248304,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.577741935483871e-05,
+      "loss": 0.152,
+      "step": 15519
+    },
+    {
+      "epoch": 0.24832,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.577580645161291e-05,
+      "loss": 0.1372,
+      "step": 15520
+    },
+    {
+      "epoch": 0.248336,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.57741935483871e-05,
+      "loss": 0.1869,
+      "step": 15521
+    },
+    {
+      "epoch": 0.248352,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.57725806451613e-05,
+      "loss": 0.1647,
+      "step": 15522
+    },
+    {
+      "epoch": 0.248368,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.577096774193548e-05,
+      "loss": 0.1574,
+      "step": 15523
+    },
+    {
+      "epoch": 0.248384,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.576935483870968e-05,
+      "loss": 0.222,
+      "step": 15524
+    },
+    {
+      "epoch": 0.2484,
+      "grad_norm": 1.25,
+      "learning_rate": 7.576774193548387e-05,
+      "loss": 0.1804,
+      "step": 15525
+    },
+    {
+      "epoch": 0.248416,
+      "grad_norm": 0.75,
+      "learning_rate": 7.576612903225807e-05,
+      "loss": 0.1974,
+      "step": 15526
+    },
+    {
+      "epoch": 0.248432,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.576451612903226e-05,
+      "loss": 0.1372,
+      "step": 15527
+    },
+    {
+      "epoch": 0.248448,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.576290322580646e-05,
+      "loss": 0.1431,
+      "step": 15528
+    },
+    {
+      "epoch": 0.248464,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.576129032258064e-05,
+      "loss": 0.17,
+      "step": 15529
+    },
+    {
+      "epoch": 0.24848,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.575967741935484e-05,
+      "loss": 0.1185,
+      "step": 15530
+    },
+    {
+      "epoch": 0.248496,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.575806451612904e-05,
+      "loss": 0.1464,
+      "step": 15531
+    },
+    {
+      "epoch": 0.248512,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.575645161290324e-05,
+      "loss": 0.1574,
+      "step": 15532
+    },
+    {
+      "epoch": 0.248528,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.575483870967743e-05,
+      "loss": 0.1541,
+      "step": 15533
+    },
+    {
+      "epoch": 0.248544,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.575322580645163e-05,
+      "loss": 0.1619,
+      "step": 15534
+    },
+    {
+      "epoch": 0.24856,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.575161290322581e-05,
+      "loss": 0.1584,
+      "step": 15535
+    },
+    {
+      "epoch": 0.248576,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.575e-05,
+      "loss": 0.1401,
+      "step": 15536
+    },
+    {
+      "epoch": 0.248592,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.57483870967742e-05,
+      "loss": 0.1867,
+      "step": 15537
+    },
+    {
+      "epoch": 0.248608,
+      "grad_norm": 0.625,
+      "learning_rate": 7.574677419354838e-05,
+      "loss": 0.179,
+      "step": 15538
+    },
+    {
+      "epoch": 0.248624,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.574516129032258e-05,
+      "loss": 0.1691,
+      "step": 15539
+    },
+    {
+      "epoch": 0.24864,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.574354838709677e-05,
+      "loss": 0.1802,
+      "step": 15540
+    },
+    {
+      "epoch": 0.248656,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.574193548387097e-05,
+      "loss": 0.1958,
+      "step": 15541
+    },
+    {
+      "epoch": 0.248672,
+      "grad_norm": 1.0,
+      "learning_rate": 7.574032258064517e-05,
+      "loss": 0.1907,
+      "step": 15542
+    },
+    {
+      "epoch": 0.248688,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.573870967741935e-05,
+      "loss": 0.1401,
+      "step": 15543
+    },
+    {
+      "epoch": 0.248704,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.573709677419355e-05,
+      "loss": 0.1702,
+      "step": 15544
+    },
+    {
+      "epoch": 0.24872,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.573548387096775e-05,
+      "loss": 0.1852,
+      "step": 15545
+    },
+    {
+      "epoch": 0.248736,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.573387096774194e-05,
+      "loss": 0.1748,
+      "step": 15546
+    },
+    {
+      "epoch": 0.248752,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.573225806451614e-05,
+      "loss": 0.1618,
+      "step": 15547
+    },
+    {
+      "epoch": 0.248768,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.573064516129033e-05,
+      "loss": 0.1611,
+      "step": 15548
+    },
+    {
+      "epoch": 0.248784,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.572903225806453e-05,
+      "loss": 0.1745,
+      "step": 15549
+    },
+    {
+      "epoch": 0.2488,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.572741935483871e-05,
+      "loss": 0.1285,
+      "step": 15550
+    },
+    {
+      "epoch": 0.248816,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.57258064516129e-05,
+      "loss": 0.1884,
+      "step": 15551
+    },
+    {
+      "epoch": 0.248832,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.57241935483871e-05,
+      "loss": 0.1887,
+      "step": 15552
+    },
+    {
+      "epoch": 0.248848,
+      "grad_norm": 1.125,
+      "learning_rate": 7.572258064516128e-05,
+      "loss": 0.1588,
+      "step": 15553
+    },
+    {
+      "epoch": 0.248864,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.572096774193548e-05,
+      "loss": 0.1721,
+      "step": 15554
+    },
+    {
+      "epoch": 0.24888,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.571935483870968e-05,
+      "loss": 0.1859,
+      "step": 15555
+    },
+    {
+      "epoch": 0.248896,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.571774193548388e-05,
+      "loss": 0.1585,
+      "step": 15556
+    },
+    {
+      "epoch": 0.248912,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.571612903225807e-05,
+      "loss": 0.154,
+      "step": 15557
+    },
+    {
+      "epoch": 0.248928,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.571451612903227e-05,
+      "loss": 0.2226,
+      "step": 15558
+    },
+    {
+      "epoch": 0.248944,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.571290322580645e-05,
+      "loss": 0.1852,
+      "step": 15559
+    },
+    {
+      "epoch": 0.24896,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.571129032258065e-05,
+      "loss": 0.1886,
+      "step": 15560
+    },
+    {
+      "epoch": 0.248976,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.570967741935484e-05,
+      "loss": 0.1825,
+      "step": 15561
+    },
+    {
+      "epoch": 0.248992,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.570806451612904e-05,
+      "loss": 0.1721,
+      "step": 15562
+    },
+    {
+      "epoch": 0.249008,
+      "grad_norm": 1.3359375,
+      "learning_rate": 7.570645161290323e-05,
+      "loss": 0.1929,
+      "step": 15563
+    },
+    {
+      "epoch": 0.249024,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.570483870967743e-05,
+      "loss": 0.1677,
+      "step": 15564
+    },
+    {
+      "epoch": 0.24904,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.570322580645161e-05,
+      "loss": 0.1807,
+      "step": 15565
+    },
+    {
+      "epoch": 0.249056,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.570161290322581e-05,
+      "loss": 0.1805,
+      "step": 15566
+    },
+    {
+      "epoch": 0.249072,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.570000000000001e-05,
+      "loss": 0.1651,
+      "step": 15567
+    },
+    {
+      "epoch": 0.249088,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.56983870967742e-05,
+      "loss": 0.1375,
+      "step": 15568
+    },
+    {
+      "epoch": 0.249104,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.56967741935484e-05,
+      "loss": 0.1683,
+      "step": 15569
+    },
+    {
+      "epoch": 0.24912,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.569516129032258e-05,
+      "loss": 0.1696,
+      "step": 15570
+    },
+    {
+      "epoch": 0.249136,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.569354838709678e-05,
+      "loss": 0.1514,
+      "step": 15571
+    },
+    {
+      "epoch": 0.249152,
+      "grad_norm": 1.2890625,
+      "learning_rate": 7.569193548387097e-05,
+      "loss": 0.189,
+      "step": 15572
+    },
+    {
+      "epoch": 0.249168,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.569032258064517e-05,
+      "loss": 0.197,
+      "step": 15573
+    },
+    {
+      "epoch": 0.249184,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.568870967741935e-05,
+      "loss": 0.147,
+      "step": 15574
+    },
+    {
+      "epoch": 0.2492,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.568709677419355e-05,
+      "loss": 0.1537,
+      "step": 15575
+    },
+    {
+      "epoch": 0.249216,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.568548387096774e-05,
+      "loss": 0.184,
+      "step": 15576
+    },
+    {
+      "epoch": 0.249232,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.568387096774194e-05,
+      "loss": 0.1417,
+      "step": 15577
+    },
+    {
+      "epoch": 0.249248,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.568225806451613e-05,
+      "loss": 0.1261,
+      "step": 15578
+    },
+    {
+      "epoch": 0.249264,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.568064516129032e-05,
+      "loss": 0.1758,
+      "step": 15579
+    },
+    {
+      "epoch": 0.24928,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.567903225806452e-05,
+      "loss": 0.1803,
+      "step": 15580
+    },
+    {
+      "epoch": 0.249296,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.567741935483871e-05,
+      "loss": 0.1668,
+      "step": 15581
+    },
+    {
+      "epoch": 0.249312,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.567580645161291e-05,
+      "loss": 0.116,
+      "step": 15582
+    },
+    {
+      "epoch": 0.249328,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.56741935483871e-05,
+      "loss": 0.1531,
+      "step": 15583
+    },
+    {
+      "epoch": 0.249344,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.56725806451613e-05,
+      "loss": 0.1888,
+      "step": 15584
+    },
+    {
+      "epoch": 0.24936,
+      "grad_norm": 0.875,
+      "learning_rate": 7.567096774193548e-05,
+      "loss": 0.1759,
+      "step": 15585
+    },
+    {
+      "epoch": 0.249376,
+      "grad_norm": 0.75,
+      "learning_rate": 7.566935483870968e-05,
+      "loss": 0.1679,
+      "step": 15586
+    },
+    {
+      "epoch": 0.249392,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.566774193548387e-05,
+      "loss": 0.1895,
+      "step": 15587
+    },
+    {
+      "epoch": 0.249408,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.566612903225807e-05,
+      "loss": 0.1902,
+      "step": 15588
+    },
+    {
+      "epoch": 0.249424,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.566451612903225e-05,
+      "loss": 0.1671,
+      "step": 15589
+    },
+    {
+      "epoch": 0.24944,
+      "grad_norm": 0.75,
+      "learning_rate": 7.566290322580645e-05,
+      "loss": 0.149,
+      "step": 15590
+    },
+    {
+      "epoch": 0.249456,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.566129032258065e-05,
+      "loss": 0.1674,
+      "step": 15591
+    },
+    {
+      "epoch": 0.249472,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.565967741935485e-05,
+      "loss": 0.154,
+      "step": 15592
+    },
+    {
+      "epoch": 0.249488,
+      "grad_norm": 1.125,
+      "learning_rate": 7.565806451612904e-05,
+      "loss": 0.1649,
+      "step": 15593
+    },
+    {
+      "epoch": 0.249504,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.565645161290324e-05,
+      "loss": 0.1179,
+      "step": 15594
+    },
+    {
+      "epoch": 0.24952,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.565483870967742e-05,
+      "loss": 0.1523,
+      "step": 15595
+    },
+    {
+      "epoch": 0.249536,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.565322580645162e-05,
+      "loss": 0.1673,
+      "step": 15596
+    },
+    {
+      "epoch": 0.249552,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.565161290322581e-05,
+      "loss": 0.1564,
+      "step": 15597
+    },
+    {
+      "epoch": 0.249568,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.565e-05,
+      "loss": 0.1296,
+      "step": 15598
+    },
+    {
+      "epoch": 0.249584,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.56483870967742e-05,
+      "loss": 0.1665,
+      "step": 15599
+    },
+    {
+      "epoch": 0.2496,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.564677419354838e-05,
+      "loss": 0.1268,
+      "step": 15600
+    },
+    {
+      "epoch": 0.249616,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.564516129032258e-05,
+      "loss": 0.1843,
+      "step": 15601
+    },
+    {
+      "epoch": 0.249632,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.564354838709678e-05,
+      "loss": 0.1679,
+      "step": 15602
+    },
+    {
+      "epoch": 0.249648,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.564193548387098e-05,
+      "loss": 0.1802,
+      "step": 15603
+    },
+    {
+      "epoch": 0.249664,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.564032258064517e-05,
+      "loss": 0.1574,
+      "step": 15604
+    },
+    {
+      "epoch": 0.24968,
+      "grad_norm": 1.4140625,
+      "learning_rate": 7.563870967741937e-05,
+      "loss": 0.1597,
+      "step": 15605
+    },
+    {
+      "epoch": 0.249696,
+      "grad_norm": 1.34375,
+      "learning_rate": 7.563709677419355e-05,
+      "loss": 0.1832,
+      "step": 15606
+    },
+    {
+      "epoch": 0.249712,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.563548387096775e-05,
+      "loss": 0.215,
+      "step": 15607
+    },
+    {
+      "epoch": 0.249728,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.563387096774194e-05,
+      "loss": 0.172,
+      "step": 15608
+    },
+    {
+      "epoch": 0.249744,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.563225806451614e-05,
+      "loss": 0.2041,
+      "step": 15609
+    },
+    {
+      "epoch": 0.24976,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.563064516129032e-05,
+      "loss": 0.1507,
+      "step": 15610
+    },
+    {
+      "epoch": 0.249776,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.562903225806452e-05,
+      "loss": 0.145,
+      "step": 15611
+    },
+    {
+      "epoch": 0.249792,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.562741935483871e-05,
+      "loss": 0.1365,
+      "step": 15612
+    },
+    {
+      "epoch": 0.249808,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.56258064516129e-05,
+      "loss": 0.1374,
+      "step": 15613
+    },
+    {
+      "epoch": 0.249824,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.56241935483871e-05,
+      "loss": 0.2205,
+      "step": 15614
+    },
+    {
+      "epoch": 0.24984,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.56225806451613e-05,
+      "loss": 0.2392,
+      "step": 15615
+    },
+    {
+      "epoch": 0.249856,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.56209677419355e-05,
+      "loss": 0.2159,
+      "step": 15616
+    },
+    {
+      "epoch": 0.249872,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.561935483870968e-05,
+      "loss": 0.1985,
+      "step": 15617
+    },
+    {
+      "epoch": 0.249888,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.561774193548388e-05,
+      "loss": 0.1584,
+      "step": 15618
+    },
+    {
+      "epoch": 0.249904,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.561612903225807e-05,
+      "loss": 0.171,
+      "step": 15619
+    },
+    {
+      "epoch": 0.24992,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.561451612903227e-05,
+      "loss": 0.1501,
+      "step": 15620
+    },
+    {
+      "epoch": 0.249936,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.561290322580645e-05,
+      "loss": 0.1595,
+      "step": 15621
+    },
+    {
+      "epoch": 0.249952,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.561129032258065e-05,
+      "loss": 0.1518,
+      "step": 15622
+    },
+    {
+      "epoch": 0.249968,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.560967741935484e-05,
+      "loss": 0.1802,
+      "step": 15623
+    },
+    {
+      "epoch": 0.249984,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.560806451612904e-05,
+      "loss": 0.1513,
+      "step": 15624
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.560645161290322e-05,
+      "loss": 0.1828,
+      "step": 15625
+    },
+    {
+      "epoch": 0.250016,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.560483870967742e-05,
+      "loss": 0.1675,
+      "step": 15626
+    },
+    {
+      "epoch": 0.250032,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.560322580645162e-05,
+      "loss": 0.1686,
+      "step": 15627
+    },
+    {
+      "epoch": 0.250048,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.560161290322581e-05,
+      "loss": 0.155,
+      "step": 15628
+    },
+    {
+      "epoch": 0.250064,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.560000000000001e-05,
+      "loss": 0.1438,
+      "step": 15629
+    },
+    {
+      "epoch": 0.25008,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.55983870967742e-05,
+      "loss": 0.1918,
+      "step": 15630
+    },
+    {
+      "epoch": 0.250096,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.55967741935484e-05,
+      "loss": 0.1796,
+      "step": 15631
+    },
+    {
+      "epoch": 0.250112,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.559516129032258e-05,
+      "loss": 0.1921,
+      "step": 15632
+    },
+    {
+      "epoch": 0.250128,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.559354838709678e-05,
+      "loss": 0.1744,
+      "step": 15633
+    },
+    {
+      "epoch": 0.250144,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.559193548387097e-05,
+      "loss": 0.1762,
+      "step": 15634
+    },
+    {
+      "epoch": 0.25016,
+      "grad_norm": 1.4609375,
+      "learning_rate": 7.559032258064517e-05,
+      "loss": 0.1763,
+      "step": 15635
+    },
+    {
+      "epoch": 0.250176,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.558870967741935e-05,
+      "loss": 0.1712,
+      "step": 15636
+    },
+    {
+      "epoch": 0.250192,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.558709677419355e-05,
+      "loss": 0.2091,
+      "step": 15637
+    },
+    {
+      "epoch": 0.250208,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.558548387096775e-05,
+      "loss": 0.1788,
+      "step": 15638
+    },
+    {
+      "epoch": 0.250224,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.558387096774194e-05,
+      "loss": 0.1627,
+      "step": 15639
+    },
+    {
+      "epoch": 0.25024,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.558225806451614e-05,
+      "loss": 0.1875,
+      "step": 15640
+    },
+    {
+      "epoch": 0.250256,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.558064516129034e-05,
+      "loss": 0.1577,
+      "step": 15641
+    },
+    {
+      "epoch": 0.250272,
+      "grad_norm": 1.3515625,
+      "learning_rate": 7.557903225806452e-05,
+      "loss": 0.194,
+      "step": 15642
+    },
+    {
+      "epoch": 0.250288,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.557741935483872e-05,
+      "loss": 0.1932,
+      "step": 15643
+    },
+    {
+      "epoch": 0.250304,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.557580645161291e-05,
+      "loss": 0.1163,
+      "step": 15644
+    },
+    {
+      "epoch": 0.25032,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.55741935483871e-05,
+      "loss": 0.1579,
+      "step": 15645
+    },
+    {
+      "epoch": 0.250336,
+      "grad_norm": 0.875,
+      "learning_rate": 7.55725806451613e-05,
+      "loss": 0.1827,
+      "step": 15646
+    },
+    {
+      "epoch": 0.250352,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.557096774193548e-05,
+      "loss": 0.1617,
+      "step": 15647
+    },
+    {
+      "epoch": 0.250368,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.556935483870968e-05,
+      "loss": 0.1751,
+      "step": 15648
+    },
+    {
+      "epoch": 0.250384,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.556774193548387e-05,
+      "loss": 0.1494,
+      "step": 15649
+    },
+    {
+      "epoch": 0.2504,
+      "grad_norm": 1.3671875,
+      "learning_rate": 7.556612903225806e-05,
+      "loss": 0.1958,
+      "step": 15650
+    },
+    {
+      "epoch": 0.250416,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.556451612903226e-05,
+      "loss": 0.1694,
+      "step": 15651
+    },
+    {
+      "epoch": 0.250432,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.556290322580646e-05,
+      "loss": 0.2127,
+      "step": 15652
+    },
+    {
+      "epoch": 0.250448,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.556129032258065e-05,
+      "loss": 0.1858,
+      "step": 15653
+    },
+    {
+      "epoch": 0.250464,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.555967741935485e-05,
+      "loss": 0.1313,
+      "step": 15654
+    },
+    {
+      "epoch": 0.25048,
+      "grad_norm": 1.390625,
+      "learning_rate": 7.555806451612904e-05,
+      "loss": 0.2087,
+      "step": 15655
+    },
+    {
+      "epoch": 0.250496,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.555645161290324e-05,
+      "loss": 0.1369,
+      "step": 15656
+    },
+    {
+      "epoch": 0.250512,
+      "grad_norm": 1.3828125,
+      "learning_rate": 7.555483870967742e-05,
+      "loss": 0.2005,
+      "step": 15657
+    },
+    {
+      "epoch": 0.250528,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.555322580645162e-05,
+      "loss": 0.1616,
+      "step": 15658
+    },
+    {
+      "epoch": 0.250544,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.555161290322581e-05,
+      "loss": 0.1454,
+      "step": 15659
+    },
+    {
+      "epoch": 0.25056,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.555e-05,
+      "loss": 0.2002,
+      "step": 15660
+    },
+    {
+      "epoch": 0.250576,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.554838709677419e-05,
+      "loss": 0.1614,
+      "step": 15661
+    },
+    {
+      "epoch": 0.250592,
+      "grad_norm": 1.484375,
+      "learning_rate": 7.554677419354839e-05,
+      "loss": 0.2334,
+      "step": 15662
+    },
+    {
+      "epoch": 0.250608,
+      "grad_norm": 1.296875,
+      "learning_rate": 7.554516129032259e-05,
+      "loss": 0.1956,
+      "step": 15663
+    },
+    {
+      "epoch": 0.250624,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.554354838709678e-05,
+      "loss": 0.1861,
+      "step": 15664
+    },
+    {
+      "epoch": 0.25064,
+      "grad_norm": 0.75,
+      "learning_rate": 7.554193548387098e-05,
+      "loss": 0.17,
+      "step": 15665
+    },
+    {
+      "epoch": 0.250656,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.554032258064516e-05,
+      "loss": 0.1781,
+      "step": 15666
+    },
+    {
+      "epoch": 0.250672,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.553870967741936e-05,
+      "loss": 0.1557,
+      "step": 15667
+    },
+    {
+      "epoch": 0.250688,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.553709677419355e-05,
+      "loss": 0.1601,
+      "step": 15668
+    },
+    {
+      "epoch": 0.250704,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.553548387096775e-05,
+      "loss": 0.1975,
+      "step": 15669
+    },
+    {
+      "epoch": 0.25072,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.553387096774194e-05,
+      "loss": 0.1534,
+      "step": 15670
+    },
+    {
+      "epoch": 0.250736,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.553225806451613e-05,
+      "loss": 0.178,
+      "step": 15671
+    },
+    {
+      "epoch": 0.250752,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.553064516129032e-05,
+      "loss": 0.2269,
+      "step": 15672
+    },
+    {
+      "epoch": 0.250768,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.552903225806452e-05,
+      "loss": 0.1432,
+      "step": 15673
+    },
+    {
+      "epoch": 0.250784,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.55274193548387e-05,
+      "loss": 0.1981,
+      "step": 15674
+    },
+    {
+      "epoch": 0.2508,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.55258064516129e-05,
+      "loss": 0.17,
+      "step": 15675
+    },
+    {
+      "epoch": 0.250816,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.55241935483871e-05,
+      "loss": 0.1931,
+      "step": 15676
+    },
+    {
+      "epoch": 0.250832,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.552258064516129e-05,
+      "loss": 0.1736,
+      "step": 15677
+    },
+    {
+      "epoch": 0.250848,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.552096774193549e-05,
+      "loss": 0.2116,
+      "step": 15678
+    },
+    {
+      "epoch": 0.250864,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.551935483870968e-05,
+      "loss": 0.1374,
+      "step": 15679
+    },
+    {
+      "epoch": 0.25088,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.551774193548388e-05,
+      "loss": 0.1537,
+      "step": 15680
+    },
+    {
+      "epoch": 0.250896,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.551612903225806e-05,
+      "loss": 0.1553,
+      "step": 15681
+    },
+    {
+      "epoch": 0.250912,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.551451612903226e-05,
+      "loss": 0.1579,
+      "step": 15682
+    },
+    {
+      "epoch": 0.250928,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.551290322580645e-05,
+      "loss": 0.1722,
+      "step": 15683
+    },
+    {
+      "epoch": 0.250944,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.551129032258065e-05,
+      "loss": 0.1626,
+      "step": 15684
+    },
+    {
+      "epoch": 0.25096,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.550967741935483e-05,
+      "loss": 0.2037,
+      "step": 15685
+    },
+    {
+      "epoch": 0.250976,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.550806451612903e-05,
+      "loss": 0.2111,
+      "step": 15686
+    },
+    {
+      "epoch": 0.250992,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.550645161290323e-05,
+      "loss": 0.1635,
+      "step": 15687
+    },
+    {
+      "epoch": 0.251008,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.550483870967743e-05,
+      "loss": 0.174,
+      "step": 15688
+    },
+    {
+      "epoch": 0.251024,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.550322580645162e-05,
+      "loss": 0.2155,
+      "step": 15689
+    },
+    {
+      "epoch": 0.25104,
+      "grad_norm": 0.49609375,
+      "learning_rate": 7.55016129032258e-05,
+      "loss": 0.1711,
+      "step": 15690
+    },
+    {
+      "epoch": 0.251056,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.55e-05,
+      "loss": 0.1764,
+      "step": 15691
+    },
+    {
+      "epoch": 0.251072,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.549838709677419e-05,
+      "loss": 0.1389,
+      "step": 15692
+    },
+    {
+      "epoch": 0.251088,
+      "grad_norm": 0.5,
+      "learning_rate": 7.549677419354839e-05,
+      "loss": 0.1462,
+      "step": 15693
+    },
+    {
+      "epoch": 0.251104,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.549516129032258e-05,
+      "loss": 0.1673,
+      "step": 15694
+    },
+    {
+      "epoch": 0.25112,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.549354838709678e-05,
+      "loss": 0.1744,
+      "step": 15695
+    },
+    {
+      "epoch": 0.251136,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.549193548387096e-05,
+      "loss": 0.1747,
+      "step": 15696
+    },
+    {
+      "epoch": 0.251152,
+      "grad_norm": 0.625,
+      "learning_rate": 7.549032258064516e-05,
+      "loss": 0.1197,
+      "step": 15697
+    },
+    {
+      "epoch": 0.251168,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.548870967741936e-05,
+      "loss": 0.1459,
+      "step": 15698
+    },
+    {
+      "epoch": 0.251184,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.548709677419356e-05,
+      "loss": 0.2256,
+      "step": 15699
+    },
+    {
+      "epoch": 0.2512,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.548548387096775e-05,
+      "loss": 0.1422,
+      "step": 15700
+    },
+    {
+      "epoch": 0.251216,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.548387096774195e-05,
+      "loss": 0.1641,
+      "step": 15701
+    },
+    {
+      "epoch": 0.251232,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.548225806451613e-05,
+      "loss": 0.1708,
+      "step": 15702
+    },
+    {
+      "epoch": 0.251248,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.548064516129033e-05,
+      "loss": 0.1465,
+      "step": 15703
+    },
+    {
+      "epoch": 0.251264,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.547903225806452e-05,
+      "loss": 0.1413,
+      "step": 15704
+    },
+    {
+      "epoch": 0.25128,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.547741935483872e-05,
+      "loss": 0.1715,
+      "step": 15705
+    },
+    {
+      "epoch": 0.251296,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.54758064516129e-05,
+      "loss": 0.1787,
+      "step": 15706
+    },
+    {
+      "epoch": 0.251312,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.547419354838709e-05,
+      "loss": 0.2126,
+      "step": 15707
+    },
+    {
+      "epoch": 0.251328,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.547258064516129e-05,
+      "loss": 0.1597,
+      "step": 15708
+    },
+    {
+      "epoch": 0.251344,
+      "grad_norm": 0.5,
+      "learning_rate": 7.547096774193548e-05,
+      "loss": 0.1826,
+      "step": 15709
+    },
+    {
+      "epoch": 0.25136,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.546935483870968e-05,
+      "loss": 0.2261,
+      "step": 15710
+    },
+    {
+      "epoch": 0.251376,
+      "grad_norm": 1.2578125,
+      "learning_rate": 7.546774193548388e-05,
+      "loss": 0.2152,
+      "step": 15711
+    },
+    {
+      "epoch": 0.251392,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.546612903225808e-05,
+      "loss": 0.1756,
+      "step": 15712
+    },
+    {
+      "epoch": 0.251408,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.546451612903226e-05,
+      "loss": 0.1751,
+      "step": 15713
+    },
+    {
+      "epoch": 0.251424,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.546290322580646e-05,
+      "loss": 0.2301,
+      "step": 15714
+    },
+    {
+      "epoch": 0.25144,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.546129032258065e-05,
+      "loss": 0.154,
+      "step": 15715
+    },
+    {
+      "epoch": 0.251456,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.545967741935485e-05,
+      "loss": 0.1355,
+      "step": 15716
+    },
+    {
+      "epoch": 0.251472,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.545806451612903e-05,
+      "loss": 0.1659,
+      "step": 15717
+    },
+    {
+      "epoch": 0.251488,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.545645161290323e-05,
+      "loss": 0.1419,
+      "step": 15718
+    },
+    {
+      "epoch": 0.251504,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.545483870967742e-05,
+      "loss": 0.1796,
+      "step": 15719
+    },
+    {
+      "epoch": 0.25152,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.545322580645162e-05,
+      "loss": 0.1382,
+      "step": 15720
+    },
+    {
+      "epoch": 0.251536,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.54516129032258e-05,
+      "loss": 0.1898,
+      "step": 15721
+    },
+    {
+      "epoch": 0.251552,
+      "grad_norm": 0.490234375,
+      "learning_rate": 7.545e-05,
+      "loss": 0.1384,
+      "step": 15722
+    },
+    {
+      "epoch": 0.251568,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.54483870967742e-05,
+      "loss": 0.2073,
+      "step": 15723
+    },
+    {
+      "epoch": 0.251584,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.544677419354839e-05,
+      "loss": 0.1973,
+      "step": 15724
+    },
+    {
+      "epoch": 0.2516,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.544516129032259e-05,
+      "loss": 0.1832,
+      "step": 15725
+    },
+    {
+      "epoch": 0.251616,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.544354838709678e-05,
+      "loss": 0.1652,
+      "step": 15726
+    },
+    {
+      "epoch": 0.251632,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.544193548387098e-05,
+      "loss": 0.1929,
+      "step": 15727
+    },
+    {
+      "epoch": 0.251648,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.544032258064516e-05,
+      "loss": 0.1723,
+      "step": 15728
+    },
+    {
+      "epoch": 0.251664,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.543870967741936e-05,
+      "loss": 0.1418,
+      "step": 15729
+    },
+    {
+      "epoch": 0.25168,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.543709677419355e-05,
+      "loss": 0.1463,
+      "step": 15730
+    },
+    {
+      "epoch": 0.251696,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.543548387096775e-05,
+      "loss": 0.162,
+      "step": 15731
+    },
+    {
+      "epoch": 0.251712,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.543387096774193e-05,
+      "loss": 0.1524,
+      "step": 15732
+    },
+    {
+      "epoch": 0.251728,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.543225806451613e-05,
+      "loss": 0.1961,
+      "step": 15733
+    },
+    {
+      "epoch": 0.251744,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.543064516129032e-05,
+      "loss": 0.1562,
+      "step": 15734
+    },
+    {
+      "epoch": 0.25176,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.542903225806452e-05,
+      "loss": 0.182,
+      "step": 15735
+    },
+    {
+      "epoch": 0.251776,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.542741935483872e-05,
+      "loss": 0.1762,
+      "step": 15736
+    },
+    {
+      "epoch": 0.251792,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.54258064516129e-05,
+      "loss": 0.1986,
+      "step": 15737
+    },
+    {
+      "epoch": 0.251808,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.54241935483871e-05,
+      "loss": 0.1584,
+      "step": 15738
+    },
+    {
+      "epoch": 0.251824,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.542258064516129e-05,
+      "loss": 0.1658,
+      "step": 15739
+    },
+    {
+      "epoch": 0.25184,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.542096774193549e-05,
+      "loss": 0.1773,
+      "step": 15740
+    },
+    {
+      "epoch": 0.251856,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.541935483870968e-05,
+      "loss": 0.1331,
+      "step": 15741
+    },
+    {
+      "epoch": 0.251872,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.541774193548387e-05,
+      "loss": 0.145,
+      "step": 15742
+    },
+    {
+      "epoch": 0.251888,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.541612903225806e-05,
+      "loss": 0.1667,
+      "step": 15743
+    },
+    {
+      "epoch": 0.251904,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.541451612903226e-05,
+      "loss": 0.1771,
+      "step": 15744
+    },
+    {
+      "epoch": 0.25192,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.541290322580645e-05,
+      "loss": 0.1889,
+      "step": 15745
+    },
+    {
+      "epoch": 0.251936,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.541129032258065e-05,
+      "loss": 0.1428,
+      "step": 15746
+    },
+    {
+      "epoch": 0.251952,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.540967741935485e-05,
+      "loss": 0.1382,
+      "step": 15747
+    },
+    {
+      "epoch": 0.251968,
+      "grad_norm": 0.5,
+      "learning_rate": 7.540806451612905e-05,
+      "loss": 0.1146,
+      "step": 15748
+    },
+    {
+      "epoch": 0.251984,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.540645161290323e-05,
+      "loss": 0.1267,
+      "step": 15749
+    },
+    {
+      "epoch": 0.252,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.540483870967743e-05,
+      "loss": 0.1545,
+      "step": 15750
+    },
+    {
+      "epoch": 0.252016,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.540322580645162e-05,
+      "loss": 0.1653,
+      "step": 15751
+    },
+    {
+      "epoch": 0.252032,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.540161290322582e-05,
+      "loss": 0.1449,
+      "step": 15752
+    },
+    {
+      "epoch": 0.252048,
+      "grad_norm": 1.2109375,
+      "learning_rate": 7.54e-05,
+      "loss": 0.2202,
+      "step": 15753
+    },
+    {
+      "epoch": 0.252064,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.539838709677419e-05,
+      "loss": 0.1452,
+      "step": 15754
+    },
+    {
+      "epoch": 0.25208,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.539677419354839e-05,
+      "loss": 0.1922,
+      "step": 15755
+    },
+    {
+      "epoch": 0.252096,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.539516129032257e-05,
+      "loss": 0.2128,
+      "step": 15756
+    },
+    {
+      "epoch": 0.252112,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.539354838709677e-05,
+      "loss": 0.1419,
+      "step": 15757
+    },
+    {
+      "epoch": 0.252128,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.539193548387097e-05,
+      "loss": 0.1446,
+      "step": 15758
+    },
+    {
+      "epoch": 0.252144,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.539032258064517e-05,
+      "loss": 0.1732,
+      "step": 15759
+    },
+    {
+      "epoch": 0.25216,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.538870967741936e-05,
+      "loss": 0.192,
+      "step": 15760
+    },
+    {
+      "epoch": 0.252176,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.538709677419356e-05,
+      "loss": 0.1529,
+      "step": 15761
+    },
+    {
+      "epoch": 0.252192,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.538548387096775e-05,
+      "loss": 0.1804,
+      "step": 15762
+    },
+    {
+      "epoch": 0.252208,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.538387096774195e-05,
+      "loss": 0.1677,
+      "step": 15763
+    },
+    {
+      "epoch": 0.252224,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.538225806451613e-05,
+      "loss": 0.1357,
+      "step": 15764
+    },
+    {
+      "epoch": 0.25224,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.538064516129033e-05,
+      "loss": 0.1753,
+      "step": 15765
+    },
+    {
+      "epoch": 0.252256,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.537903225806452e-05,
+      "loss": 0.163,
+      "step": 15766
+    },
+    {
+      "epoch": 0.252272,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.537741935483872e-05,
+      "loss": 0.1999,
+      "step": 15767
+    },
+    {
+      "epoch": 0.252288,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.53758064516129e-05,
+      "loss": 0.1704,
+      "step": 15768
+    },
+    {
+      "epoch": 0.252304,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.537419354838709e-05,
+      "loss": 0.1916,
+      "step": 15769
+    },
+    {
+      "epoch": 0.25232,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.537258064516129e-05,
+      "loss": 0.1401,
+      "step": 15770
+    },
+    {
+      "epoch": 0.252336,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.537096774193549e-05,
+      "loss": 0.1441,
+      "step": 15771
+    },
+    {
+      "epoch": 0.252352,
+      "grad_norm": 1.328125,
+      "learning_rate": 7.536935483870969e-05,
+      "loss": 0.2108,
+      "step": 15772
+    },
+    {
+      "epoch": 0.252368,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.536774193548387e-05,
+      "loss": 0.1862,
+      "step": 15773
+    },
+    {
+      "epoch": 0.252384,
+      "grad_norm": 1.25,
+      "learning_rate": 7.536612903225807e-05,
+      "loss": 0.1946,
+      "step": 15774
+    },
+    {
+      "epoch": 0.2524,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.536451612903226e-05,
+      "loss": 0.181,
+      "step": 15775
+    },
+    {
+      "epoch": 0.252416,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.536290322580646e-05,
+      "loss": 0.1253,
+      "step": 15776
+    },
+    {
+      "epoch": 0.252432,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.536129032258065e-05,
+      "loss": 0.1799,
+      "step": 15777
+    },
+    {
+      "epoch": 0.252448,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.535967741935484e-05,
+      "loss": 0.1643,
+      "step": 15778
+    },
+    {
+      "epoch": 0.252464,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.535806451612903e-05,
+      "loss": 0.203,
+      "step": 15779
+    },
+    {
+      "epoch": 0.25248,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.535645161290323e-05,
+      "loss": 0.1669,
+      "step": 15780
+    },
+    {
+      "epoch": 0.252496,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.535483870967742e-05,
+      "loss": 0.1851,
+      "step": 15781
+    },
+    {
+      "epoch": 0.252512,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.535322580645162e-05,
+      "loss": 0.1738,
+      "step": 15782
+    },
+    {
+      "epoch": 0.252528,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.535161290322582e-05,
+      "loss": 0.1775,
+      "step": 15783
+    },
+    {
+      "epoch": 0.252544,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.535e-05,
+      "loss": 0.1921,
+      "step": 15784
+    },
+    {
+      "epoch": 0.25256,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.53483870967742e-05,
+      "loss": 0.1655,
+      "step": 15785
+    },
+    {
+      "epoch": 0.252576,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.534677419354839e-05,
+      "loss": 0.1371,
+      "step": 15786
+    },
+    {
+      "epoch": 0.252592,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.534516129032259e-05,
+      "loss": 0.1977,
+      "step": 15787
+    },
+    {
+      "epoch": 0.252608,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.534354838709677e-05,
+      "loss": 0.1709,
+      "step": 15788
+    },
+    {
+      "epoch": 0.252624,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.534193548387097e-05,
+      "loss": 0.1679,
+      "step": 15789
+    },
+    {
+      "epoch": 0.25264,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.534032258064516e-05,
+      "loss": 0.1927,
+      "step": 15790
+    },
+    {
+      "epoch": 0.252656,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.533870967741936e-05,
+      "loss": 0.1892,
+      "step": 15791
+    },
+    {
+      "epoch": 0.252672,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.533709677419354e-05,
+      "loss": 0.1731,
+      "step": 15792
+    },
+    {
+      "epoch": 0.252688,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.533548387096774e-05,
+      "loss": 0.1585,
+      "step": 15793
+    },
+    {
+      "epoch": 0.252704,
+      "grad_norm": 1.3671875,
+      "learning_rate": 7.533387096774194e-05,
+      "loss": 0.1641,
+      "step": 15794
+    },
+    {
+      "epoch": 0.25272,
+      "grad_norm": 1.3515625,
+      "learning_rate": 7.533225806451613e-05,
+      "loss": 0.1753,
+      "step": 15795
+    },
+    {
+      "epoch": 0.252736,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.533064516129033e-05,
+      "loss": 0.171,
+      "step": 15796
+    },
+    {
+      "epoch": 0.252752,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.532903225806453e-05,
+      "loss": 0.1883,
+      "step": 15797
+    },
+    {
+      "epoch": 0.252768,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.532741935483872e-05,
+      "loss": 0.1782,
+      "step": 15798
+    },
+    {
+      "epoch": 0.252784,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.53258064516129e-05,
+      "loss": 0.1827,
+      "step": 15799
+    },
+    {
+      "epoch": 0.2528,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.53241935483871e-05,
+      "loss": 0.1951,
+      "step": 15800
+    },
+    {
+      "epoch": 0.252816,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.532258064516129e-05,
+      "loss": 0.1706,
+      "step": 15801
+    },
+    {
+      "epoch": 0.252832,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.532096774193549e-05,
+      "loss": 0.175,
+      "step": 15802
+    },
+    {
+      "epoch": 0.252848,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.531935483870967e-05,
+      "loss": 0.203,
+      "step": 15803
+    },
+    {
+      "epoch": 0.252864,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.531774193548387e-05,
+      "loss": 0.1818,
+      "step": 15804
+    },
+    {
+      "epoch": 0.25288,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.531612903225806e-05,
+      "loss": 0.1363,
+      "step": 15805
+    },
+    {
+      "epoch": 0.252896,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.531451612903226e-05,
+      "loss": 0.1655,
+      "step": 15806
+    },
+    {
+      "epoch": 0.252912,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.531290322580646e-05,
+      "loss": 0.2114,
+      "step": 15807
+    },
+    {
+      "epoch": 0.252928,
+      "grad_norm": 0.625,
+      "learning_rate": 7.531129032258066e-05,
+      "loss": 0.1459,
+      "step": 15808
+    },
+    {
+      "epoch": 0.252944,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.530967741935484e-05,
+      "loss": 0.1725,
+      "step": 15809
+    },
+    {
+      "epoch": 0.25296,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.530806451612904e-05,
+      "loss": 0.1681,
+      "step": 15810
+    },
+    {
+      "epoch": 0.252976,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.530645161290323e-05,
+      "loss": 0.1794,
+      "step": 15811
+    },
+    {
+      "epoch": 0.252992,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.530483870967743e-05,
+      "loss": 0.1559,
+      "step": 15812
+    },
+    {
+      "epoch": 0.253008,
+      "grad_norm": 1.390625,
+      "learning_rate": 7.530322580645162e-05,
+      "loss": 0.1666,
+      "step": 15813
+    },
+    {
+      "epoch": 0.253024,
+      "grad_norm": 0.421875,
+      "learning_rate": 7.530161290322581e-05,
+      "loss": 0.1481,
+      "step": 15814
+    },
+    {
+      "epoch": 0.25304,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.53e-05,
+      "loss": 0.1931,
+      "step": 15815
+    },
+    {
+      "epoch": 0.253056,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.529838709677419e-05,
+      "loss": 0.198,
+      "step": 15816
+    },
+    {
+      "epoch": 0.253072,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.529677419354839e-05,
+      "loss": 0.1789,
+      "step": 15817
+    },
+    {
+      "epoch": 0.253088,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.529516129032259e-05,
+      "loss": 0.2342,
+      "step": 15818
+    },
+    {
+      "epoch": 0.253104,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.529354838709679e-05,
+      "loss": 0.1922,
+      "step": 15819
+    },
+    {
+      "epoch": 0.25312,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.529193548387097e-05,
+      "loss": 0.1637,
+      "step": 15820
+    },
+    {
+      "epoch": 0.253136,
+      "grad_norm": 1.3984375,
+      "learning_rate": 7.529032258064517e-05,
+      "loss": 0.1838,
+      "step": 15821
+    },
+    {
+      "epoch": 0.253152,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.528870967741936e-05,
+      "loss": 0.1783,
+      "step": 15822
+    },
+    {
+      "epoch": 0.253168,
+      "grad_norm": 0.484375,
+      "learning_rate": 7.528709677419356e-05,
+      "loss": 0.1551,
+      "step": 15823
+    },
+    {
+      "epoch": 0.253184,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.528548387096774e-05,
+      "loss": 0.231,
+      "step": 15824
+    },
+    {
+      "epoch": 0.2532,
+      "grad_norm": 1.2890625,
+      "learning_rate": 7.528387096774194e-05,
+      "loss": 0.1915,
+      "step": 15825
+    },
+    {
+      "epoch": 0.253216,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.528225806451613e-05,
+      "loss": 0.1753,
+      "step": 15826
+    },
+    {
+      "epoch": 0.253232,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.528064516129033e-05,
+      "loss": 0.2345,
+      "step": 15827
+    },
+    {
+      "epoch": 0.253248,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.527903225806451e-05,
+      "loss": 0.2114,
+      "step": 15828
+    },
+    {
+      "epoch": 0.253264,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.527741935483871e-05,
+      "loss": 0.1637,
+      "step": 15829
+    },
+    {
+      "epoch": 0.25328,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.52758064516129e-05,
+      "loss": 0.1468,
+      "step": 15830
+    },
+    {
+      "epoch": 0.253296,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.52741935483871e-05,
+      "loss": 0.1395,
+      "step": 15831
+    },
+    {
+      "epoch": 0.253312,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.52725806451613e-05,
+      "loss": 0.1481,
+      "step": 15832
+    },
+    {
+      "epoch": 0.253328,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.527096774193549e-05,
+      "loss": 0.1561,
+      "step": 15833
+    },
+    {
+      "epoch": 0.253344,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.526935483870969e-05,
+      "loss": 0.1765,
+      "step": 15834
+    },
+    {
+      "epoch": 0.25336,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.526774193548387e-05,
+      "loss": 0.2098,
+      "step": 15835
+    },
+    {
+      "epoch": 0.253376,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.526612903225807e-05,
+      "loss": 0.1466,
+      "step": 15836
+    },
+    {
+      "epoch": 0.253392,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.526451612903226e-05,
+      "loss": 0.198,
+      "step": 15837
+    },
+    {
+      "epoch": 0.253408,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.526290322580646e-05,
+      "loss": 0.1792,
+      "step": 15838
+    },
+    {
+      "epoch": 0.253424,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.526129032258064e-05,
+      "loss": 0.1544,
+      "step": 15839
+    },
+    {
+      "epoch": 0.25344,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.525967741935484e-05,
+      "loss": 0.1393,
+      "step": 15840
+    },
+    {
+      "epoch": 0.253456,
+      "grad_norm": 1.28125,
+      "learning_rate": 7.525806451612903e-05,
+      "loss": 0.1757,
+      "step": 15841
+    },
+    {
+      "epoch": 0.253472,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.525645161290323e-05,
+      "loss": 0.2267,
+      "step": 15842
+    },
+    {
+      "epoch": 0.253488,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.525483870967743e-05,
+      "loss": 0.1561,
+      "step": 15843
+    },
+    {
+      "epoch": 0.253504,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.525322580645163e-05,
+      "loss": 0.1618,
+      "step": 15844
+    },
+    {
+      "epoch": 0.25352,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.525161290322581e-05,
+      "loss": 0.23,
+      "step": 15845
+    },
+    {
+      "epoch": 0.253536,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.525e-05,
+      "loss": 0.2046,
+      "step": 15846
+    },
+    {
+      "epoch": 0.253552,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.52483870967742e-05,
+      "loss": 0.1803,
+      "step": 15847
+    },
+    {
+      "epoch": 0.253568,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.524677419354839e-05,
+      "loss": 0.1353,
+      "step": 15848
+    },
+    {
+      "epoch": 0.253584,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.524516129032258e-05,
+      "loss": 0.1719,
+      "step": 15849
+    },
+    {
+      "epoch": 0.2536,
+      "grad_norm": 1.5703125,
+      "learning_rate": 7.524354838709677e-05,
+      "loss": 0.2374,
+      "step": 15850
+    },
+    {
+      "epoch": 0.253616,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.524193548387097e-05,
+      "loss": 0.1549,
+      "step": 15851
+    },
+    {
+      "epoch": 0.253632,
+      "grad_norm": 0.46875,
+      "learning_rate": 7.524032258064516e-05,
+      "loss": 0.1439,
+      "step": 15852
+    },
+    {
+      "epoch": 0.253648,
+      "grad_norm": 1.25,
+      "learning_rate": 7.523870967741936e-05,
+      "loss": 0.1928,
+      "step": 15853
+    },
+    {
+      "epoch": 0.253664,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.523709677419356e-05,
+      "loss": 0.177,
+      "step": 15854
+    },
+    {
+      "epoch": 0.25368,
+      "grad_norm": 0.625,
+      "learning_rate": 7.523548387096776e-05,
+      "loss": 0.1806,
+      "step": 15855
+    },
+    {
+      "epoch": 0.253696,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.523387096774194e-05,
+      "loss": 0.1754,
+      "step": 15856
+    },
+    {
+      "epoch": 0.253712,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.523225806451614e-05,
+      "loss": 0.1602,
+      "step": 15857
+    },
+    {
+      "epoch": 0.253728,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.523064516129033e-05,
+      "loss": 0.1537,
+      "step": 15858
+    },
+    {
+      "epoch": 0.253744,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.522903225806453e-05,
+      "loss": 0.1724,
+      "step": 15859
+    },
+    {
+      "epoch": 0.25376,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.522741935483871e-05,
+      "loss": 0.1583,
+      "step": 15860
+    },
+    {
+      "epoch": 0.253776,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.52258064516129e-05,
+      "loss": 0.1391,
+      "step": 15861
+    },
+    {
+      "epoch": 0.253792,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.52241935483871e-05,
+      "loss": 0.164,
+      "step": 15862
+    },
+    {
+      "epoch": 0.253808,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.522258064516128e-05,
+      "loss": 0.1912,
+      "step": 15863
+    },
+    {
+      "epoch": 0.253824,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.522096774193548e-05,
+      "loss": 0.1701,
+      "step": 15864
+    },
+    {
+      "epoch": 0.25384,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.521935483870967e-05,
+      "loss": 0.1468,
+      "step": 15865
+    },
+    {
+      "epoch": 0.253856,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.521774193548387e-05,
+      "loss": 0.1422,
+      "step": 15866
+    },
+    {
+      "epoch": 0.253872,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.521612903225807e-05,
+      "loss": 0.1911,
+      "step": 15867
+    },
+    {
+      "epoch": 0.253888,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.521451612903227e-05,
+      "loss": 0.2003,
+      "step": 15868
+    },
+    {
+      "epoch": 0.253904,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.521290322580646e-05,
+      "loss": 0.1382,
+      "step": 15869
+    },
+    {
+      "epoch": 0.25392,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.521129032258066e-05,
+      "loss": 0.1871,
+      "step": 15870
+    },
+    {
+      "epoch": 0.253936,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.520967741935484e-05,
+      "loss": 0.1497,
+      "step": 15871
+    },
+    {
+      "epoch": 0.253952,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.520806451612904e-05,
+      "loss": 0.1664,
+      "step": 15872
+    },
+    {
+      "epoch": 0.253968,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.520645161290323e-05,
+      "loss": 0.1703,
+      "step": 15873
+    },
+    {
+      "epoch": 0.253984,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.520483870967743e-05,
+      "loss": 0.1976,
+      "step": 15874
+    },
+    {
+      "epoch": 0.254,
+      "grad_norm": 1.4453125,
+      "learning_rate": 7.520322580645161e-05,
+      "loss": 0.1556,
+      "step": 15875
+    },
+    {
+      "epoch": 0.254016,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.520161290322581e-05,
+      "loss": 0.2276,
+      "step": 15876
+    },
+    {
+      "epoch": 0.254032,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.52e-05,
+      "loss": 0.154,
+      "step": 15877
+    },
+    {
+      "epoch": 0.254048,
+      "grad_norm": 1.328125,
+      "learning_rate": 7.51983870967742e-05,
+      "loss": 0.1714,
+      "step": 15878
+    },
+    {
+      "epoch": 0.254064,
+      "grad_norm": 0.625,
+      "learning_rate": 7.51967741935484e-05,
+      "loss": 0.1394,
+      "step": 15879
+    },
+    {
+      "epoch": 0.25408,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.519516129032258e-05,
+      "loss": 0.1537,
+      "step": 15880
+    },
+    {
+      "epoch": 0.254096,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.519354838709678e-05,
+      "loss": 0.1743,
+      "step": 15881
+    },
+    {
+      "epoch": 0.254112,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.519193548387097e-05,
+      "loss": 0.2044,
+      "step": 15882
+    },
+    {
+      "epoch": 0.254128,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.519032258064517e-05,
+      "loss": 0.1628,
+      "step": 15883
+    },
+    {
+      "epoch": 0.254144,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.518870967741936e-05,
+      "loss": 0.191,
+      "step": 15884
+    },
+    {
+      "epoch": 0.25416,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.518709677419355e-05,
+      "loss": 0.1658,
+      "step": 15885
+    },
+    {
+      "epoch": 0.254176,
+      "grad_norm": 0.625,
+      "learning_rate": 7.518548387096774e-05,
+      "loss": 0.1652,
+      "step": 15886
+    },
+    {
+      "epoch": 0.254192,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.518387096774194e-05,
+      "loss": 0.1271,
+      "step": 15887
+    },
+    {
+      "epoch": 0.254208,
+      "grad_norm": 1.265625,
+      "learning_rate": 7.518225806451613e-05,
+      "loss": 0.1967,
+      "step": 15888
+    },
+    {
+      "epoch": 0.254224,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.518064516129033e-05,
+      "loss": 0.1475,
+      "step": 15889
+    },
+    {
+      "epoch": 0.25424,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.517903225806453e-05,
+      "loss": 0.1674,
+      "step": 15890
+    },
+    {
+      "epoch": 0.254256,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.517741935483871e-05,
+      "loss": 0.1787,
+      "step": 15891
+    },
+    {
+      "epoch": 0.254272,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.517580645161291e-05,
+      "loss": 0.118,
+      "step": 15892
+    },
+    {
+      "epoch": 0.254288,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.51741935483871e-05,
+      "loss": 0.1707,
+      "step": 15893
+    },
+    {
+      "epoch": 0.254304,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.51725806451613e-05,
+      "loss": 0.1942,
+      "step": 15894
+    },
+    {
+      "epoch": 0.25432,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.517096774193548e-05,
+      "loss": 0.2095,
+      "step": 15895
+    },
+    {
+      "epoch": 0.254336,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.516935483870968e-05,
+      "loss": 0.1947,
+      "step": 15896
+    },
+    {
+      "epoch": 0.254352,
+      "grad_norm": 1.2578125,
+      "learning_rate": 7.516774193548387e-05,
+      "loss": 0.1914,
+      "step": 15897
+    },
+    {
+      "epoch": 0.254368,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.516612903225807e-05,
+      "loss": 0.1552,
+      "step": 15898
+    },
+    {
+      "epoch": 0.254384,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.516451612903225e-05,
+      "loss": 0.1888,
+      "step": 15899
+    },
+    {
+      "epoch": 0.2544,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.516290322580645e-05,
+      "loss": 0.1791,
+      "step": 15900
+    },
+    {
+      "epoch": 0.254416,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.516129032258064e-05,
+      "loss": 0.1947,
+      "step": 15901
+    },
+    {
+      "epoch": 0.254432,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.515967741935484e-05,
+      "loss": 0.2067,
+      "step": 15902
+    },
+    {
+      "epoch": 0.254448,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.515806451612904e-05,
+      "loss": 0.1737,
+      "step": 15903
+    },
+    {
+      "epoch": 0.254464,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.515645161290324e-05,
+      "loss": 0.158,
+      "step": 15904
+    },
+    {
+      "epoch": 0.25448,
+      "grad_norm": 1.2578125,
+      "learning_rate": 7.515483870967743e-05,
+      "loss": 0.1624,
+      "step": 15905
+    },
+    {
+      "epoch": 0.254496,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.515322580645162e-05,
+      "loss": 0.203,
+      "step": 15906
+    },
+    {
+      "epoch": 0.254512,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.515161290322581e-05,
+      "loss": 0.1544,
+      "step": 15907
+    },
+    {
+      "epoch": 0.254528,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.515e-05,
+      "loss": 0.1972,
+      "step": 15908
+    },
+    {
+      "epoch": 0.254544,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.51483870967742e-05,
+      "loss": 0.1585,
+      "step": 15909
+    },
+    {
+      "epoch": 0.25456,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.514677419354838e-05,
+      "loss": 0.1356,
+      "step": 15910
+    },
+    {
+      "epoch": 0.254576,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.514516129032258e-05,
+      "loss": 0.1677,
+      "step": 15911
+    },
+    {
+      "epoch": 0.254592,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.514354838709677e-05,
+      "loss": 0.1861,
+      "step": 15912
+    },
+    {
+      "epoch": 0.254608,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.514193548387097e-05,
+      "loss": 0.2374,
+      "step": 15913
+    },
+    {
+      "epoch": 0.254624,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.514032258064517e-05,
+      "loss": 0.1633,
+      "step": 15914
+    },
+    {
+      "epoch": 0.25464,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.513870967741937e-05,
+      "loss": 0.1869,
+      "step": 15915
+    },
+    {
+      "epoch": 0.254656,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.513709677419355e-05,
+      "loss": 0.1947,
+      "step": 15916
+    },
+    {
+      "epoch": 0.254672,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.513548387096775e-05,
+      "loss": 0.1937,
+      "step": 15917
+    },
+    {
+      "epoch": 0.254688,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.513387096774194e-05,
+      "loss": 0.1545,
+      "step": 15918
+    },
+    {
+      "epoch": 0.254704,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.513225806451614e-05,
+      "loss": 0.157,
+      "step": 15919
+    },
+    {
+      "epoch": 0.25472,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.513064516129032e-05,
+      "loss": 0.1743,
+      "step": 15920
+    },
+    {
+      "epoch": 0.254736,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.512903225806452e-05,
+      "loss": 0.1761,
+      "step": 15921
+    },
+    {
+      "epoch": 0.254752,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.512741935483871e-05,
+      "loss": 0.1559,
+      "step": 15922
+    },
+    {
+      "epoch": 0.254768,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.512580645161291e-05,
+      "loss": 0.176,
+      "step": 15923
+    },
+    {
+      "epoch": 0.254784,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.51241935483871e-05,
+      "loss": 0.1617,
+      "step": 15924
+    },
+    {
+      "epoch": 0.2548,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.512258064516128e-05,
+      "loss": 0.1821,
+      "step": 15925
+    },
+    {
+      "epoch": 0.254816,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.512096774193548e-05,
+      "loss": 0.2351,
+      "step": 15926
+    },
+    {
+      "epoch": 0.254832,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.511935483870968e-05,
+      "loss": 0.1643,
+      "step": 15927
+    },
+    {
+      "epoch": 0.254848,
+      "grad_norm": 1.2578125,
+      "learning_rate": 7.511774193548388e-05,
+      "loss": 0.2188,
+      "step": 15928
+    },
+    {
+      "epoch": 0.254864,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.511612903225807e-05,
+      "loss": 0.1525,
+      "step": 15929
+    },
+    {
+      "epoch": 0.25488,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.511451612903227e-05,
+      "loss": 0.1492,
+      "step": 15930
+    },
+    {
+      "epoch": 0.254896,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.511290322580645e-05,
+      "loss": 0.1543,
+      "step": 15931
+    },
+    {
+      "epoch": 0.254912,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.511129032258065e-05,
+      "loss": 0.1824,
+      "step": 15932
+    },
+    {
+      "epoch": 0.254928,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.510967741935484e-05,
+      "loss": 0.1723,
+      "step": 15933
+    },
+    {
+      "epoch": 0.254944,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.510806451612904e-05,
+      "loss": 0.1552,
+      "step": 15934
+    },
+    {
+      "epoch": 0.25496,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.510645161290322e-05,
+      "loss": 0.1744,
+      "step": 15935
+    },
+    {
+      "epoch": 0.254976,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.510483870967742e-05,
+      "loss": 0.2097,
+      "step": 15936
+    },
+    {
+      "epoch": 0.254992,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.510322580645161e-05,
+      "loss": 0.1622,
+      "step": 15937
+    },
+    {
+      "epoch": 0.255008,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.510161290322581e-05,
+      "loss": 0.1396,
+      "step": 15938
+    },
+    {
+      "epoch": 0.255024,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.510000000000001e-05,
+      "loss": 0.1573,
+      "step": 15939
+    },
+    {
+      "epoch": 0.25504,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.50983870967742e-05,
+      "loss": 0.2079,
+      "step": 15940
+    },
+    {
+      "epoch": 0.255056,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.50967741935484e-05,
+      "loss": 0.1605,
+      "step": 15941
+    },
+    {
+      "epoch": 0.255072,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.509516129032258e-05,
+      "loss": 0.1547,
+      "step": 15942
+    },
+    {
+      "epoch": 0.255088,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.509354838709678e-05,
+      "loss": 0.1759,
+      "step": 15943
+    },
+    {
+      "epoch": 0.255104,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.509193548387097e-05,
+      "loss": 0.1352,
+      "step": 15944
+    },
+    {
+      "epoch": 0.25512,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.509032258064517e-05,
+      "loss": 0.1576,
+      "step": 15945
+    },
+    {
+      "epoch": 0.255136,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.508870967741935e-05,
+      "loss": 0.1912,
+      "step": 15946
+    },
+    {
+      "epoch": 0.255152,
+      "grad_norm": 0.625,
+      "learning_rate": 7.508709677419355e-05,
+      "loss": 0.1953,
+      "step": 15947
+    },
+    {
+      "epoch": 0.255168,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.508548387096774e-05,
+      "loss": 0.2028,
+      "step": 15948
+    },
+    {
+      "epoch": 0.255184,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.508387096774194e-05,
+      "loss": 0.1753,
+      "step": 15949
+    },
+    {
+      "epoch": 0.2552,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.508225806451614e-05,
+      "loss": 0.1284,
+      "step": 15950
+    },
+    {
+      "epoch": 0.255216,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.508064516129034e-05,
+      "loss": 0.2011,
+      "step": 15951
+    },
+    {
+      "epoch": 0.255232,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.507903225806452e-05,
+      "loss": 0.1374,
+      "step": 15952
+    },
+    {
+      "epoch": 0.255248,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.507741935483872e-05,
+      "loss": 0.1547,
+      "step": 15953
+    },
+    {
+      "epoch": 0.255264,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.507580645161291e-05,
+      "loss": 0.1389,
+      "step": 15954
+    },
+    {
+      "epoch": 0.25528,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.50741935483871e-05,
+      "loss": 0.2024,
+      "step": 15955
+    },
+    {
+      "epoch": 0.255296,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.50725806451613e-05,
+      "loss": 0.1496,
+      "step": 15956
+    },
+    {
+      "epoch": 0.255312,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.507096774193548e-05,
+      "loss": 0.1766,
+      "step": 15957
+    },
+    {
+      "epoch": 0.255328,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.506935483870968e-05,
+      "loss": 0.1651,
+      "step": 15958
+    },
+    {
+      "epoch": 0.255344,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.506774193548387e-05,
+      "loss": 0.1812,
+      "step": 15959
+    },
+    {
+      "epoch": 0.25536,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.506612903225807e-05,
+      "loss": 0.1748,
+      "step": 15960
+    },
+    {
+      "epoch": 0.255376,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.506451612903225e-05,
+      "loss": 0.1546,
+      "step": 15961
+    },
+    {
+      "epoch": 0.255392,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.506290322580645e-05,
+      "loss": 0.177,
+      "step": 15962
+    },
+    {
+      "epoch": 0.255408,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.506129032258065e-05,
+      "loss": 0.1515,
+      "step": 15963
+    },
+    {
+      "epoch": 0.255424,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.505967741935485e-05,
+      "loss": 0.1875,
+      "step": 15964
+    },
+    {
+      "epoch": 0.25544,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.505806451612904e-05,
+      "loss": 0.1795,
+      "step": 15965
+    },
+    {
+      "epoch": 0.255456,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.505645161290324e-05,
+      "loss": 0.2144,
+      "step": 15966
+    },
+    {
+      "epoch": 0.255472,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.505483870967742e-05,
+      "loss": 0.1994,
+      "step": 15967
+    },
+    {
+      "epoch": 0.255488,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.505322580645162e-05,
+      "loss": 0.1675,
+      "step": 15968
+    },
+    {
+      "epoch": 0.255504,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.505161290322581e-05,
+      "loss": 0.1703,
+      "step": 15969
+    },
+    {
+      "epoch": 0.25552,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.505e-05,
+      "loss": 0.1598,
+      "step": 15970
+    },
+    {
+      "epoch": 0.255536,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.50483870967742e-05,
+      "loss": 0.1523,
+      "step": 15971
+    },
+    {
+      "epoch": 0.255552,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.504677419354838e-05,
+      "loss": 0.1732,
+      "step": 15972
+    },
+    {
+      "epoch": 0.255568,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.504516129032258e-05,
+      "loss": 0.1829,
+      "step": 15973
+    },
+    {
+      "epoch": 0.255584,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.504354838709678e-05,
+      "loss": 0.1306,
+      "step": 15974
+    },
+    {
+      "epoch": 0.2556,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.504193548387098e-05,
+      "loss": 0.1601,
+      "step": 15975
+    },
+    {
+      "epoch": 0.255616,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.504032258064517e-05,
+      "loss": 0.1514,
+      "step": 15976
+    },
+    {
+      "epoch": 0.255632,
+      "grad_norm": 1.390625,
+      "learning_rate": 7.503870967741936e-05,
+      "loss": 0.1283,
+      "step": 15977
+    },
+    {
+      "epoch": 0.255648,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.503709677419355e-05,
+      "loss": 0.1578,
+      "step": 15978
+    },
+    {
+      "epoch": 0.255664,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.503548387096775e-05,
+      "loss": 0.231,
+      "step": 15979
+    },
+    {
+      "epoch": 0.25568,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.503387096774194e-05,
+      "loss": 0.1591,
+      "step": 15980
+    },
+    {
+      "epoch": 0.255696,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.503225806451614e-05,
+      "loss": 0.1618,
+      "step": 15981
+    },
+    {
+      "epoch": 0.255712,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.503064516129032e-05,
+      "loss": 0.146,
+      "step": 15982
+    },
+    {
+      "epoch": 0.255728,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.502903225806452e-05,
+      "loss": 0.1665,
+      "step": 15983
+    },
+    {
+      "epoch": 0.255744,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.502741935483871e-05,
+      "loss": 0.165,
+      "step": 15984
+    },
+    {
+      "epoch": 0.25576,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.502580645161291e-05,
+      "loss": 0.1988,
+      "step": 15985
+    },
+    {
+      "epoch": 0.255776,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.50241935483871e-05,
+      "loss": 0.1507,
+      "step": 15986
+    },
+    {
+      "epoch": 0.255792,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.50225806451613e-05,
+      "loss": 0.191,
+      "step": 15987
+    },
+    {
+      "epoch": 0.255808,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.502096774193549e-05,
+      "loss": 0.1878,
+      "step": 15988
+    },
+    {
+      "epoch": 0.255824,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.501935483870968e-05,
+      "loss": 0.1722,
+      "step": 15989
+    },
+    {
+      "epoch": 0.25584,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.501774193548388e-05,
+      "loss": 0.1638,
+      "step": 15990
+    },
+    {
+      "epoch": 0.255856,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.501612903225806e-05,
+      "loss": 0.141,
+      "step": 15991
+    },
+    {
+      "epoch": 0.255872,
+      "grad_norm": 1.0,
+      "learning_rate": 7.501451612903226e-05,
+      "loss": 0.1424,
+      "step": 15992
+    },
+    {
+      "epoch": 0.255888,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.501290322580645e-05,
+      "loss": 0.204,
+      "step": 15993
+    },
+    {
+      "epoch": 0.255904,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.501129032258065e-05,
+      "loss": 0.161,
+      "step": 15994
+    },
+    {
+      "epoch": 0.25592,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.500967741935484e-05,
+      "loss": 0.1849,
+      "step": 15995
+    },
+    {
+      "epoch": 0.255936,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.500806451612904e-05,
+      "loss": 0.1691,
+      "step": 15996
+    },
+    {
+      "epoch": 0.255952,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.500645161290322e-05,
+      "loss": 0.1935,
+      "step": 15997
+    },
+    {
+      "epoch": 0.255968,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.500483870967742e-05,
+      "loss": 0.1533,
+      "step": 15998
+    },
+    {
+      "epoch": 0.255984,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.500322580645162e-05,
+      "loss": 0.1439,
+      "step": 15999
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.500161290322582e-05,
+      "loss": 0.1721,
+      "step": 16000
+    },
+    {
+      "epoch": 0.256016,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 0.1815,
+      "step": 16001
+    },
+    {
+      "epoch": 0.256032,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.499838709677419e-05,
+      "loss": 0.1863,
+      "step": 16002
+    },
+    {
+      "epoch": 0.256048,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.499677419354839e-05,
+      "loss": 0.1893,
+      "step": 16003
+    },
+    {
+      "epoch": 0.256064,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.499516129032258e-05,
+      "loss": 0.1902,
+      "step": 16004
+    },
+    {
+      "epoch": 0.25608,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.499354838709678e-05,
+      "loss": 0.1812,
+      "step": 16005
+    },
+    {
+      "epoch": 0.256096,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.499193548387096e-05,
+      "loss": 0.1761,
+      "step": 16006
+    },
+    {
+      "epoch": 0.256112,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.499032258064516e-05,
+      "loss": 0.1343,
+      "step": 16007
+    },
+    {
+      "epoch": 0.256128,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.498870967741935e-05,
+      "loss": 0.1426,
+      "step": 16008
+    },
+    {
+      "epoch": 0.256144,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.498709677419355e-05,
+      "loss": 0.1281,
+      "step": 16009
+    },
+    {
+      "epoch": 0.25616,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.498548387096775e-05,
+      "loss": 0.1539,
+      "step": 16010
+    },
+    {
+      "epoch": 0.256176,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.498387096774195e-05,
+      "loss": 0.1489,
+      "step": 16011
+    },
+    {
+      "epoch": 0.256192,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.498225806451614e-05,
+      "loss": 0.1684,
+      "step": 16012
+    },
+    {
+      "epoch": 0.256208,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.498064516129033e-05,
+      "loss": 0.1295,
+      "step": 16013
+    },
+    {
+      "epoch": 0.256224,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.497903225806452e-05,
+      "loss": 0.1542,
+      "step": 16014
+    },
+    {
+      "epoch": 0.25624,
+      "grad_norm": 1.3828125,
+      "learning_rate": 7.497741935483872e-05,
+      "loss": 0.1976,
+      "step": 16015
+    },
+    {
+      "epoch": 0.256256,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.49758064516129e-05,
+      "loss": 0.1409,
+      "step": 16016
+    },
+    {
+      "epoch": 0.256272,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.497419354838709e-05,
+      "loss": 0.1803,
+      "step": 16017
+    },
+    {
+      "epoch": 0.256288,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.497258064516129e-05,
+      "loss": 0.1641,
+      "step": 16018
+    },
+    {
+      "epoch": 0.256304,
+      "grad_norm": 1.8515625,
+      "learning_rate": 7.497096774193548e-05,
+      "loss": 0.1947,
+      "step": 16019
+    },
+    {
+      "epoch": 0.25632,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.496935483870968e-05,
+      "loss": 0.1697,
+      "step": 16020
+    },
+    {
+      "epoch": 0.256336,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.496774193548386e-05,
+      "loss": 0.1781,
+      "step": 16021
+    },
+    {
+      "epoch": 0.256352,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.496612903225806e-05,
+      "loss": 0.181,
+      "step": 16022
+    },
+    {
+      "epoch": 0.256368,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.496451612903226e-05,
+      "loss": 0.169,
+      "step": 16023
+    },
+    {
+      "epoch": 0.256384,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.496290322580646e-05,
+      "loss": 0.1525,
+      "step": 16024
+    },
+    {
+      "epoch": 0.2564,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.496129032258065e-05,
+      "loss": 0.1521,
+      "step": 16025
+    },
+    {
+      "epoch": 0.256416,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.495967741935485e-05,
+      "loss": 0.1802,
+      "step": 16026
+    },
+    {
+      "epoch": 0.256432,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.495806451612903e-05,
+      "loss": 0.1908,
+      "step": 16027
+    },
+    {
+      "epoch": 0.256448,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.495645161290323e-05,
+      "loss": 0.2112,
+      "step": 16028
+    },
+    {
+      "epoch": 0.256464,
+      "grad_norm": 0.75,
+      "learning_rate": 7.495483870967742e-05,
+      "loss": 0.1763,
+      "step": 16029
+    },
+    {
+      "epoch": 0.25648,
+      "grad_norm": 0.625,
+      "learning_rate": 7.495322580645162e-05,
+      "loss": 0.142,
+      "step": 16030
+    },
+    {
+      "epoch": 0.256496,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.49516129032258e-05,
+      "loss": 0.1534,
+      "step": 16031
+    },
+    {
+      "epoch": 0.256512,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.495e-05,
+      "loss": 0.1851,
+      "step": 16032
+    },
+    {
+      "epoch": 0.256528,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.494838709677419e-05,
+      "loss": 0.171,
+      "step": 16033
+    },
+    {
+      "epoch": 0.256544,
+      "grad_norm": 1.2578125,
+      "learning_rate": 7.494677419354839e-05,
+      "loss": 0.1083,
+      "step": 16034
+    },
+    {
+      "epoch": 0.25656,
+      "grad_norm": 0.625,
+      "learning_rate": 7.494516129032259e-05,
+      "loss": 0.165,
+      "step": 16035
+    },
+    {
+      "epoch": 0.256576,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.494354838709678e-05,
+      "loss": 0.1499,
+      "step": 16036
+    },
+    {
+      "epoch": 0.256592,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.494193548387098e-05,
+      "loss": 0.1594,
+      "step": 16037
+    },
+    {
+      "epoch": 0.256608,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.494032258064516e-05,
+      "loss": 0.1675,
+      "step": 16038
+    },
+    {
+      "epoch": 0.256624,
+      "grad_norm": 1.3046875,
+      "learning_rate": 7.493870967741936e-05,
+      "loss": 0.1565,
+      "step": 16039
+    },
+    {
+      "epoch": 0.25664,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.493709677419355e-05,
+      "loss": 0.1662,
+      "step": 16040
+    },
+    {
+      "epoch": 0.256656,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.493548387096775e-05,
+      "loss": 0.1674,
+      "step": 16041
+    },
+    {
+      "epoch": 0.256672,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.493387096774193e-05,
+      "loss": 0.176,
+      "step": 16042
+    },
+    {
+      "epoch": 0.256688,
+      "grad_norm": 1.296875,
+      "learning_rate": 7.493225806451613e-05,
+      "loss": 0.2028,
+      "step": 16043
+    },
+    {
+      "epoch": 0.256704,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.493064516129032e-05,
+      "loss": 0.1475,
+      "step": 16044
+    },
+    {
+      "epoch": 0.25672,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.492903225806452e-05,
+      "loss": 0.1439,
+      "step": 16045
+    },
+    {
+      "epoch": 0.256736,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.492741935483872e-05,
+      "loss": 0.1608,
+      "step": 16046
+    },
+    {
+      "epoch": 0.256752,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.492580645161292e-05,
+      "loss": 0.1846,
+      "step": 16047
+    },
+    {
+      "epoch": 0.256768,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.49241935483871e-05,
+      "loss": 0.1404,
+      "step": 16048
+    },
+    {
+      "epoch": 0.256784,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.492258064516129e-05,
+      "loss": 0.1689,
+      "step": 16049
+    },
+    {
+      "epoch": 0.2568,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.492096774193549e-05,
+      "loss": 0.1626,
+      "step": 16050
+    },
+    {
+      "epoch": 0.256816,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.491935483870968e-05,
+      "loss": 0.1464,
+      "step": 16051
+    },
+    {
+      "epoch": 0.256832,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.491774193548388e-05,
+      "loss": 0.1474,
+      "step": 16052
+    },
+    {
+      "epoch": 0.256848,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.491612903225806e-05,
+      "loss": 0.1739,
+      "step": 16053
+    },
+    {
+      "epoch": 0.256864,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.491451612903226e-05,
+      "loss": 0.1641,
+      "step": 16054
+    },
+    {
+      "epoch": 0.25688,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.491290322580645e-05,
+      "loss": 0.1496,
+      "step": 16055
+    },
+    {
+      "epoch": 0.256896,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.491129032258065e-05,
+      "loss": 0.1696,
+      "step": 16056
+    },
+    {
+      "epoch": 0.256912,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.490967741935483e-05,
+      "loss": 0.1718,
+      "step": 16057
+    },
+    {
+      "epoch": 0.256928,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.490806451612903e-05,
+      "loss": 0.1616,
+      "step": 16058
+    },
+    {
+      "epoch": 0.256944,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.490645161290323e-05,
+      "loss": 0.1536,
+      "step": 16059
+    },
+    {
+      "epoch": 0.25696,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.490483870967743e-05,
+      "loss": 0.199,
+      "step": 16060
+    },
+    {
+      "epoch": 0.256976,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.490322580645162e-05,
+      "loss": 0.1651,
+      "step": 16061
+    },
+    {
+      "epoch": 0.256992,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.490161290322582e-05,
+      "loss": 0.167,
+      "step": 16062
+    },
+    {
+      "epoch": 0.257008,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.49e-05,
+      "loss": 0.1583,
+      "step": 16063
+    },
+    {
+      "epoch": 0.257024,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.489838709677419e-05,
+      "loss": 0.1553,
+      "step": 16064
+    },
+    {
+      "epoch": 0.25704,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.489677419354839e-05,
+      "loss": 0.1649,
+      "step": 16065
+    },
+    {
+      "epoch": 0.257056,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.489516129032258e-05,
+      "loss": 0.1637,
+      "step": 16066
+    },
+    {
+      "epoch": 0.257072,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.489354838709678e-05,
+      "loss": 0.1568,
+      "step": 16067
+    },
+    {
+      "epoch": 0.257088,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.489193548387096e-05,
+      "loss": 0.1719,
+      "step": 16068
+    },
+    {
+      "epoch": 0.257104,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.489032258064516e-05,
+      "loss": 0.1589,
+      "step": 16069
+    },
+    {
+      "epoch": 0.25712,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.488870967741936e-05,
+      "loss": 0.1979,
+      "step": 16070
+    },
+    {
+      "epoch": 0.257136,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.488709677419356e-05,
+      "loss": 0.1733,
+      "step": 16071
+    },
+    {
+      "epoch": 0.257152,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.488548387096775e-05,
+      "loss": 0.2144,
+      "step": 16072
+    },
+    {
+      "epoch": 0.257168,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.488387096774195e-05,
+      "loss": 0.1569,
+      "step": 16073
+    },
+    {
+      "epoch": 0.257184,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.488225806451613e-05,
+      "loss": 0.1488,
+      "step": 16074
+    },
+    {
+      "epoch": 0.2572,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.488064516129033e-05,
+      "loss": 0.1488,
+      "step": 16075
+    },
+    {
+      "epoch": 0.257216,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.487903225806452e-05,
+      "loss": 0.1538,
+      "step": 16076
+    },
+    {
+      "epoch": 0.257232,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.487741935483872e-05,
+      "loss": 0.1744,
+      "step": 16077
+    },
+    {
+      "epoch": 0.257248,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.48758064516129e-05,
+      "loss": 0.1501,
+      "step": 16078
+    },
+    {
+      "epoch": 0.257264,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.487419354838709e-05,
+      "loss": 0.1786,
+      "step": 16079
+    },
+    {
+      "epoch": 0.25728,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.487258064516129e-05,
+      "loss": 0.1588,
+      "step": 16080
+    },
+    {
+      "epoch": 0.257296,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.487096774193548e-05,
+      "loss": 0.1576,
+      "step": 16081
+    },
+    {
+      "epoch": 0.257312,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.486935483870968e-05,
+      "loss": 0.1756,
+      "step": 16082
+    },
+    {
+      "epoch": 0.257328,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.486774193548388e-05,
+      "loss": 0.1812,
+      "step": 16083
+    },
+    {
+      "epoch": 0.257344,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.486612903225807e-05,
+      "loss": 0.2006,
+      "step": 16084
+    },
+    {
+      "epoch": 0.25736,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.486451612903226e-05,
+      "loss": 0.1512,
+      "step": 16085
+    },
+    {
+      "epoch": 0.257376,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.486290322580646e-05,
+      "loss": 0.154,
+      "step": 16086
+    },
+    {
+      "epoch": 0.257392,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.486129032258065e-05,
+      "loss": 0.1761,
+      "step": 16087
+    },
+    {
+      "epoch": 0.257408,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.485967741935485e-05,
+      "loss": 0.1827,
+      "step": 16088
+    },
+    {
+      "epoch": 0.257424,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.485806451612903e-05,
+      "loss": 0.1757,
+      "step": 16089
+    },
+    {
+      "epoch": 0.25744,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.485645161290323e-05,
+      "loss": 0.1766,
+      "step": 16090
+    },
+    {
+      "epoch": 0.257456,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.485483870967742e-05,
+      "loss": 0.1227,
+      "step": 16091
+    },
+    {
+      "epoch": 0.257472,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.485322580645162e-05,
+      "loss": 0.1556,
+      "step": 16092
+    },
+    {
+      "epoch": 0.257488,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.48516129032258e-05,
+      "loss": 0.1748,
+      "step": 16093
+    },
+    {
+      "epoch": 0.257504,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.485e-05,
+      "loss": 0.1858,
+      "step": 16094
+    },
+    {
+      "epoch": 0.25752,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.48483870967742e-05,
+      "loss": 0.1978,
+      "step": 16095
+    },
+    {
+      "epoch": 0.257536,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.484677419354839e-05,
+      "loss": 0.1993,
+      "step": 16096
+    },
+    {
+      "epoch": 0.257552,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.484516129032259e-05,
+      "loss": 0.1475,
+      "step": 16097
+    },
+    {
+      "epoch": 0.257568,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.484354838709677e-05,
+      "loss": 0.1503,
+      "step": 16098
+    },
+    {
+      "epoch": 0.257584,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.484193548387097e-05,
+      "loss": 0.2112,
+      "step": 16099
+    },
+    {
+      "epoch": 0.2576,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.484032258064516e-05,
+      "loss": 0.1584,
+      "step": 16100
+    },
+    {
+      "epoch": 0.257616,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.483870967741936e-05,
+      "loss": 0.1526,
+      "step": 16101
+    },
+    {
+      "epoch": 0.257632,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.483709677419355e-05,
+      "loss": 0.1427,
+      "step": 16102
+    },
+    {
+      "epoch": 0.257648,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.483548387096775e-05,
+      "loss": 0.1579,
+      "step": 16103
+    },
+    {
+      "epoch": 0.257664,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.483387096774193e-05,
+      "loss": 0.1609,
+      "step": 16104
+    },
+    {
+      "epoch": 0.25768,
+      "grad_norm": 1.515625,
+      "learning_rate": 7.483225806451613e-05,
+      "loss": 0.1878,
+      "step": 16105
+    },
+    {
+      "epoch": 0.257696,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.483064516129033e-05,
+      "loss": 0.1657,
+      "step": 16106
+    },
+    {
+      "epoch": 0.257712,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.482903225806453e-05,
+      "loss": 0.149,
+      "step": 16107
+    },
+    {
+      "epoch": 0.257728,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.482741935483872e-05,
+      "loss": 0.2004,
+      "step": 16108
+    },
+    {
+      "epoch": 0.257744,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.482580645161292e-05,
+      "loss": 0.1842,
+      "step": 16109
+    },
+    {
+      "epoch": 0.25776,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.48241935483871e-05,
+      "loss": 0.1478,
+      "step": 16110
+    },
+    {
+      "epoch": 0.257776,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.482258064516129e-05,
+      "loss": 0.1894,
+      "step": 16111
+    },
+    {
+      "epoch": 0.257792,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.482096774193549e-05,
+      "loss": 0.161,
+      "step": 16112
+    },
+    {
+      "epoch": 0.257808,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.481935483870967e-05,
+      "loss": 0.1441,
+      "step": 16113
+    },
+    {
+      "epoch": 0.257824,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.481774193548387e-05,
+      "loss": 0.1844,
+      "step": 16114
+    },
+    {
+      "epoch": 0.25784,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.481612903225806e-05,
+      "loss": 0.1691,
+      "step": 16115
+    },
+    {
+      "epoch": 0.257856,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.481451612903226e-05,
+      "loss": 0.172,
+      "step": 16116
+    },
+    {
+      "epoch": 0.257872,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.481290322580645e-05,
+      "loss": 0.1935,
+      "step": 16117
+    },
+    {
+      "epoch": 0.257888,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.481129032258065e-05,
+      "loss": 0.1895,
+      "step": 16118
+    },
+    {
+      "epoch": 0.257904,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.480967741935485e-05,
+      "loss": 0.1567,
+      "step": 16119
+    },
+    {
+      "epoch": 0.25792,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.480806451612904e-05,
+      "loss": 0.1731,
+      "step": 16120
+    },
+    {
+      "epoch": 0.257936,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.480645161290323e-05,
+      "loss": 0.1787,
+      "step": 16121
+    },
+    {
+      "epoch": 0.257952,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.480483870967743e-05,
+      "loss": 0.183,
+      "step": 16122
+    },
+    {
+      "epoch": 0.257968,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.480322580645162e-05,
+      "loss": 0.151,
+      "step": 16123
+    },
+    {
+      "epoch": 0.257984,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.480161290322582e-05,
+      "loss": 0.1471,
+      "step": 16124
+    },
+    {
+      "epoch": 0.258,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.48e-05,
+      "loss": 0.2019,
+      "step": 16125
+    },
+    {
+      "epoch": 0.258016,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.479838709677419e-05,
+      "loss": 0.1521,
+      "step": 16126
+    },
+    {
+      "epoch": 0.258032,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.479677419354839e-05,
+      "loss": 0.1794,
+      "step": 16127
+    },
+    {
+      "epoch": 0.258048,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.479516129032257e-05,
+      "loss": 0.1222,
+      "step": 16128
+    },
+    {
+      "epoch": 0.258064,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.479354838709677e-05,
+      "loss": 0.1818,
+      "step": 16129
+    },
+    {
+      "epoch": 0.25808,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.479193548387097e-05,
+      "loss": 0.2008,
+      "step": 16130
+    },
+    {
+      "epoch": 0.258096,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.479032258064517e-05,
+      "loss": 0.1648,
+      "step": 16131
+    },
+    {
+      "epoch": 0.258112,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.478870967741936e-05,
+      "loss": 0.1601,
+      "step": 16132
+    },
+    {
+      "epoch": 0.258128,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.478709677419356e-05,
+      "loss": 0.164,
+      "step": 16133
+    },
+    {
+      "epoch": 0.258144,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.478548387096774e-05,
+      "loss": 0.1834,
+      "step": 16134
+    },
+    {
+      "epoch": 0.25816,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.478387096774194e-05,
+      "loss": 0.1571,
+      "step": 16135
+    },
+    {
+      "epoch": 0.258176,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.478225806451613e-05,
+      "loss": 0.1428,
+      "step": 16136
+    },
+    {
+      "epoch": 0.258192,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.478064516129033e-05,
+      "loss": 0.1961,
+      "step": 16137
+    },
+    {
+      "epoch": 0.258208,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.477903225806452e-05,
+      "loss": 0.1921,
+      "step": 16138
+    },
+    {
+      "epoch": 0.258224,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.477741935483872e-05,
+      "loss": 0.1686,
+      "step": 16139
+    },
+    {
+      "epoch": 0.25824,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.47758064516129e-05,
+      "loss": 0.1778,
+      "step": 16140
+    },
+    {
+      "epoch": 0.258256,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.47741935483871e-05,
+      "loss": 0.1584,
+      "step": 16141
+    },
+    {
+      "epoch": 0.258272,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.47725806451613e-05,
+      "loss": 0.1672,
+      "step": 16142
+    },
+    {
+      "epoch": 0.258288,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.477096774193549e-05,
+      "loss": 0.1552,
+      "step": 16143
+    },
+    {
+      "epoch": 0.258304,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.476935483870969e-05,
+      "loss": 0.1285,
+      "step": 16144
+    },
+    {
+      "epoch": 0.25832,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.476774193548387e-05,
+      "loss": 0.1732,
+      "step": 16145
+    },
+    {
+      "epoch": 0.258336,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.476612903225807e-05,
+      "loss": 0.1724,
+      "step": 16146
+    },
+    {
+      "epoch": 0.258352,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.476451612903226e-05,
+      "loss": 0.1909,
+      "step": 16147
+    },
+    {
+      "epoch": 0.258368,
+      "grad_norm": 1.328125,
+      "learning_rate": 7.476290322580646e-05,
+      "loss": 0.1883,
+      "step": 16148
+    },
+    {
+      "epoch": 0.258384,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.476129032258064e-05,
+      "loss": 0.1756,
+      "step": 16149
+    },
+    {
+      "epoch": 0.2584,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.475967741935484e-05,
+      "loss": 0.1789,
+      "step": 16150
+    },
+    {
+      "epoch": 0.258416,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.475806451612903e-05,
+      "loss": 0.1932,
+      "step": 16151
+    },
+    {
+      "epoch": 0.258432,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.475645161290323e-05,
+      "loss": 0.1758,
+      "step": 16152
+    },
+    {
+      "epoch": 0.258448,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.475483870967742e-05,
+      "loss": 0.1173,
+      "step": 16153
+    },
+    {
+      "epoch": 0.258464,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.475322580645162e-05,
+      "loss": 0.1529,
+      "step": 16154
+    },
+    {
+      "epoch": 0.25848,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.475161290322581e-05,
+      "loss": 0.1813,
+      "step": 16155
+    },
+    {
+      "epoch": 0.258496,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.475000000000001e-05,
+      "loss": 0.1286,
+      "step": 16156
+    },
+    {
+      "epoch": 0.258512,
+      "grad_norm": 1.3359375,
+      "learning_rate": 7.47483870967742e-05,
+      "loss": 0.207,
+      "step": 16157
+    },
+    {
+      "epoch": 0.258528,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.474677419354839e-05,
+      "loss": 0.1557,
+      "step": 16158
+    },
+    {
+      "epoch": 0.258544,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.474516129032259e-05,
+      "loss": 0.1493,
+      "step": 16159
+    },
+    {
+      "epoch": 0.25856,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.474354838709677e-05,
+      "loss": 0.1742,
+      "step": 16160
+    },
+    {
+      "epoch": 0.258576,
+      "grad_norm": 1.34375,
+      "learning_rate": 7.474193548387097e-05,
+      "loss": 0.1763,
+      "step": 16161
+    },
+    {
+      "epoch": 0.258592,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.474032258064516e-05,
+      "loss": 0.1837,
+      "step": 16162
+    },
+    {
+      "epoch": 0.258608,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.473870967741936e-05,
+      "loss": 0.1575,
+      "step": 16163
+    },
+    {
+      "epoch": 0.258624,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.473709677419354e-05,
+      "loss": 0.171,
+      "step": 16164
+    },
+    {
+      "epoch": 0.25864,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.473548387096774e-05,
+      "loss": 0.1852,
+      "step": 16165
+    },
+    {
+      "epoch": 0.258656,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.473387096774194e-05,
+      "loss": 0.1699,
+      "step": 16166
+    },
+    {
+      "epoch": 0.258672,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.473225806451614e-05,
+      "loss": 0.1902,
+      "step": 16167
+    },
+    {
+      "epoch": 0.258688,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.473064516129033e-05,
+      "loss": 0.1756,
+      "step": 16168
+    },
+    {
+      "epoch": 0.258704,
+      "grad_norm": 1.375,
+      "learning_rate": 7.472903225806453e-05,
+      "loss": 0.1377,
+      "step": 16169
+    },
+    {
+      "epoch": 0.25872,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.472741935483871e-05,
+      "loss": 0.2068,
+      "step": 16170
+    },
+    {
+      "epoch": 0.258736,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.472580645161291e-05,
+      "loss": 0.1941,
+      "step": 16171
+    },
+    {
+      "epoch": 0.258752,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.47241935483871e-05,
+      "loss": 0.1693,
+      "step": 16172
+    },
+    {
+      "epoch": 0.258768,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.472258064516129e-05,
+      "loss": 0.2054,
+      "step": 16173
+    },
+    {
+      "epoch": 0.258784,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.472096774193549e-05,
+      "loss": 0.136,
+      "step": 16174
+    },
+    {
+      "epoch": 0.2588,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.471935483870967e-05,
+      "loss": 0.1963,
+      "step": 16175
+    },
+    {
+      "epoch": 0.258816,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.471774193548387e-05,
+      "loss": 0.2025,
+      "step": 16176
+    },
+    {
+      "epoch": 0.258832,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.471612903225806e-05,
+      "loss": 0.1708,
+      "step": 16177
+    },
+    {
+      "epoch": 0.258848,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.471451612903226e-05,
+      "loss": 0.1594,
+      "step": 16178
+    },
+    {
+      "epoch": 0.258864,
+      "grad_norm": 1.3359375,
+      "learning_rate": 7.471290322580646e-05,
+      "loss": 0.2457,
+      "step": 16179
+    },
+    {
+      "epoch": 0.25888,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.471129032258066e-05,
+      "loss": 0.1819,
+      "step": 16180
+    },
+    {
+      "epoch": 0.258896,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.470967741935484e-05,
+      "loss": 0.1594,
+      "step": 16181
+    },
+    {
+      "epoch": 0.258912,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.470806451612904e-05,
+      "loss": 0.1619,
+      "step": 16182
+    },
+    {
+      "epoch": 0.258928,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.470645161290323e-05,
+      "loss": 0.1599,
+      "step": 16183
+    },
+    {
+      "epoch": 0.258944,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.470483870967743e-05,
+      "loss": 0.1622,
+      "step": 16184
+    },
+    {
+      "epoch": 0.25896,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.470322580645161e-05,
+      "loss": 0.1771,
+      "step": 16185
+    },
+    {
+      "epoch": 0.258976,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.470161290322581e-05,
+      "loss": 0.1536,
+      "step": 16186
+    },
+    {
+      "epoch": 0.258992,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.47e-05,
+      "loss": 0.1315,
+      "step": 16187
+    },
+    {
+      "epoch": 0.259008,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.469838709677419e-05,
+      "loss": 0.1711,
+      "step": 16188
+    },
+    {
+      "epoch": 0.259024,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.469677419354839e-05,
+      "loss": 0.1744,
+      "step": 16189
+    },
+    {
+      "epoch": 0.25904,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.469516129032259e-05,
+      "loss": 0.2508,
+      "step": 16190
+    },
+    {
+      "epoch": 0.259056,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.469354838709678e-05,
+      "loss": 0.2024,
+      "step": 16191
+    },
+    {
+      "epoch": 0.259072,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.469193548387097e-05,
+      "loss": 0.2018,
+      "step": 16192
+    },
+    {
+      "epoch": 0.259088,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.469032258064517e-05,
+      "loss": 0.2219,
+      "step": 16193
+    },
+    {
+      "epoch": 0.259104,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.468870967741936e-05,
+      "loss": 0.178,
+      "step": 16194
+    },
+    {
+      "epoch": 0.25912,
+      "grad_norm": 0.75,
+      "learning_rate": 7.468709677419356e-05,
+      "loss": 0.1572,
+      "step": 16195
+    },
+    {
+      "epoch": 0.259136,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.468548387096774e-05,
+      "loss": 0.1238,
+      "step": 16196
+    },
+    {
+      "epoch": 0.259152,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.468387096774194e-05,
+      "loss": 0.1629,
+      "step": 16197
+    },
+    {
+      "epoch": 0.259168,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.468225806451613e-05,
+      "loss": 0.1642,
+      "step": 16198
+    },
+    {
+      "epoch": 0.259184,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.468064516129033e-05,
+      "loss": 0.1957,
+      "step": 16199
+    },
+    {
+      "epoch": 0.2592,
+      "grad_norm": 1.0,
+      "learning_rate": 7.467903225806451e-05,
+      "loss": 0.1562,
+      "step": 16200
+    },
+    {
+      "epoch": 0.259216,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.467741935483871e-05,
+      "loss": 0.1747,
+      "step": 16201
+    },
+    {
+      "epoch": 0.259232,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.467580645161291e-05,
+      "loss": 0.1547,
+      "step": 16202
+    },
+    {
+      "epoch": 0.259248,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.467419354838711e-05,
+      "loss": 0.1836,
+      "step": 16203
+    },
+    {
+      "epoch": 0.259264,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.46725806451613e-05,
+      "loss": 0.1592,
+      "step": 16204
+    },
+    {
+      "epoch": 0.25928,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.467096774193548e-05,
+      "loss": 0.1315,
+      "step": 16205
+    },
+    {
+      "epoch": 0.259296,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.466935483870968e-05,
+      "loss": 0.1507,
+      "step": 16206
+    },
+    {
+      "epoch": 0.259312,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.466774193548387e-05,
+      "loss": 0.1833,
+      "step": 16207
+    },
+    {
+      "epoch": 0.259328,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.466612903225807e-05,
+      "loss": 0.1963,
+      "step": 16208
+    },
+    {
+      "epoch": 0.259344,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.466451612903226e-05,
+      "loss": 0.1733,
+      "step": 16209
+    },
+    {
+      "epoch": 0.25936,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.466290322580646e-05,
+      "loss": 0.1883,
+      "step": 16210
+    },
+    {
+      "epoch": 0.259376,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.466129032258064e-05,
+      "loss": 0.1492,
+      "step": 16211
+    },
+    {
+      "epoch": 0.259392,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.465967741935484e-05,
+      "loss": 0.1554,
+      "step": 16212
+    },
+    {
+      "epoch": 0.259408,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.465806451612903e-05,
+      "loss": 0.205,
+      "step": 16213
+    },
+    {
+      "epoch": 0.259424,
+      "grad_norm": 0.5078125,
+      "learning_rate": 7.465645161290323e-05,
+      "loss": 0.1422,
+      "step": 16214
+    },
+    {
+      "epoch": 0.25944,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.465483870967743e-05,
+      "loss": 0.1832,
+      "step": 16215
+    },
+    {
+      "epoch": 0.259456,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.465322580645163e-05,
+      "loss": 0.2005,
+      "step": 16216
+    },
+    {
+      "epoch": 0.259472,
+      "grad_norm": 1.4921875,
+      "learning_rate": 7.465161290322581e-05,
+      "loss": 0.1668,
+      "step": 16217
+    },
+    {
+      "epoch": 0.259488,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.465000000000001e-05,
+      "loss": 0.1726,
+      "step": 16218
+    },
+    {
+      "epoch": 0.259504,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.46483870967742e-05,
+      "loss": 0.173,
+      "step": 16219
+    },
+    {
+      "epoch": 0.25952,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.464677419354838e-05,
+      "loss": 0.1411,
+      "step": 16220
+    },
+    {
+      "epoch": 0.259536,
+      "grad_norm": 1.25,
+      "learning_rate": 7.464516129032258e-05,
+      "loss": 0.2081,
+      "step": 16221
+    },
+    {
+      "epoch": 0.259552,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.464354838709677e-05,
+      "loss": 0.1777,
+      "step": 16222
+    },
+    {
+      "epoch": 0.259568,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.464193548387097e-05,
+      "loss": 0.1978,
+      "step": 16223
+    },
+    {
+      "epoch": 0.259584,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.464032258064516e-05,
+      "loss": 0.1351,
+      "step": 16224
+    },
+    {
+      "epoch": 0.2596,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.463870967741936e-05,
+      "loss": 0.1384,
+      "step": 16225
+    },
+    {
+      "epoch": 0.259616,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.463709677419355e-05,
+      "loss": 0.1684,
+      "step": 16226
+    },
+    {
+      "epoch": 0.259632,
+      "grad_norm": 1.359375,
+      "learning_rate": 7.463548387096775e-05,
+      "loss": 0.2025,
+      "step": 16227
+    },
+    {
+      "epoch": 0.259648,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.463387096774194e-05,
+      "loss": 0.1774,
+      "step": 16228
+    },
+    {
+      "epoch": 0.259664,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.463225806451614e-05,
+      "loss": 0.159,
+      "step": 16229
+    },
+    {
+      "epoch": 0.25968,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.463064516129033e-05,
+      "loss": 0.1751,
+      "step": 16230
+    },
+    {
+      "epoch": 0.259696,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.462903225806453e-05,
+      "loss": 0.173,
+      "step": 16231
+    },
+    {
+      "epoch": 0.259712,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.462741935483871e-05,
+      "loss": 0.2068,
+      "step": 16232
+    },
+    {
+      "epoch": 0.259728,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.462580645161291e-05,
+      "loss": 0.1624,
+      "step": 16233
+    },
+    {
+      "epoch": 0.259744,
+      "grad_norm": 1.5078125,
+      "learning_rate": 7.46241935483871e-05,
+      "loss": 0.231,
+      "step": 16234
+    },
+    {
+      "epoch": 0.25976,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.462258064516128e-05,
+      "loss": 0.1521,
+      "step": 16235
+    },
+    {
+      "epoch": 0.259776,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.462096774193548e-05,
+      "loss": 0.1653,
+      "step": 16236
+    },
+    {
+      "epoch": 0.259792,
+      "grad_norm": 1.0,
+      "learning_rate": 7.461935483870968e-05,
+      "loss": 0.1693,
+      "step": 16237
+    },
+    {
+      "epoch": 0.259808,
+      "grad_norm": 0.75,
+      "learning_rate": 7.461774193548387e-05,
+      "loss": 0.1753,
+      "step": 16238
+    },
+    {
+      "epoch": 0.259824,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.461612903225807e-05,
+      "loss": 0.178,
+      "step": 16239
+    },
+    {
+      "epoch": 0.25984,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.461451612903227e-05,
+      "loss": 0.2086,
+      "step": 16240
+    },
+    {
+      "epoch": 0.259856,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.461290322580645e-05,
+      "loss": 0.1658,
+      "step": 16241
+    },
+    {
+      "epoch": 0.259872,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.461129032258065e-05,
+      "loss": 0.1598,
+      "step": 16242
+    },
+    {
+      "epoch": 0.259888,
+      "grad_norm": 0.75,
+      "learning_rate": 7.460967741935484e-05,
+      "loss": 0.1984,
+      "step": 16243
+    },
+    {
+      "epoch": 0.259904,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.460806451612904e-05,
+      "loss": 0.1614,
+      "step": 16244
+    },
+    {
+      "epoch": 0.25992,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.460645161290323e-05,
+      "loss": 0.1799,
+      "step": 16245
+    },
+    {
+      "epoch": 0.259936,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.460483870967743e-05,
+      "loss": 0.1306,
+      "step": 16246
+    },
+    {
+      "epoch": 0.259952,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.460322580645161e-05,
+      "loss": 0.193,
+      "step": 16247
+    },
+    {
+      "epoch": 0.259968,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.460161290322581e-05,
+      "loss": 0.172,
+      "step": 16248
+    },
+    {
+      "epoch": 0.259984,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.46e-05,
+      "loss": 0.1904,
+      "step": 16249
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.75,
+      "learning_rate": 7.45983870967742e-05,
+      "loss": 0.143,
+      "step": 16250
+    },
+    {
+      "epoch": 0.260016,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.45967741935484e-05,
+      "loss": 0.1671,
+      "step": 16251
+    },
+    {
+      "epoch": 0.260032,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.459516129032258e-05,
+      "loss": 0.1857,
+      "step": 16252
+    },
+    {
+      "epoch": 0.260048,
+      "grad_norm": 1.3046875,
+      "learning_rate": 7.459354838709678e-05,
+      "loss": 0.1969,
+      "step": 16253
+    },
+    {
+      "epoch": 0.260064,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.459193548387097e-05,
+      "loss": 0.1725,
+      "step": 16254
+    },
+    {
+      "epoch": 0.26008,
+      "grad_norm": 0.47265625,
+      "learning_rate": 7.459032258064517e-05,
+      "loss": 0.138,
+      "step": 16255
+    },
+    {
+      "epoch": 0.260096,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.458870967741935e-05,
+      "loss": 0.1758,
+      "step": 16256
+    },
+    {
+      "epoch": 0.260112,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.458709677419355e-05,
+      "loss": 0.171,
+      "step": 16257
+    },
+    {
+      "epoch": 0.260128,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.458548387096774e-05,
+      "loss": 0.1658,
+      "step": 16258
+    },
+    {
+      "epoch": 0.260144,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.458387096774194e-05,
+      "loss": 0.181,
+      "step": 16259
+    },
+    {
+      "epoch": 0.26016,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.458225806451613e-05,
+      "loss": 0.1575,
+      "step": 16260
+    },
+    {
+      "epoch": 0.260176,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.458064516129033e-05,
+      "loss": 0.1623,
+      "step": 16261
+    },
+    {
+      "epoch": 0.260192,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.457903225806452e-05,
+      "loss": 0.1514,
+      "step": 16262
+    },
+    {
+      "epoch": 0.260208,
+      "grad_norm": 0.75,
+      "learning_rate": 7.457741935483872e-05,
+      "loss": 0.1548,
+      "step": 16263
+    },
+    {
+      "epoch": 0.260224,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.457580645161291e-05,
+      "loss": 0.1866,
+      "step": 16264
+    },
+    {
+      "epoch": 0.26024,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.457419354838711e-05,
+      "loss": 0.1964,
+      "step": 16265
+    },
+    {
+      "epoch": 0.260256,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.45725806451613e-05,
+      "loss": 0.1815,
+      "step": 16266
+    },
+    {
+      "epoch": 0.260272,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.457096774193548e-05,
+      "loss": 0.1599,
+      "step": 16267
+    },
+    {
+      "epoch": 0.260288,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.456935483870968e-05,
+      "loss": 0.1634,
+      "step": 16268
+    },
+    {
+      "epoch": 0.260304,
+      "grad_norm": 0.625,
+      "learning_rate": 7.456774193548387e-05,
+      "loss": 0.136,
+      "step": 16269
+    },
+    {
+      "epoch": 0.26032,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.456612903225807e-05,
+      "loss": 0.1478,
+      "step": 16270
+    },
+    {
+      "epoch": 0.260336,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.456451612903225e-05,
+      "loss": 0.1465,
+      "step": 16271
+    },
+    {
+      "epoch": 0.260352,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.456290322580645e-05,
+      "loss": 0.1436,
+      "step": 16272
+    },
+    {
+      "epoch": 0.260368,
+      "grad_norm": 0.470703125,
+      "learning_rate": 7.456129032258064e-05,
+      "loss": 0.1702,
+      "step": 16273
+    },
+    {
+      "epoch": 0.260384,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.455967741935484e-05,
+      "loss": 0.2148,
+      "step": 16274
+    },
+    {
+      "epoch": 0.2604,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.455806451612904e-05,
+      "loss": 0.164,
+      "step": 16275
+    },
+    {
+      "epoch": 0.260416,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.455645161290324e-05,
+      "loss": 0.1655,
+      "step": 16276
+    },
+    {
+      "epoch": 0.260432,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.455483870967742e-05,
+      "loss": 0.204,
+      "step": 16277
+    },
+    {
+      "epoch": 0.260448,
+      "grad_norm": 1.25,
+      "learning_rate": 7.455322580645162e-05,
+      "loss": 0.1738,
+      "step": 16278
+    },
+    {
+      "epoch": 0.260464,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.455161290322581e-05,
+      "loss": 0.167,
+      "step": 16279
+    },
+    {
+      "epoch": 0.26048,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.455000000000001e-05,
+      "loss": 0.166,
+      "step": 16280
+    },
+    {
+      "epoch": 0.260496,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.45483870967742e-05,
+      "loss": 0.1627,
+      "step": 16281
+    },
+    {
+      "epoch": 0.260512,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.454677419354838e-05,
+      "loss": 0.1694,
+      "step": 16282
+    },
+    {
+      "epoch": 0.260528,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.454516129032258e-05,
+      "loss": 0.1736,
+      "step": 16283
+    },
+    {
+      "epoch": 0.260544,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.454354838709677e-05,
+      "loss": 0.1918,
+      "step": 16284
+    },
+    {
+      "epoch": 0.26056,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.454193548387097e-05,
+      "loss": 0.2129,
+      "step": 16285
+    },
+    {
+      "epoch": 0.260576,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.454032258064517e-05,
+      "loss": 0.1902,
+      "step": 16286
+    },
+    {
+      "epoch": 0.260592,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.453870967741937e-05,
+      "loss": 0.1375,
+      "step": 16287
+    },
+    {
+      "epoch": 0.260608,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.453709677419355e-05,
+      "loss": 0.171,
+      "step": 16288
+    },
+    {
+      "epoch": 0.260624,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.453548387096775e-05,
+      "loss": 0.1378,
+      "step": 16289
+    },
+    {
+      "epoch": 0.26064,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.453387096774194e-05,
+      "loss": 0.1605,
+      "step": 16290
+    },
+    {
+      "epoch": 0.260656,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.453225806451614e-05,
+      "loss": 0.1578,
+      "step": 16291
+    },
+    {
+      "epoch": 0.260672,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.453064516129032e-05,
+      "loss": 0.1611,
+      "step": 16292
+    },
+    {
+      "epoch": 0.260688,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.452903225806452e-05,
+      "loss": 0.2072,
+      "step": 16293
+    },
+    {
+      "epoch": 0.260704,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.452741935483871e-05,
+      "loss": 0.219,
+      "step": 16294
+    },
+    {
+      "epoch": 0.26072,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.452580645161291e-05,
+      "loss": 0.1555,
+      "step": 16295
+    },
+    {
+      "epoch": 0.260736,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.45241935483871e-05,
+      "loss": 0.1849,
+      "step": 16296
+    },
+    {
+      "epoch": 0.260752,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.45225806451613e-05,
+      "loss": 0.1912,
+      "step": 16297
+    },
+    {
+      "epoch": 0.260768,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.45209677419355e-05,
+      "loss": 0.2084,
+      "step": 16298
+    },
+    {
+      "epoch": 0.260784,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.451935483870968e-05,
+      "loss": 0.134,
+      "step": 16299
+    },
+    {
+      "epoch": 0.2608,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.451774193548388e-05,
+      "loss": 0.1712,
+      "step": 16300
+    },
+    {
+      "epoch": 0.260816,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.451612903225807e-05,
+      "loss": 0.2089,
+      "step": 16301
+    },
+    {
+      "epoch": 0.260832,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.451451612903227e-05,
+      "loss": 0.1904,
+      "step": 16302
+    },
+    {
+      "epoch": 0.260848,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.451290322580645e-05,
+      "loss": 0.154,
+      "step": 16303
+    },
+    {
+      "epoch": 0.260864,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.451129032258065e-05,
+      "loss": 0.1436,
+      "step": 16304
+    },
+    {
+      "epoch": 0.26088,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.450967741935484e-05,
+      "loss": 0.146,
+      "step": 16305
+    },
+    {
+      "epoch": 0.260896,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.450806451612904e-05,
+      "loss": 0.156,
+      "step": 16306
+    },
+    {
+      "epoch": 0.260912,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.450645161290322e-05,
+      "loss": 0.1785,
+      "step": 16307
+    },
+    {
+      "epoch": 0.260928,
+      "grad_norm": 1.4296875,
+      "learning_rate": 7.450483870967742e-05,
+      "loss": 0.1931,
+      "step": 16308
+    },
+    {
+      "epoch": 0.260944,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.450322580645161e-05,
+      "loss": 0.1551,
+      "step": 16309
+    },
+    {
+      "epoch": 0.26096,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.450161290322581e-05,
+      "loss": 0.1828,
+      "step": 16310
+    },
+    {
+      "epoch": 0.260976,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.450000000000001e-05,
+      "loss": 0.1723,
+      "step": 16311
+    },
+    {
+      "epoch": 0.260992,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.449838709677421e-05,
+      "loss": 0.1654,
+      "step": 16312
+    },
+    {
+      "epoch": 0.261008,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.44967741935484e-05,
+      "loss": 0.1656,
+      "step": 16313
+    },
+    {
+      "epoch": 0.261024,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.449516129032258e-05,
+      "loss": 0.2101,
+      "step": 16314
+    },
+    {
+      "epoch": 0.26104,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.449354838709678e-05,
+      "loss": 0.2028,
+      "step": 16315
+    },
+    {
+      "epoch": 0.261056,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.449193548387097e-05,
+      "loss": 0.149,
+      "step": 16316
+    },
+    {
+      "epoch": 0.261072,
+      "grad_norm": 0.875,
+      "learning_rate": 7.449032258064517e-05,
+      "loss": 0.1478,
+      "step": 16317
+    },
+    {
+      "epoch": 0.261088,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.448870967741935e-05,
+      "loss": 0.1673,
+      "step": 16318
+    },
+    {
+      "epoch": 0.261104,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.448709677419355e-05,
+      "loss": 0.1442,
+      "step": 16319
+    },
+    {
+      "epoch": 0.26112,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.448548387096774e-05,
+      "loss": 0.1633,
+      "step": 16320
+    },
+    {
+      "epoch": 0.261136,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.448387096774194e-05,
+      "loss": 0.2019,
+      "step": 16321
+    },
+    {
+      "epoch": 0.261152,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.448225806451614e-05,
+      "loss": 0.1647,
+      "step": 16322
+    },
+    {
+      "epoch": 0.261168,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.448064516129034e-05,
+      "loss": 0.1936,
+      "step": 16323
+    },
+    {
+      "epoch": 0.261184,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.447903225806452e-05,
+      "loss": 0.1571,
+      "step": 16324
+    },
+    {
+      "epoch": 0.2612,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.447741935483872e-05,
+      "loss": 0.1838,
+      "step": 16325
+    },
+    {
+      "epoch": 0.261216,
+      "grad_norm": 0.875,
+      "learning_rate": 7.447580645161291e-05,
+      "loss": 0.2085,
+      "step": 16326
+    },
+    {
+      "epoch": 0.261232,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.447419354838711e-05,
+      "loss": 0.1633,
+      "step": 16327
+    },
+    {
+      "epoch": 0.261248,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.44725806451613e-05,
+      "loss": 0.1642,
+      "step": 16328
+    },
+    {
+      "epoch": 0.261264,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.447096774193548e-05,
+      "loss": 0.1823,
+      "step": 16329
+    },
+    {
+      "epoch": 0.26128,
+      "grad_norm": 0.625,
+      "learning_rate": 7.446935483870968e-05,
+      "loss": 0.1574,
+      "step": 16330
+    },
+    {
+      "epoch": 0.261296,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.446774193548387e-05,
+      "loss": 0.1827,
+      "step": 16331
+    },
+    {
+      "epoch": 0.261312,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.446612903225807e-05,
+      "loss": 0.1604,
+      "step": 16332
+    },
+    {
+      "epoch": 0.261328,
+      "grad_norm": 1.296875,
+      "learning_rate": 7.446451612903226e-05,
+      "loss": 0.2195,
+      "step": 16333
+    },
+    {
+      "epoch": 0.261344,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.446290322580645e-05,
+      "loss": 0.1718,
+      "step": 16334
+    },
+    {
+      "epoch": 0.26136,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.446129032258065e-05,
+      "loss": 0.2078,
+      "step": 16335
+    },
+    {
+      "epoch": 0.261376,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.445967741935485e-05,
+      "loss": 0.1909,
+      "step": 16336
+    },
+    {
+      "epoch": 0.261392,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.445806451612904e-05,
+      "loss": 0.1907,
+      "step": 16337
+    },
+    {
+      "epoch": 0.261408,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.445645161290324e-05,
+      "loss": 0.1634,
+      "step": 16338
+    },
+    {
+      "epoch": 0.261424,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.445483870967742e-05,
+      "loss": 0.1902,
+      "step": 16339
+    },
+    {
+      "epoch": 0.26144,
+      "grad_norm": 0.49609375,
+      "learning_rate": 7.445322580645162e-05,
+      "loss": 0.1747,
+      "step": 16340
+    },
+    {
+      "epoch": 0.261456,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.445161290322581e-05,
+      "loss": 0.1936,
+      "step": 16341
+    },
+    {
+      "epoch": 0.261472,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.445000000000001e-05,
+      "loss": 0.1405,
+      "step": 16342
+    },
+    {
+      "epoch": 0.261488,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.44483870967742e-05,
+      "loss": 0.1996,
+      "step": 16343
+    },
+    {
+      "epoch": 0.261504,
+      "grad_norm": 1.375,
+      "learning_rate": 7.444677419354838e-05,
+      "loss": 0.1623,
+      "step": 16344
+    },
+    {
+      "epoch": 0.26152,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.444516129032258e-05,
+      "loss": 0.18,
+      "step": 16345
+    },
+    {
+      "epoch": 0.261536,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.444354838709678e-05,
+      "loss": 0.1805,
+      "step": 16346
+    },
+    {
+      "epoch": 0.261552,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.444193548387098e-05,
+      "loss": 0.206,
+      "step": 16347
+    },
+    {
+      "epoch": 0.261568,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.444032258064516e-05,
+      "loss": 0.1581,
+      "step": 16348
+    },
+    {
+      "epoch": 0.261584,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.443870967741936e-05,
+      "loss": 0.1767,
+      "step": 16349
+    },
+    {
+      "epoch": 0.2616,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.443709677419355e-05,
+      "loss": 0.1558,
+      "step": 16350
+    },
+    {
+      "epoch": 0.261616,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.443548387096775e-05,
+      "loss": 0.1888,
+      "step": 16351
+    },
+    {
+      "epoch": 0.261632,
+      "grad_norm": 0.47265625,
+      "learning_rate": 7.443387096774194e-05,
+      "loss": 0.1293,
+      "step": 16352
+    },
+    {
+      "epoch": 0.261648,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.443225806451614e-05,
+      "loss": 0.1436,
+      "step": 16353
+    },
+    {
+      "epoch": 0.261664,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.443064516129032e-05,
+      "loss": 0.2221,
+      "step": 16354
+    },
+    {
+      "epoch": 0.26168,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.442903225806452e-05,
+      "loss": 0.1904,
+      "step": 16355
+    },
+    {
+      "epoch": 0.261696,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.442741935483871e-05,
+      "loss": 0.167,
+      "step": 16356
+    },
+    {
+      "epoch": 0.261712,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.44258064516129e-05,
+      "loss": 0.162,
+      "step": 16357
+    },
+    {
+      "epoch": 0.261728,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.44241935483871e-05,
+      "loss": 0.1367,
+      "step": 16358
+    },
+    {
+      "epoch": 0.261744,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.442258064516129e-05,
+      "loss": 0.1433,
+      "step": 16359
+    },
+    {
+      "epoch": 0.26176,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.442096774193549e-05,
+      "loss": 0.1984,
+      "step": 16360
+    },
+    {
+      "epoch": 0.261776,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.441935483870968e-05,
+      "loss": 0.1871,
+      "step": 16361
+    },
+    {
+      "epoch": 0.261792,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.441774193548388e-05,
+      "loss": 0.1745,
+      "step": 16362
+    },
+    {
+      "epoch": 0.261808,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.441612903225806e-05,
+      "loss": 0.1752,
+      "step": 16363
+    },
+    {
+      "epoch": 0.261824,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.441451612903226e-05,
+      "loss": 0.1777,
+      "step": 16364
+    },
+    {
+      "epoch": 0.26184,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.441290322580645e-05,
+      "loss": 0.1882,
+      "step": 16365
+    },
+    {
+      "epoch": 0.261856,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.441129032258065e-05,
+      "loss": 0.1681,
+      "step": 16366
+    },
+    {
+      "epoch": 0.261872,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.440967741935484e-05,
+      "loss": 0.1984,
+      "step": 16367
+    },
+    {
+      "epoch": 0.261888,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.440806451612903e-05,
+      "loss": 0.1592,
+      "step": 16368
+    },
+    {
+      "epoch": 0.261904,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.440645161290322e-05,
+      "loss": 0.2068,
+      "step": 16369
+    },
+    {
+      "epoch": 0.26192,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.440483870967742e-05,
+      "loss": 0.1803,
+      "step": 16370
+    },
+    {
+      "epoch": 0.261936,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.440322580645162e-05,
+      "loss": 0.1589,
+      "step": 16371
+    },
+    {
+      "epoch": 0.261952,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.440161290322582e-05,
+      "loss": 0.1732,
+      "step": 16372
+    },
+    {
+      "epoch": 0.261968,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.44e-05,
+      "loss": 0.2306,
+      "step": 16373
+    },
+    {
+      "epoch": 0.261984,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.43983870967742e-05,
+      "loss": 0.1373,
+      "step": 16374
+    },
+    {
+      "epoch": 0.262,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.439677419354839e-05,
+      "loss": 0.215,
+      "step": 16375
+    },
+    {
+      "epoch": 0.262016,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.439516129032258e-05,
+      "loss": 0.178,
+      "step": 16376
+    },
+    {
+      "epoch": 0.262032,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.439354838709678e-05,
+      "loss": 0.1603,
+      "step": 16377
+    },
+    {
+      "epoch": 0.262048,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.439193548387096e-05,
+      "loss": 0.188,
+      "step": 16378
+    },
+    {
+      "epoch": 0.262064,
+      "grad_norm": 1.40625,
+      "learning_rate": 7.439032258064516e-05,
+      "loss": 0.1538,
+      "step": 16379
+    },
+    {
+      "epoch": 0.26208,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.438870967741935e-05,
+      "loss": 0.1893,
+      "step": 16380
+    },
+    {
+      "epoch": 0.262096,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.438709677419355e-05,
+      "loss": 0.1683,
+      "step": 16381
+    },
+    {
+      "epoch": 0.262112,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.438548387096775e-05,
+      "loss": 0.1524,
+      "step": 16382
+    },
+    {
+      "epoch": 0.262128,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.438387096774195e-05,
+      "loss": 0.1945,
+      "step": 16383
+    },
+    {
+      "epoch": 0.262144,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.438225806451613e-05,
+      "loss": 0.1638,
+      "step": 16384
+    },
+    {
+      "epoch": 0.26216,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.438064516129033e-05,
+      "loss": 0.1683,
+      "step": 16385
+    },
+    {
+      "epoch": 0.262176,
+      "grad_norm": 1.2890625,
+      "learning_rate": 7.437903225806452e-05,
+      "loss": 0.1747,
+      "step": 16386
+    },
+    {
+      "epoch": 0.262192,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.437741935483872e-05,
+      "loss": 0.1459,
+      "step": 16387
+    },
+    {
+      "epoch": 0.262208,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.43758064516129e-05,
+      "loss": 0.1727,
+      "step": 16388
+    },
+    {
+      "epoch": 0.262224,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.43741935483871e-05,
+      "loss": 0.1825,
+      "step": 16389
+    },
+    {
+      "epoch": 0.26224,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.437258064516129e-05,
+      "loss": 0.1631,
+      "step": 16390
+    },
+    {
+      "epoch": 0.262256,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.437096774193548e-05,
+      "loss": 0.158,
+      "step": 16391
+    },
+    {
+      "epoch": 0.262272,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.436935483870968e-05,
+      "loss": 0.1606,
+      "step": 16392
+    },
+    {
+      "epoch": 0.262288,
+      "grad_norm": 0.625,
+      "learning_rate": 7.436774193548388e-05,
+      "loss": 0.1265,
+      "step": 16393
+    },
+    {
+      "epoch": 0.262304,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.436612903225808e-05,
+      "loss": 0.1796,
+      "step": 16394
+    },
+    {
+      "epoch": 0.26232,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.436451612903226e-05,
+      "loss": 0.1598,
+      "step": 16395
+    },
+    {
+      "epoch": 0.262336,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.436290322580646e-05,
+      "loss": 0.2056,
+      "step": 16396
+    },
+    {
+      "epoch": 0.262352,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.436129032258065e-05,
+      "loss": 0.1648,
+      "step": 16397
+    },
+    {
+      "epoch": 0.262368,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.435967741935485e-05,
+      "loss": 0.1752,
+      "step": 16398
+    },
+    {
+      "epoch": 0.262384,
+      "grad_norm": 1.40625,
+      "learning_rate": 7.435806451612903e-05,
+      "loss": 0.191,
+      "step": 16399
+    },
+    {
+      "epoch": 0.2624,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.435645161290323e-05,
+      "loss": 0.1517,
+      "step": 16400
+    },
+    {
+      "epoch": 0.262416,
+      "grad_norm": 1.3046875,
+      "learning_rate": 7.435483870967742e-05,
+      "loss": 0.1646,
+      "step": 16401
+    },
+    {
+      "epoch": 0.262432,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.435322580645162e-05,
+      "loss": 0.1598,
+      "step": 16402
+    },
+    {
+      "epoch": 0.262448,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.43516129032258e-05,
+      "loss": 0.1926,
+      "step": 16403
+    },
+    {
+      "epoch": 0.262464,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.435e-05,
+      "loss": 0.1648,
+      "step": 16404
+    },
+    {
+      "epoch": 0.26248,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.434838709677419e-05,
+      "loss": 0.1899,
+      "step": 16405
+    },
+    {
+      "epoch": 0.262496,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.434677419354839e-05,
+      "loss": 0.1688,
+      "step": 16406
+    },
+    {
+      "epoch": 0.262512,
+      "grad_norm": 1.75,
+      "learning_rate": 7.434516129032259e-05,
+      "loss": 0.2347,
+      "step": 16407
+    },
+    {
+      "epoch": 0.262528,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.434354838709678e-05,
+      "loss": 0.1761,
+      "step": 16408
+    },
+    {
+      "epoch": 0.262544,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.434193548387098e-05,
+      "loss": 0.1353,
+      "step": 16409
+    },
+    {
+      "epoch": 0.26256,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.434032258064516e-05,
+      "loss": 0.1905,
+      "step": 16410
+    },
+    {
+      "epoch": 0.262576,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.433870967741936e-05,
+      "loss": 0.1383,
+      "step": 16411
+    },
+    {
+      "epoch": 0.262592,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.433709677419355e-05,
+      "loss": 0.1545,
+      "step": 16412
+    },
+    {
+      "epoch": 0.262608,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.433548387096775e-05,
+      "loss": 0.1351,
+      "step": 16413
+    },
+    {
+      "epoch": 0.262624,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.433387096774193e-05,
+      "loss": 0.1645,
+      "step": 16414
+    },
+    {
+      "epoch": 0.26264,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.433225806451613e-05,
+      "loss": 0.175,
+      "step": 16415
+    },
+    {
+      "epoch": 0.262656,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.433064516129032e-05,
+      "loss": 0.1911,
+      "step": 16416
+    },
+    {
+      "epoch": 0.262672,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.432903225806452e-05,
+      "loss": 0.1588,
+      "step": 16417
+    },
+    {
+      "epoch": 0.262688,
+      "grad_norm": 1.4140625,
+      "learning_rate": 7.432741935483872e-05,
+      "loss": 0.1512,
+      "step": 16418
+    },
+    {
+      "epoch": 0.262704,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.432580645161292e-05,
+      "loss": 0.1616,
+      "step": 16419
+    },
+    {
+      "epoch": 0.26272,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.43241935483871e-05,
+      "loss": 0.1688,
+      "step": 16420
+    },
+    {
+      "epoch": 0.262736,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.43225806451613e-05,
+      "loss": 0.1797,
+      "step": 16421
+    },
+    {
+      "epoch": 0.262752,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.432096774193549e-05,
+      "loss": 0.1656,
+      "step": 16422
+    },
+    {
+      "epoch": 0.262768,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.431935483870968e-05,
+      "loss": 0.1655,
+      "step": 16423
+    },
+    {
+      "epoch": 0.262784,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.431774193548388e-05,
+      "loss": 0.1573,
+      "step": 16424
+    },
+    {
+      "epoch": 0.2628,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.431612903225806e-05,
+      "loss": 0.1698,
+      "step": 16425
+    },
+    {
+      "epoch": 0.262816,
+      "grad_norm": 1.34375,
+      "learning_rate": 7.431451612903226e-05,
+      "loss": 0.1867,
+      "step": 16426
+    },
+    {
+      "epoch": 0.262832,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.431290322580645e-05,
+      "loss": 0.2005,
+      "step": 16427
+    },
+    {
+      "epoch": 0.262848,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.431129032258065e-05,
+      "loss": 0.164,
+      "step": 16428
+    },
+    {
+      "epoch": 0.262864,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.430967741935483e-05,
+      "loss": 0.1892,
+      "step": 16429
+    },
+    {
+      "epoch": 0.26288,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.430806451612903e-05,
+      "loss": 0.1608,
+      "step": 16430
+    },
+    {
+      "epoch": 0.262896,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.430645161290323e-05,
+      "loss": 0.1555,
+      "step": 16431
+    },
+    {
+      "epoch": 0.262912,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.430483870967743e-05,
+      "loss": 0.2019,
+      "step": 16432
+    },
+    {
+      "epoch": 0.262928,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.430322580645162e-05,
+      "loss": 0.1395,
+      "step": 16433
+    },
+    {
+      "epoch": 0.262944,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.430161290322582e-05,
+      "loss": 0.148,
+      "step": 16434
+    },
+    {
+      "epoch": 0.26296,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.43e-05,
+      "loss": 0.1566,
+      "step": 16435
+    },
+    {
+      "epoch": 0.262976,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.42983870967742e-05,
+      "loss": 0.1451,
+      "step": 16436
+    },
+    {
+      "epoch": 0.262992,
+      "grad_norm": 1.3359375,
+      "learning_rate": 7.429677419354839e-05,
+      "loss": 0.2,
+      "step": 16437
+    },
+    {
+      "epoch": 0.263008,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.429516129032258e-05,
+      "loss": 0.1945,
+      "step": 16438
+    },
+    {
+      "epoch": 0.263024,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.429354838709677e-05,
+      "loss": 0.1418,
+      "step": 16439
+    },
+    {
+      "epoch": 0.26304,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.429193548387096e-05,
+      "loss": 0.144,
+      "step": 16440
+    },
+    {
+      "epoch": 0.263056,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.429032258064516e-05,
+      "loss": 0.1501,
+      "step": 16441
+    },
+    {
+      "epoch": 0.263072,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.428870967741936e-05,
+      "loss": 0.1643,
+      "step": 16442
+    },
+    {
+      "epoch": 0.263088,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.428709677419356e-05,
+      "loss": 0.1429,
+      "step": 16443
+    },
+    {
+      "epoch": 0.263104,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.428548387096775e-05,
+      "loss": 0.1759,
+      "step": 16444
+    },
+    {
+      "epoch": 0.26312,
+      "grad_norm": 1.2578125,
+      "learning_rate": 7.428387096774195e-05,
+      "loss": 0.1662,
+      "step": 16445
+    },
+    {
+      "epoch": 0.263136,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.428225806451613e-05,
+      "loss": 0.2095,
+      "step": 16446
+    },
+    {
+      "epoch": 0.263152,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.428064516129033e-05,
+      "loss": 0.1632,
+      "step": 16447
+    },
+    {
+      "epoch": 0.263168,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.427903225806452e-05,
+      "loss": 0.1442,
+      "step": 16448
+    },
+    {
+      "epoch": 0.263184,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.427741935483872e-05,
+      "loss": 0.1888,
+      "step": 16449
+    },
+    {
+      "epoch": 0.2632,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.42758064516129e-05,
+      "loss": 0.1988,
+      "step": 16450
+    },
+    {
+      "epoch": 0.263216,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.42741935483871e-05,
+      "loss": 0.1605,
+      "step": 16451
+    },
+    {
+      "epoch": 0.263232,
+      "grad_norm": 1.0,
+      "learning_rate": 7.427258064516129e-05,
+      "loss": 0.1931,
+      "step": 16452
+    },
+    {
+      "epoch": 0.263248,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.427096774193549e-05,
+      "loss": 0.1483,
+      "step": 16453
+    },
+    {
+      "epoch": 0.263264,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.426935483870969e-05,
+      "loss": 0.1542,
+      "step": 16454
+    },
+    {
+      "epoch": 0.26328,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.426774193548387e-05,
+      "loss": 0.1185,
+      "step": 16455
+    },
+    {
+      "epoch": 0.263296,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.426612903225807e-05,
+      "loss": 0.1956,
+      "step": 16456
+    },
+    {
+      "epoch": 0.263312,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.426451612903226e-05,
+      "loss": 0.1893,
+      "step": 16457
+    },
+    {
+      "epoch": 0.263328,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.426290322580646e-05,
+      "loss": 0.1554,
+      "step": 16458
+    },
+    {
+      "epoch": 0.263344,
+      "grad_norm": 0.423828125,
+      "learning_rate": 7.426129032258065e-05,
+      "loss": 0.1359,
+      "step": 16459
+    },
+    {
+      "epoch": 0.26336,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.425967741935485e-05,
+      "loss": 0.1923,
+      "step": 16460
+    },
+    {
+      "epoch": 0.263376,
+      "grad_norm": 1.265625,
+      "learning_rate": 7.425806451612903e-05,
+      "loss": 0.199,
+      "step": 16461
+    },
+    {
+      "epoch": 0.263392,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.425645161290323e-05,
+      "loss": 0.1553,
+      "step": 16462
+    },
+    {
+      "epoch": 0.263408,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.425483870967742e-05,
+      "loss": 0.1468,
+      "step": 16463
+    },
+    {
+      "epoch": 0.263424,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.425322580645162e-05,
+      "loss": 0.197,
+      "step": 16464
+    },
+    {
+      "epoch": 0.26344,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.42516129032258e-05,
+      "loss": 0.1758,
+      "step": 16465
+    },
+    {
+      "epoch": 0.263456,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.425e-05,
+      "loss": 0.1952,
+      "step": 16466
+    },
+    {
+      "epoch": 0.263472,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.42483870967742e-05,
+      "loss": 0.162,
+      "step": 16467
+    },
+    {
+      "epoch": 0.263488,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.424677419354839e-05,
+      "loss": 0.1891,
+      "step": 16468
+    },
+    {
+      "epoch": 0.263504,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.424516129032259e-05,
+      "loss": 0.1632,
+      "step": 16469
+    },
+    {
+      "epoch": 0.26352,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.424354838709677e-05,
+      "loss": 0.1959,
+      "step": 16470
+    },
+    {
+      "epoch": 0.263536,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.424193548387097e-05,
+      "loss": 0.1375,
+      "step": 16471
+    },
+    {
+      "epoch": 0.263552,
+      "grad_norm": 1.328125,
+      "learning_rate": 7.424032258064516e-05,
+      "loss": 0.1716,
+      "step": 16472
+    },
+    {
+      "epoch": 0.263568,
+      "grad_norm": 1.0,
+      "learning_rate": 7.423870967741936e-05,
+      "loss": 0.1618,
+      "step": 16473
+    },
+    {
+      "epoch": 0.263584,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.423709677419355e-05,
+      "loss": 0.1548,
+      "step": 16474
+    },
+    {
+      "epoch": 0.2636,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.423548387096774e-05,
+      "loss": 0.1803,
+      "step": 16475
+    },
+    {
+      "epoch": 0.263616,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.423387096774193e-05,
+      "loss": 0.1587,
+      "step": 16476
+    },
+    {
+      "epoch": 0.263632,
+      "grad_norm": 0.4609375,
+      "learning_rate": 7.423225806451613e-05,
+      "loss": 0.1541,
+      "step": 16477
+    },
+    {
+      "epoch": 0.263648,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.423064516129033e-05,
+      "loss": 0.2123,
+      "step": 16478
+    },
+    {
+      "epoch": 0.263664,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.422903225806453e-05,
+      "loss": 0.1464,
+      "step": 16479
+    },
+    {
+      "epoch": 0.26368,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.422741935483872e-05,
+      "loss": 0.1543,
+      "step": 16480
+    },
+    {
+      "epoch": 0.263696,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.422580645161292e-05,
+      "loss": 0.2437,
+      "step": 16481
+    },
+    {
+      "epoch": 0.263712,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.42241935483871e-05,
+      "loss": 0.2008,
+      "step": 16482
+    },
+    {
+      "epoch": 0.263728,
+      "grad_norm": 1.265625,
+      "learning_rate": 7.42225806451613e-05,
+      "loss": 0.1471,
+      "step": 16483
+    },
+    {
+      "epoch": 0.263744,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.422096774193549e-05,
+      "loss": 0.1356,
+      "step": 16484
+    },
+    {
+      "epoch": 0.26376,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.421935483870967e-05,
+      "loss": 0.182,
+      "step": 16485
+    },
+    {
+      "epoch": 0.263776,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.421774193548387e-05,
+      "loss": 0.1344,
+      "step": 16486
+    },
+    {
+      "epoch": 0.263792,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.421612903225806e-05,
+      "loss": 0.1194,
+      "step": 16487
+    },
+    {
+      "epoch": 0.263808,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.421451612903226e-05,
+      "loss": 0.1646,
+      "step": 16488
+    },
+    {
+      "epoch": 0.263824,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.421290322580646e-05,
+      "loss": 0.1624,
+      "step": 16489
+    },
+    {
+      "epoch": 0.26384,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.421129032258064e-05,
+      "loss": 0.1621,
+      "step": 16490
+    },
+    {
+      "epoch": 0.263856,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.420967741935484e-05,
+      "loss": 0.1816,
+      "step": 16491
+    },
+    {
+      "epoch": 0.263872,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.420806451612904e-05,
+      "loss": 0.157,
+      "step": 16492
+    },
+    {
+      "epoch": 0.263888,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.420645161290323e-05,
+      "loss": 0.1544,
+      "step": 16493
+    },
+    {
+      "epoch": 0.263904,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.420483870967743e-05,
+      "loss": 0.1648,
+      "step": 16494
+    },
+    {
+      "epoch": 0.26392,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.420322580645162e-05,
+      "loss": 0.1828,
+      "step": 16495
+    },
+    {
+      "epoch": 0.263936,
+      "grad_norm": 1.0,
+      "learning_rate": 7.420161290322582e-05,
+      "loss": 0.1821,
+      "step": 16496
+    },
+    {
+      "epoch": 0.263952,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.42e-05,
+      "loss": 0.1739,
+      "step": 16497
+    },
+    {
+      "epoch": 0.263968,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.41983870967742e-05,
+      "loss": 0.1869,
+      "step": 16498
+    },
+    {
+      "epoch": 0.263984,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.419677419354839e-05,
+      "loss": 0.1965,
+      "step": 16499
+    },
+    {
+      "epoch": 0.264,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.419516129032257e-05,
+      "loss": 0.1712,
+      "step": 16500
+    },
+    {
+      "epoch": 0.264016,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.419354838709677e-05,
+      "loss": 0.1489,
+      "step": 16501
+    },
+    {
+      "epoch": 0.264032,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.419193548387097e-05,
+      "loss": 0.1486,
+      "step": 16502
+    },
+    {
+      "epoch": 0.264048,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.419032258064517e-05,
+      "loss": 0.162,
+      "step": 16503
+    },
+    {
+      "epoch": 0.264064,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.418870967741936e-05,
+      "loss": 0.148,
+      "step": 16504
+    },
+    {
+      "epoch": 0.26408,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.418709677419356e-05,
+      "loss": 0.1575,
+      "step": 16505
+    },
+    {
+      "epoch": 0.264096,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.418548387096774e-05,
+      "loss": 0.1518,
+      "step": 16506
+    },
+    {
+      "epoch": 0.264112,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.418387096774194e-05,
+      "loss": 0.1485,
+      "step": 16507
+    },
+    {
+      "epoch": 0.264128,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.418225806451613e-05,
+      "loss": 0.1432,
+      "step": 16508
+    },
+    {
+      "epoch": 0.264144,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.418064516129033e-05,
+      "loss": 0.1559,
+      "step": 16509
+    },
+    {
+      "epoch": 0.26416,
+      "grad_norm": 1.25,
+      "learning_rate": 7.417903225806451e-05,
+      "loss": 0.1254,
+      "step": 16510
+    },
+    {
+      "epoch": 0.264176,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.417741935483871e-05,
+      "loss": 0.196,
+      "step": 16511
+    },
+    {
+      "epoch": 0.264192,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.41758064516129e-05,
+      "loss": 0.1404,
+      "step": 16512
+    },
+    {
+      "epoch": 0.264208,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.41741935483871e-05,
+      "loss": 0.1684,
+      "step": 16513
+    },
+    {
+      "epoch": 0.264224,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.41725806451613e-05,
+      "loss": 0.1715,
+      "step": 16514
+    },
+    {
+      "epoch": 0.26424,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.417096774193549e-05,
+      "loss": 0.1533,
+      "step": 16515
+    },
+    {
+      "epoch": 0.264256,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.416935483870969e-05,
+      "loss": 0.1713,
+      "step": 16516
+    },
+    {
+      "epoch": 0.264272,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.416774193548387e-05,
+      "loss": 0.1276,
+      "step": 16517
+    },
+    {
+      "epoch": 0.264288,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.416612903225807e-05,
+      "loss": 0.1734,
+      "step": 16518
+    },
+    {
+      "epoch": 0.264304,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.416451612903226e-05,
+      "loss": 0.1474,
+      "step": 16519
+    },
+    {
+      "epoch": 0.26432,
+      "grad_norm": 1.953125,
+      "learning_rate": 7.416290322580646e-05,
+      "loss": 0.1799,
+      "step": 16520
+    },
+    {
+      "epoch": 0.264336,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.416129032258064e-05,
+      "loss": 0.1631,
+      "step": 16521
+    },
+    {
+      "epoch": 0.264352,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.415967741935484e-05,
+      "loss": 0.1588,
+      "step": 16522
+    },
+    {
+      "epoch": 0.264368,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.415806451612903e-05,
+      "loss": 0.1538,
+      "step": 16523
+    },
+    {
+      "epoch": 0.264384,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.415645161290323e-05,
+      "loss": 0.1562,
+      "step": 16524
+    },
+    {
+      "epoch": 0.2644,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.415483870967741e-05,
+      "loss": 0.1697,
+      "step": 16525
+    },
+    {
+      "epoch": 0.264416,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.415322580645161e-05,
+      "loss": 0.2066,
+      "step": 16526
+    },
+    {
+      "epoch": 0.264432,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.415161290322581e-05,
+      "loss": 0.1599,
+      "step": 16527
+    },
+    {
+      "epoch": 0.264448,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.415000000000001e-05,
+      "loss": 0.1746,
+      "step": 16528
+    },
+    {
+      "epoch": 0.264464,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.41483870967742e-05,
+      "loss": 0.1361,
+      "step": 16529
+    },
+    {
+      "epoch": 0.26448,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.41467741935484e-05,
+      "loss": 0.1576,
+      "step": 16530
+    },
+    {
+      "epoch": 0.264496,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.414516129032259e-05,
+      "loss": 0.1949,
+      "step": 16531
+    },
+    {
+      "epoch": 0.264512,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.414354838709677e-05,
+      "loss": 0.1755,
+      "step": 16532
+    },
+    {
+      "epoch": 0.264528,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.414193548387097e-05,
+      "loss": 0.1901,
+      "step": 16533
+    },
+    {
+      "epoch": 0.264544,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.414032258064516e-05,
+      "loss": 0.1985,
+      "step": 16534
+    },
+    {
+      "epoch": 0.26456,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.413870967741936e-05,
+      "loss": 0.1681,
+      "step": 16535
+    },
+    {
+      "epoch": 0.264576,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.413709677419354e-05,
+      "loss": 0.16,
+      "step": 16536
+    },
+    {
+      "epoch": 0.264592,
+      "grad_norm": 0.75,
+      "learning_rate": 7.413548387096774e-05,
+      "loss": 0.1576,
+      "step": 16537
+    },
+    {
+      "epoch": 0.264608,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.413387096774194e-05,
+      "loss": 0.1249,
+      "step": 16538
+    },
+    {
+      "epoch": 0.264624,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.413225806451614e-05,
+      "loss": 0.166,
+      "step": 16539
+    },
+    {
+      "epoch": 0.26464,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.413064516129033e-05,
+      "loss": 0.1767,
+      "step": 16540
+    },
+    {
+      "epoch": 0.264656,
+      "grad_norm": 0.625,
+      "learning_rate": 7.412903225806453e-05,
+      "loss": 0.1616,
+      "step": 16541
+    },
+    {
+      "epoch": 0.264672,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.412741935483871e-05,
+      "loss": 0.1601,
+      "step": 16542
+    },
+    {
+      "epoch": 0.264688,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.412580645161291e-05,
+      "loss": 0.1593,
+      "step": 16543
+    },
+    {
+      "epoch": 0.264704,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.41241935483871e-05,
+      "loss": 0.1372,
+      "step": 16544
+    },
+    {
+      "epoch": 0.26472,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.41225806451613e-05,
+      "loss": 0.1716,
+      "step": 16545
+    },
+    {
+      "epoch": 0.264736,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.412096774193548e-05,
+      "loss": 0.1595,
+      "step": 16546
+    },
+    {
+      "epoch": 0.264752,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.411935483870967e-05,
+      "loss": 0.1917,
+      "step": 16547
+    },
+    {
+      "epoch": 0.264768,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.411774193548387e-05,
+      "loss": 0.1962,
+      "step": 16548
+    },
+    {
+      "epoch": 0.264784,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.411612903225807e-05,
+      "loss": 0.1225,
+      "step": 16549
+    },
+    {
+      "epoch": 0.2648,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.411451612903227e-05,
+      "loss": 0.1965,
+      "step": 16550
+    },
+    {
+      "epoch": 0.264816,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.411290322580646e-05,
+      "loss": 0.1703,
+      "step": 16551
+    },
+    {
+      "epoch": 0.264832,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.411129032258066e-05,
+      "loss": 0.1854,
+      "step": 16552
+    },
+    {
+      "epoch": 0.264848,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.410967741935484e-05,
+      "loss": 0.1625,
+      "step": 16553
+    },
+    {
+      "epoch": 0.264864,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.410806451612904e-05,
+      "loss": 0.1993,
+      "step": 16554
+    },
+    {
+      "epoch": 0.26488,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.410645161290323e-05,
+      "loss": 0.1452,
+      "step": 16555
+    },
+    {
+      "epoch": 0.264896,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.410483870967743e-05,
+      "loss": 0.1642,
+      "step": 16556
+    },
+    {
+      "epoch": 0.264912,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.410322580645161e-05,
+      "loss": 0.1589,
+      "step": 16557
+    },
+    {
+      "epoch": 0.264928,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.410161290322581e-05,
+      "loss": 0.1555,
+      "step": 16558
+    },
+    {
+      "epoch": 0.264944,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.41e-05,
+      "loss": 0.1447,
+      "step": 16559
+    },
+    {
+      "epoch": 0.26496,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.40983870967742e-05,
+      "loss": 0.1688,
+      "step": 16560
+    },
+    {
+      "epoch": 0.264976,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.409677419354838e-05,
+      "loss": 0.1674,
+      "step": 16561
+    },
+    {
+      "epoch": 0.264992,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.409516129032258e-05,
+      "loss": 0.1222,
+      "step": 16562
+    },
+    {
+      "epoch": 0.265008,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.409354838709678e-05,
+      "loss": 0.1838,
+      "step": 16563
+    },
+    {
+      "epoch": 0.265024,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.409193548387097e-05,
+      "loss": 0.1581,
+      "step": 16564
+    },
+    {
+      "epoch": 0.26504,
+      "grad_norm": 0.875,
+      "learning_rate": 7.409032258064517e-05,
+      "loss": 0.1863,
+      "step": 16565
+    },
+    {
+      "epoch": 0.265056,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.408870967741936e-05,
+      "loss": 0.1327,
+      "step": 16566
+    },
+    {
+      "epoch": 0.265072,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.408709677419356e-05,
+      "loss": 0.1629,
+      "step": 16567
+    },
+    {
+      "epoch": 0.265088,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.408548387096774e-05,
+      "loss": 0.1962,
+      "step": 16568
+    },
+    {
+      "epoch": 0.265104,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.408387096774194e-05,
+      "loss": 0.1466,
+      "step": 16569
+    },
+    {
+      "epoch": 0.26512,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.408225806451613e-05,
+      "loss": 0.1425,
+      "step": 16570
+    },
+    {
+      "epoch": 0.265136,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.408064516129033e-05,
+      "loss": 0.1678,
+      "step": 16571
+    },
+    {
+      "epoch": 0.265152,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.407903225806451e-05,
+      "loss": 0.1467,
+      "step": 16572
+    },
+    {
+      "epoch": 0.265168,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.407741935483871e-05,
+      "loss": 0.1539,
+      "step": 16573
+    },
+    {
+      "epoch": 0.265184,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.407580645161291e-05,
+      "loss": 0.1731,
+      "step": 16574
+    },
+    {
+      "epoch": 0.2652,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.407419354838711e-05,
+      "loss": 0.1391,
+      "step": 16575
+    },
+    {
+      "epoch": 0.265216,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.40725806451613e-05,
+      "loss": 0.1231,
+      "step": 16576
+    },
+    {
+      "epoch": 0.265232,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.407096774193548e-05,
+      "loss": 0.118,
+      "step": 16577
+    },
+    {
+      "epoch": 0.265248,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.406935483870968e-05,
+      "loss": 0.1905,
+      "step": 16578
+    },
+    {
+      "epoch": 0.265264,
+      "grad_norm": 1.453125,
+      "learning_rate": 7.406774193548387e-05,
+      "loss": 0.2201,
+      "step": 16579
+    },
+    {
+      "epoch": 0.26528,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.406612903225807e-05,
+      "loss": 0.1658,
+      "step": 16580
+    },
+    {
+      "epoch": 0.265296,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.406451612903225e-05,
+      "loss": 0.1453,
+      "step": 16581
+    },
+    {
+      "epoch": 0.265312,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.406290322580645e-05,
+      "loss": 0.1743,
+      "step": 16582
+    },
+    {
+      "epoch": 0.265328,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.406129032258064e-05,
+      "loss": 0.1714,
+      "step": 16583
+    },
+    {
+      "epoch": 0.265344,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.405967741935484e-05,
+      "loss": 0.1361,
+      "step": 16584
+    },
+    {
+      "epoch": 0.26536,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.405806451612904e-05,
+      "loss": 0.1609,
+      "step": 16585
+    },
+    {
+      "epoch": 0.265376,
+      "grad_norm": 0.75,
+      "learning_rate": 7.405645161290323e-05,
+      "loss": 0.1632,
+      "step": 16586
+    },
+    {
+      "epoch": 0.265392,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.405483870967743e-05,
+      "loss": 0.153,
+      "step": 16587
+    },
+    {
+      "epoch": 0.265408,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.405322580645163e-05,
+      "loss": 0.1571,
+      "step": 16588
+    },
+    {
+      "epoch": 0.265424,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.405161290322581e-05,
+      "loss": 0.2007,
+      "step": 16589
+    },
+    {
+      "epoch": 0.26544,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.405000000000001e-05,
+      "loss": 0.1424,
+      "step": 16590
+    },
+    {
+      "epoch": 0.265456,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.40483870967742e-05,
+      "loss": 0.1675,
+      "step": 16591
+    },
+    {
+      "epoch": 0.265472,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.40467741935484e-05,
+      "loss": 0.1114,
+      "step": 16592
+    },
+    {
+      "epoch": 0.265488,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.404516129032258e-05,
+      "loss": 0.2032,
+      "step": 16593
+    },
+    {
+      "epoch": 0.265504,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.404354838709677e-05,
+      "loss": 0.1239,
+      "step": 16594
+    },
+    {
+      "epoch": 0.26552,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.404193548387097e-05,
+      "loss": 0.176,
+      "step": 16595
+    },
+    {
+      "epoch": 0.265536,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.404032258064515e-05,
+      "loss": 0.1717,
+      "step": 16596
+    },
+    {
+      "epoch": 0.265552,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.403870967741935e-05,
+      "loss": 0.1761,
+      "step": 16597
+    },
+    {
+      "epoch": 0.265568,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.403709677419355e-05,
+      "loss": 0.192,
+      "step": 16598
+    },
+    {
+      "epoch": 0.265584,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.403548387096775e-05,
+      "loss": 0.1785,
+      "step": 16599
+    },
+    {
+      "epoch": 0.2656,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.403387096774194e-05,
+      "loss": 0.1445,
+      "step": 16600
+    },
+    {
+      "epoch": 0.265616,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.403225806451614e-05,
+      "loss": 0.1622,
+      "step": 16601
+    },
+    {
+      "epoch": 0.265632,
+      "grad_norm": 1.2734375,
+      "learning_rate": 7.403064516129033e-05,
+      "loss": 0.1893,
+      "step": 16602
+    },
+    {
+      "epoch": 0.265648,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.402903225806452e-05,
+      "loss": 0.2039,
+      "step": 16603
+    },
+    {
+      "epoch": 0.265664,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.402741935483871e-05,
+      "loss": 0.2261,
+      "step": 16604
+    },
+    {
+      "epoch": 0.26568,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.402580645161291e-05,
+      "loss": 0.1602,
+      "step": 16605
+    },
+    {
+      "epoch": 0.265696,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.40241935483871e-05,
+      "loss": 0.1833,
+      "step": 16606
+    },
+    {
+      "epoch": 0.265712,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.40225806451613e-05,
+      "loss": 0.1718,
+      "step": 16607
+    },
+    {
+      "epoch": 0.265728,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.402096774193548e-05,
+      "loss": 0.1265,
+      "step": 16608
+    },
+    {
+      "epoch": 0.265744,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.401935483870968e-05,
+      "loss": 0.1577,
+      "step": 16609
+    },
+    {
+      "epoch": 0.26576,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.401774193548388e-05,
+      "loss": 0.1772,
+      "step": 16610
+    },
+    {
+      "epoch": 0.265776,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.401612903225807e-05,
+      "loss": 0.1959,
+      "step": 16611
+    },
+    {
+      "epoch": 0.265792,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.401451612903227e-05,
+      "loss": 0.2017,
+      "step": 16612
+    },
+    {
+      "epoch": 0.265808,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.401290322580645e-05,
+      "loss": 0.1916,
+      "step": 16613
+    },
+    {
+      "epoch": 0.265824,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.401129032258065e-05,
+      "loss": 0.1597,
+      "step": 16614
+    },
+    {
+      "epoch": 0.26584,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.400967741935484e-05,
+      "loss": 0.1556,
+      "step": 16615
+    },
+    {
+      "epoch": 0.265856,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.400806451612904e-05,
+      "loss": 0.1543,
+      "step": 16616
+    },
+    {
+      "epoch": 0.265872,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.400645161290322e-05,
+      "loss": 0.1519,
+      "step": 16617
+    },
+    {
+      "epoch": 0.265888,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.400483870967742e-05,
+      "loss": 0.184,
+      "step": 16618
+    },
+    {
+      "epoch": 0.265904,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.400322580645161e-05,
+      "loss": 0.2072,
+      "step": 16619
+    },
+    {
+      "epoch": 0.26592,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.400161290322581e-05,
+      "loss": 0.1962,
+      "step": 16620
+    },
+    {
+      "epoch": 0.265936,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.4e-05,
+      "loss": 0.1191,
+      "step": 16621
+    },
+    {
+      "epoch": 0.265952,
+      "grad_norm": 1.296875,
+      "learning_rate": 7.39983870967742e-05,
+      "loss": 0.1672,
+      "step": 16622
+    },
+    {
+      "epoch": 0.265968,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.39967741935484e-05,
+      "loss": 0.1458,
+      "step": 16623
+    },
+    {
+      "epoch": 0.265984,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.399516129032258e-05,
+      "loss": 0.1704,
+      "step": 16624
+    },
+    {
+      "epoch": 0.266,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.399354838709678e-05,
+      "loss": 0.1558,
+      "step": 16625
+    },
+    {
+      "epoch": 0.266016,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.399193548387097e-05,
+      "loss": 0.2171,
+      "step": 16626
+    },
+    {
+      "epoch": 0.266032,
+      "grad_norm": 1.3359375,
+      "learning_rate": 7.399032258064517e-05,
+      "loss": 0.1746,
+      "step": 16627
+    },
+    {
+      "epoch": 0.266048,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.398870967741935e-05,
+      "loss": 0.1725,
+      "step": 16628
+    },
+    {
+      "epoch": 0.266064,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.398709677419355e-05,
+      "loss": 0.2159,
+      "step": 16629
+    },
+    {
+      "epoch": 0.26608,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.398548387096774e-05,
+      "loss": 0.1569,
+      "step": 16630
+    },
+    {
+      "epoch": 0.266096,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.398387096774194e-05,
+      "loss": 0.1291,
+      "step": 16631
+    },
+    {
+      "epoch": 0.266112,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.398225806451612e-05,
+      "loss": 0.1646,
+      "step": 16632
+    },
+    {
+      "epoch": 0.266128,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.398064516129032e-05,
+      "loss": 0.1698,
+      "step": 16633
+    },
+    {
+      "epoch": 0.266144,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.397903225806452e-05,
+      "loss": 0.1669,
+      "step": 16634
+    },
+    {
+      "epoch": 0.26616,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.397741935483872e-05,
+      "loss": 0.1342,
+      "step": 16635
+    },
+    {
+      "epoch": 0.266176,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.397580645161291e-05,
+      "loss": 0.1529,
+      "step": 16636
+    },
+    {
+      "epoch": 0.266192,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.397419354838711e-05,
+      "loss": 0.1793,
+      "step": 16637
+    },
+    {
+      "epoch": 0.266208,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.39725806451613e-05,
+      "loss": 0.1675,
+      "step": 16638
+    },
+    {
+      "epoch": 0.266224,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.39709677419355e-05,
+      "loss": 0.1644,
+      "step": 16639
+    },
+    {
+      "epoch": 0.26624,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.396935483870968e-05,
+      "loss": 0.1979,
+      "step": 16640
+    },
+    {
+      "epoch": 0.266256,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.396774193548387e-05,
+      "loss": 0.2003,
+      "step": 16641
+    },
+    {
+      "epoch": 0.266272,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.396612903225807e-05,
+      "loss": 0.2085,
+      "step": 16642
+    },
+    {
+      "epoch": 0.266288,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.396451612903225e-05,
+      "loss": 0.1729,
+      "step": 16643
+    },
+    {
+      "epoch": 0.266304,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.396290322580645e-05,
+      "loss": 0.1752,
+      "step": 16644
+    },
+    {
+      "epoch": 0.26632,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.396129032258065e-05,
+      "loss": 0.1648,
+      "step": 16645
+    },
+    {
+      "epoch": 0.266336,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.395967741935485e-05,
+      "loss": 0.171,
+      "step": 16646
+    },
+    {
+      "epoch": 0.266352,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.395806451612904e-05,
+      "loss": 0.1636,
+      "step": 16647
+    },
+    {
+      "epoch": 0.266368,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.395645161290324e-05,
+      "loss": 0.1464,
+      "step": 16648
+    },
+    {
+      "epoch": 0.266384,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.395483870967742e-05,
+      "loss": 0.1645,
+      "step": 16649
+    },
+    {
+      "epoch": 0.2664,
+      "grad_norm": 1.2109375,
+      "learning_rate": 7.395322580645162e-05,
+      "loss": 0.1984,
+      "step": 16650
+    },
+    {
+      "epoch": 0.266416,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.395161290322581e-05,
+      "loss": 0.1642,
+      "step": 16651
+    },
+    {
+      "epoch": 0.266432,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.395000000000001e-05,
+      "loss": 0.1526,
+      "step": 16652
+    },
+    {
+      "epoch": 0.266448,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.39483870967742e-05,
+      "loss": 0.167,
+      "step": 16653
+    },
+    {
+      "epoch": 0.266464,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.39467741935484e-05,
+      "loss": 0.1629,
+      "step": 16654
+    },
+    {
+      "epoch": 0.26648,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.394516129032258e-05,
+      "loss": 0.1452,
+      "step": 16655
+    },
+    {
+      "epoch": 0.266496,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.394354838709677e-05,
+      "loss": 0.1552,
+      "step": 16656
+    },
+    {
+      "epoch": 0.266512,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.394193548387097e-05,
+      "loss": 0.1823,
+      "step": 16657
+    },
+    {
+      "epoch": 0.266528,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.394032258064517e-05,
+      "loss": 0.1735,
+      "step": 16658
+    },
+    {
+      "epoch": 0.266544,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.393870967741937e-05,
+      "loss": 0.1715,
+      "step": 16659
+    },
+    {
+      "epoch": 0.26656,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.393709677419355e-05,
+      "loss": 0.1664,
+      "step": 16660
+    },
+    {
+      "epoch": 0.266576,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.393548387096775e-05,
+      "loss": 0.1428,
+      "step": 16661
+    },
+    {
+      "epoch": 0.266592,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.393387096774194e-05,
+      "loss": 0.1642,
+      "step": 16662
+    },
+    {
+      "epoch": 0.266608,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.393225806451614e-05,
+      "loss": 0.1644,
+      "step": 16663
+    },
+    {
+      "epoch": 0.266624,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.393064516129032e-05,
+      "loss": 0.1746,
+      "step": 16664
+    },
+    {
+      "epoch": 0.26664,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.392903225806452e-05,
+      "loss": 0.1375,
+      "step": 16665
+    },
+    {
+      "epoch": 0.266656,
+      "grad_norm": 1.4921875,
+      "learning_rate": 7.392741935483871e-05,
+      "loss": 0.1756,
+      "step": 16666
+    },
+    {
+      "epoch": 0.266672,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.392580645161291e-05,
+      "loss": 0.156,
+      "step": 16667
+    },
+    {
+      "epoch": 0.266688,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.39241935483871e-05,
+      "loss": 0.1614,
+      "step": 16668
+    },
+    {
+      "epoch": 0.266704,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.39225806451613e-05,
+      "loss": 0.2095,
+      "step": 16669
+    },
+    {
+      "epoch": 0.26672,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.39209677419355e-05,
+      "loss": 0.1851,
+      "step": 16670
+    },
+    {
+      "epoch": 0.266736,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.391935483870968e-05,
+      "loss": 0.2052,
+      "step": 16671
+    },
+    {
+      "epoch": 0.266752,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.391774193548388e-05,
+      "loss": 0.1958,
+      "step": 16672
+    },
+    {
+      "epoch": 0.266768,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.391612903225807e-05,
+      "loss": 0.1718,
+      "step": 16673
+    },
+    {
+      "epoch": 0.266784,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.391451612903226e-05,
+      "loss": 0.1743,
+      "step": 16674
+    },
+    {
+      "epoch": 0.2668,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.391290322580645e-05,
+      "loss": 0.153,
+      "step": 16675
+    },
+    {
+      "epoch": 0.266816,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.391129032258065e-05,
+      "loss": 0.1814,
+      "step": 16676
+    },
+    {
+      "epoch": 0.266832,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.390967741935484e-05,
+      "loss": 0.143,
+      "step": 16677
+    },
+    {
+      "epoch": 0.266848,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.390806451612904e-05,
+      "loss": 0.1725,
+      "step": 16678
+    },
+    {
+      "epoch": 0.266864,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.390645161290322e-05,
+      "loss": 0.1528,
+      "step": 16679
+    },
+    {
+      "epoch": 0.26688,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.390483870967742e-05,
+      "loss": 0.1632,
+      "step": 16680
+    },
+    {
+      "epoch": 0.266896,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.390322580645161e-05,
+      "loss": 0.1527,
+      "step": 16681
+    },
+    {
+      "epoch": 0.266912,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.390161290322581e-05,
+      "loss": 0.1583,
+      "step": 16682
+    },
+    {
+      "epoch": 0.266928,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.390000000000001e-05,
+      "loss": 0.2007,
+      "step": 16683
+    },
+    {
+      "epoch": 0.266944,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.389838709677421e-05,
+      "loss": 0.2112,
+      "step": 16684
+    },
+    {
+      "epoch": 0.26696,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.389677419354839e-05,
+      "loss": 0.1373,
+      "step": 16685
+    },
+    {
+      "epoch": 0.266976,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.389516129032258e-05,
+      "loss": 0.1685,
+      "step": 16686
+    },
+    {
+      "epoch": 0.266992,
+      "grad_norm": 1.3203125,
+      "learning_rate": 7.389354838709678e-05,
+      "loss": 0.1908,
+      "step": 16687
+    },
+    {
+      "epoch": 0.267008,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.389193548387096e-05,
+      "loss": 0.1985,
+      "step": 16688
+    },
+    {
+      "epoch": 0.267024,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.389032258064516e-05,
+      "loss": 0.1921,
+      "step": 16689
+    },
+    {
+      "epoch": 0.26704,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.388870967741935e-05,
+      "loss": 0.1317,
+      "step": 16690
+    },
+    {
+      "epoch": 0.267056,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.388709677419355e-05,
+      "loss": 0.1672,
+      "step": 16691
+    },
+    {
+      "epoch": 0.267072,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.388548387096774e-05,
+      "loss": 0.1763,
+      "step": 16692
+    },
+    {
+      "epoch": 0.267088,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.388387096774194e-05,
+      "loss": 0.1556,
+      "step": 16693
+    },
+    {
+      "epoch": 0.267104,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.388225806451614e-05,
+      "loss": 0.1478,
+      "step": 16694
+    },
+    {
+      "epoch": 0.26712,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.388064516129034e-05,
+      "loss": 0.1352,
+      "step": 16695
+    },
+    {
+      "epoch": 0.267136,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.387903225806452e-05,
+      "loss": 0.165,
+      "step": 16696
+    },
+    {
+      "epoch": 0.267152,
+      "grad_norm": 1.8125,
+      "learning_rate": 7.387741935483872e-05,
+      "loss": 0.1561,
+      "step": 16697
+    },
+    {
+      "epoch": 0.267168,
+      "grad_norm": 1.453125,
+      "learning_rate": 7.387580645161291e-05,
+      "loss": 0.1721,
+      "step": 16698
+    },
+    {
+      "epoch": 0.267184,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.38741935483871e-05,
+      "loss": 0.1607,
+      "step": 16699
+    },
+    {
+      "epoch": 0.2672,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.387258064516129e-05,
+      "loss": 0.1434,
+      "step": 16700
+    },
+    {
+      "epoch": 0.267216,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.387096774193549e-05,
+      "loss": 0.1559,
+      "step": 16701
+    },
+    {
+      "epoch": 0.267232,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.386935483870968e-05,
+      "loss": 0.1713,
+      "step": 16702
+    },
+    {
+      "epoch": 0.267248,
+      "grad_norm": 0.75,
+      "learning_rate": 7.386774193548386e-05,
+      "loss": 0.1412,
+      "step": 16703
+    },
+    {
+      "epoch": 0.267264,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.386612903225806e-05,
+      "loss": 0.1829,
+      "step": 16704
+    },
+    {
+      "epoch": 0.26728,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.386451612903226e-05,
+      "loss": 0.2286,
+      "step": 16705
+    },
+    {
+      "epoch": 0.267296,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.386290322580646e-05,
+      "loss": 0.1283,
+      "step": 16706
+    },
+    {
+      "epoch": 0.267312,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.386129032258065e-05,
+      "loss": 0.166,
+      "step": 16707
+    },
+    {
+      "epoch": 0.267328,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.385967741935485e-05,
+      "loss": 0.1711,
+      "step": 16708
+    },
+    {
+      "epoch": 0.267344,
+      "grad_norm": 1.40625,
+      "learning_rate": 7.385806451612904e-05,
+      "loss": 0.1653,
+      "step": 16709
+    },
+    {
+      "epoch": 0.26736,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.385645161290323e-05,
+      "loss": 0.1841,
+      "step": 16710
+    },
+    {
+      "epoch": 0.267376,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.385483870967742e-05,
+      "loss": 0.2216,
+      "step": 16711
+    },
+    {
+      "epoch": 0.267392,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.385322580645162e-05,
+      "loss": 0.1824,
+      "step": 16712
+    },
+    {
+      "epoch": 0.267408,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.38516129032258e-05,
+      "loss": 0.1783,
+      "step": 16713
+    },
+    {
+      "epoch": 0.267424,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.385e-05,
+      "loss": 0.1668,
+      "step": 16714
+    },
+    {
+      "epoch": 0.26744,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.384838709677419e-05,
+      "loss": 0.1691,
+      "step": 16715
+    },
+    {
+      "epoch": 0.267456,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.384677419354839e-05,
+      "loss": 0.1645,
+      "step": 16716
+    },
+    {
+      "epoch": 0.267472,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.384516129032258e-05,
+      "loss": 0.1695,
+      "step": 16717
+    },
+    {
+      "epoch": 0.267488,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.384354838709678e-05,
+      "loss": 0.1814,
+      "step": 16718
+    },
+    {
+      "epoch": 0.267504,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.384193548387098e-05,
+      "loss": 0.1763,
+      "step": 16719
+    },
+    {
+      "epoch": 0.26752,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.384032258064516e-05,
+      "loss": 0.1525,
+      "step": 16720
+    },
+    {
+      "epoch": 0.267536,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.383870967741936e-05,
+      "loss": 0.2387,
+      "step": 16721
+    },
+    {
+      "epoch": 0.267552,
+      "grad_norm": 1.4609375,
+      "learning_rate": 7.383709677419355e-05,
+      "loss": 0.1653,
+      "step": 16722
+    },
+    {
+      "epoch": 0.267568,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.383548387096775e-05,
+      "loss": 0.1587,
+      "step": 16723
+    },
+    {
+      "epoch": 0.267584,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.383387096774193e-05,
+      "loss": 0.1818,
+      "step": 16724
+    },
+    {
+      "epoch": 0.2676,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.383225806451613e-05,
+      "loss": 0.1544,
+      "step": 16725
+    },
+    {
+      "epoch": 0.267616,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.383064516129032e-05,
+      "loss": 0.2225,
+      "step": 16726
+    },
+    {
+      "epoch": 0.267632,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.382903225806452e-05,
+      "loss": 0.1776,
+      "step": 16727
+    },
+    {
+      "epoch": 0.267648,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.38274193548387e-05,
+      "loss": 0.17,
+      "step": 16728
+    },
+    {
+      "epoch": 0.267664,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.38258064516129e-05,
+      "loss": 0.2132,
+      "step": 16729
+    },
+    {
+      "epoch": 0.26768,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.38241935483871e-05,
+      "loss": 0.2105,
+      "step": 16730
+    },
+    {
+      "epoch": 0.267696,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.38225806451613e-05,
+      "loss": 0.2179,
+      "step": 16731
+    },
+    {
+      "epoch": 0.267712,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.382096774193549e-05,
+      "loss": 0.1312,
+      "step": 16732
+    },
+    {
+      "epoch": 0.267728,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.381935483870968e-05,
+      "loss": 0.1564,
+      "step": 16733
+    },
+    {
+      "epoch": 0.267744,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.381774193548388e-05,
+      "loss": 0.1687,
+      "step": 16734
+    },
+    {
+      "epoch": 0.26776,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.381612903225806e-05,
+      "loss": 0.1503,
+      "step": 16735
+    },
+    {
+      "epoch": 0.267776,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.381451612903226e-05,
+      "loss": 0.1913,
+      "step": 16736
+    },
+    {
+      "epoch": 0.267792,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.381290322580645e-05,
+      "loss": 0.1362,
+      "step": 16737
+    },
+    {
+      "epoch": 0.267808,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.381129032258065e-05,
+      "loss": 0.1427,
+      "step": 16738
+    },
+    {
+      "epoch": 0.267824,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.380967741935483e-05,
+      "loss": 0.1505,
+      "step": 16739
+    },
+    {
+      "epoch": 0.26784,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.380806451612903e-05,
+      "loss": 0.1757,
+      "step": 16740
+    },
+    {
+      "epoch": 0.267856,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.380645161290323e-05,
+      "loss": 0.1616,
+      "step": 16741
+    },
+    {
+      "epoch": 0.267872,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.380483870967742e-05,
+      "loss": 0.1511,
+      "step": 16742
+    },
+    {
+      "epoch": 0.267888,
+      "grad_norm": 1.46875,
+      "learning_rate": 7.380322580645162e-05,
+      "loss": 0.2475,
+      "step": 16743
+    },
+    {
+      "epoch": 0.267904,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.380161290322582e-05,
+      "loss": 0.1411,
+      "step": 16744
+    },
+    {
+      "epoch": 0.26792,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.38e-05,
+      "loss": 0.1313,
+      "step": 16745
+    },
+    {
+      "epoch": 0.267936,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.37983870967742e-05,
+      "loss": 0.1587,
+      "step": 16746
+    },
+    {
+      "epoch": 0.267952,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.379677419354839e-05,
+      "loss": 0.1599,
+      "step": 16747
+    },
+    {
+      "epoch": 0.267968,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.379516129032258e-05,
+      "loss": 0.187,
+      "step": 16748
+    },
+    {
+      "epoch": 0.267984,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.379354838709678e-05,
+      "loss": 0.163,
+      "step": 16749
+    },
+    {
+      "epoch": 0.268,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.379193548387096e-05,
+      "loss": 0.2081,
+      "step": 16750
+    },
+    {
+      "epoch": 0.268016,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.379032258064516e-05,
+      "loss": 0.1988,
+      "step": 16751
+    },
+    {
+      "epoch": 0.268032,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.378870967741935e-05,
+      "loss": 0.163,
+      "step": 16752
+    },
+    {
+      "epoch": 0.268048,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.378709677419355e-05,
+      "loss": 0.1307,
+      "step": 16753
+    },
+    {
+      "epoch": 0.268064,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.378548387096775e-05,
+      "loss": 0.211,
+      "step": 16754
+    },
+    {
+      "epoch": 0.26808,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.378387096774195e-05,
+      "loss": 0.2042,
+      "step": 16755
+    },
+    {
+      "epoch": 0.268096,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.378225806451613e-05,
+      "loss": 0.1686,
+      "step": 16756
+    },
+    {
+      "epoch": 0.268112,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.378064516129033e-05,
+      "loss": 0.1737,
+      "step": 16757
+    },
+    {
+      "epoch": 0.268128,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.377903225806452e-05,
+      "loss": 0.1575,
+      "step": 16758
+    },
+    {
+      "epoch": 0.268144,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.377741935483872e-05,
+      "loss": 0.193,
+      "step": 16759
+    },
+    {
+      "epoch": 0.26816,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.37758064516129e-05,
+      "loss": 0.1613,
+      "step": 16760
+    },
+    {
+      "epoch": 0.268176,
+      "grad_norm": 1.28125,
+      "learning_rate": 7.37741935483871e-05,
+      "loss": 0.1784,
+      "step": 16761
+    },
+    {
+      "epoch": 0.268192,
+      "grad_norm": 0.625,
+      "learning_rate": 7.377258064516129e-05,
+      "loss": 0.1558,
+      "step": 16762
+    },
+    {
+      "epoch": 0.268208,
+      "grad_norm": 0.625,
+      "learning_rate": 7.377096774193549e-05,
+      "loss": 0.1629,
+      "step": 16763
+    },
+    {
+      "epoch": 0.268224,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.376935483870968e-05,
+      "loss": 0.1682,
+      "step": 16764
+    },
+    {
+      "epoch": 0.26824,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.376774193548388e-05,
+      "loss": 0.1808,
+      "step": 16765
+    },
+    {
+      "epoch": 0.268256,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.376612903225808e-05,
+      "loss": 0.1449,
+      "step": 16766
+    },
+    {
+      "epoch": 0.268272,
+      "grad_norm": 1.28125,
+      "learning_rate": 7.376451612903226e-05,
+      "loss": 0.1854,
+      "step": 16767
+    },
+    {
+      "epoch": 0.268288,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.376290322580646e-05,
+      "loss": 0.1544,
+      "step": 16768
+    },
+    {
+      "epoch": 0.268304,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.376129032258065e-05,
+      "loss": 0.1744,
+      "step": 16769
+    },
+    {
+      "epoch": 0.26832,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.375967741935485e-05,
+      "loss": 0.1904,
+      "step": 16770
+    },
+    {
+      "epoch": 0.268336,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.375806451612903e-05,
+      "loss": 0.2022,
+      "step": 16771
+    },
+    {
+      "epoch": 0.268352,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.375645161290323e-05,
+      "loss": 0.1301,
+      "step": 16772
+    },
+    {
+      "epoch": 0.268368,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.375483870967742e-05,
+      "loss": 0.1583,
+      "step": 16773
+    },
+    {
+      "epoch": 0.268384,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.375322580645162e-05,
+      "loss": 0.148,
+      "step": 16774
+    },
+    {
+      "epoch": 0.2684,
+      "grad_norm": 0.625,
+      "learning_rate": 7.37516129032258e-05,
+      "loss": 0.1312,
+      "step": 16775
+    },
+    {
+      "epoch": 0.268416,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.375e-05,
+      "loss": 0.2286,
+      "step": 16776
+    },
+    {
+      "epoch": 0.268432,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.374838709677419e-05,
+      "loss": 0.1418,
+      "step": 16777
+    },
+    {
+      "epoch": 0.268448,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.374677419354839e-05,
+      "loss": 0.1816,
+      "step": 16778
+    },
+    {
+      "epoch": 0.268464,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.374516129032259e-05,
+      "loss": 0.1935,
+      "step": 16779
+    },
+    {
+      "epoch": 0.26848,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.374354838709678e-05,
+      "loss": 0.1534,
+      "step": 16780
+    },
+    {
+      "epoch": 0.268496,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.374193548387097e-05,
+      "loss": 0.1627,
+      "step": 16781
+    },
+    {
+      "epoch": 0.268512,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.374032258064516e-05,
+      "loss": 0.1597,
+      "step": 16782
+    },
+    {
+      "epoch": 0.268528,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.373870967741936e-05,
+      "loss": 0.1607,
+      "step": 16783
+    },
+    {
+      "epoch": 0.268544,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.373709677419355e-05,
+      "loss": 0.1764,
+      "step": 16784
+    },
+    {
+      "epoch": 0.26856,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.373548387096775e-05,
+      "loss": 0.22,
+      "step": 16785
+    },
+    {
+      "epoch": 0.268576,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.373387096774193e-05,
+      "loss": 0.133,
+      "step": 16786
+    },
+    {
+      "epoch": 0.268592,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.373225806451613e-05,
+      "loss": 0.1659,
+      "step": 16787
+    },
+    {
+      "epoch": 0.268608,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.373064516129032e-05,
+      "loss": 0.1742,
+      "step": 16788
+    },
+    {
+      "epoch": 0.268624,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.372903225806452e-05,
+      "loss": 0.1583,
+      "step": 16789
+    },
+    {
+      "epoch": 0.26864,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.372741935483872e-05,
+      "loss": 0.1736,
+      "step": 16790
+    },
+    {
+      "epoch": 0.268656,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.372580645161292e-05,
+      "loss": 0.1872,
+      "step": 16791
+    },
+    {
+      "epoch": 0.268672,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.37241935483871e-05,
+      "loss": 0.157,
+      "step": 16792
+    },
+    {
+      "epoch": 0.268688,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.37225806451613e-05,
+      "loss": 0.1815,
+      "step": 16793
+    },
+    {
+      "epoch": 0.268704,
+      "grad_norm": 1.484375,
+      "learning_rate": 7.372096774193549e-05,
+      "loss": 0.1627,
+      "step": 16794
+    },
+    {
+      "epoch": 0.26872,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.371935483870967e-05,
+      "loss": 0.2008,
+      "step": 16795
+    },
+    {
+      "epoch": 0.268736,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.371774193548387e-05,
+      "loss": 0.194,
+      "step": 16796
+    },
+    {
+      "epoch": 0.268752,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.371612903225806e-05,
+      "loss": 0.1708,
+      "step": 16797
+    },
+    {
+      "epoch": 0.268768,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.371451612903226e-05,
+      "loss": 0.1476,
+      "step": 16798
+    },
+    {
+      "epoch": 0.268784,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.371290322580645e-05,
+      "loss": 0.1563,
+      "step": 16799
+    },
+    {
+      "epoch": 0.2688,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.371129032258065e-05,
+      "loss": 0.1921,
+      "step": 16800
+    },
+    {
+      "epoch": 0.268816,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.370967741935485e-05,
+      "loss": 0.1832,
+      "step": 16801
+    },
+    {
+      "epoch": 0.268832,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.370806451612904e-05,
+      "loss": 0.1465,
+      "step": 16802
+    },
+    {
+      "epoch": 0.268848,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.370645161290323e-05,
+      "loss": 0.1609,
+      "step": 16803
+    },
+    {
+      "epoch": 0.268864,
+      "grad_norm": 0.875,
+      "learning_rate": 7.370483870967743e-05,
+      "loss": 0.1545,
+      "step": 16804
+    },
+    {
+      "epoch": 0.26888,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.370322580645162e-05,
+      "loss": 0.217,
+      "step": 16805
+    },
+    {
+      "epoch": 0.268896,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.370161290322582e-05,
+      "loss": 0.1628,
+      "step": 16806
+    },
+    {
+      "epoch": 0.268912,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.37e-05,
+      "loss": 0.1225,
+      "step": 16807
+    },
+    {
+      "epoch": 0.268928,
+      "grad_norm": 1.25,
+      "learning_rate": 7.36983870967742e-05,
+      "loss": 0.1702,
+      "step": 16808
+    },
+    {
+      "epoch": 0.268944,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.369677419354839e-05,
+      "loss": 0.1681,
+      "step": 16809
+    },
+    {
+      "epoch": 0.26896,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.369516129032259e-05,
+      "loss": 0.165,
+      "step": 16810
+    },
+    {
+      "epoch": 0.268976,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.369354838709677e-05,
+      "loss": 0.1554,
+      "step": 16811
+    },
+    {
+      "epoch": 0.268992,
+      "grad_norm": 0.474609375,
+      "learning_rate": 7.369193548387096e-05,
+      "loss": 0.1533,
+      "step": 16812
+    },
+    {
+      "epoch": 0.269008,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.369032258064516e-05,
+      "loss": 0.1678,
+      "step": 16813
+    },
+    {
+      "epoch": 0.269024,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.368870967741936e-05,
+      "loss": 0.2108,
+      "step": 16814
+    },
+    {
+      "epoch": 0.26904,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.368709677419356e-05,
+      "loss": 0.1233,
+      "step": 16815
+    },
+    {
+      "epoch": 0.269056,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.368548387096774e-05,
+      "loss": 0.1609,
+      "step": 16816
+    },
+    {
+      "epoch": 0.269072,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.368387096774194e-05,
+      "loss": 0.1887,
+      "step": 16817
+    },
+    {
+      "epoch": 0.269088,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.368225806451613e-05,
+      "loss": 0.2022,
+      "step": 16818
+    },
+    {
+      "epoch": 0.269104,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.368064516129033e-05,
+      "loss": 0.1808,
+      "step": 16819
+    },
+    {
+      "epoch": 0.26912,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.367903225806452e-05,
+      "loss": 0.1927,
+      "step": 16820
+    },
+    {
+      "epoch": 0.269136,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.367741935483872e-05,
+      "loss": 0.1484,
+      "step": 16821
+    },
+    {
+      "epoch": 0.269152,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.36758064516129e-05,
+      "loss": 0.16,
+      "step": 16822
+    },
+    {
+      "epoch": 0.269168,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.36741935483871e-05,
+      "loss": 0.1606,
+      "step": 16823
+    },
+    {
+      "epoch": 0.269184,
+      "grad_norm": 1.3125,
+      "learning_rate": 7.367258064516129e-05,
+      "loss": 0.1489,
+      "step": 16824
+    },
+    {
+      "epoch": 0.2692,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.367096774193549e-05,
+      "loss": 0.1631,
+      "step": 16825
+    },
+    {
+      "epoch": 0.269216,
+      "grad_norm": 1.34375,
+      "learning_rate": 7.366935483870969e-05,
+      "loss": 0.1632,
+      "step": 16826
+    },
+    {
+      "epoch": 0.269232,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.366774193548387e-05,
+      "loss": 0.1713,
+      "step": 16827
+    },
+    {
+      "epoch": 0.269248,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.366612903225807e-05,
+      "loss": 0.1581,
+      "step": 16828
+    },
+    {
+      "epoch": 0.269264,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.366451612903226e-05,
+      "loss": 0.1966,
+      "step": 16829
+    },
+    {
+      "epoch": 0.26928,
+      "grad_norm": 0.875,
+      "learning_rate": 7.366290322580646e-05,
+      "loss": 0.1712,
+      "step": 16830
+    },
+    {
+      "epoch": 0.269296,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.366129032258064e-05,
+      "loss": 0.1708,
+      "step": 16831
+    },
+    {
+      "epoch": 0.269312,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.365967741935484e-05,
+      "loss": 0.1822,
+      "step": 16832
+    },
+    {
+      "epoch": 0.269328,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.365806451612903e-05,
+      "loss": 0.1871,
+      "step": 16833
+    },
+    {
+      "epoch": 0.269344,
+      "grad_norm": 1.34375,
+      "learning_rate": 7.365645161290323e-05,
+      "loss": 0.1854,
+      "step": 16834
+    },
+    {
+      "epoch": 0.26936,
+      "grad_norm": 0.75,
+      "learning_rate": 7.365483870967742e-05,
+      "loss": 0.1356,
+      "step": 16835
+    },
+    {
+      "epoch": 0.269376,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.365322580645162e-05,
+      "loss": 0.1544,
+      "step": 16836
+    },
+    {
+      "epoch": 0.269392,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.365161290322582e-05,
+      "loss": 0.1738,
+      "step": 16837
+    },
+    {
+      "epoch": 0.269408,
+      "grad_norm": 0.75,
+      "learning_rate": 7.365e-05,
+      "loss": 0.1966,
+      "step": 16838
+    },
+    {
+      "epoch": 0.269424,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.36483870967742e-05,
+      "loss": 0.1869,
+      "step": 16839
+    },
+    {
+      "epoch": 0.26944,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.36467741935484e-05,
+      "loss": 0.1883,
+      "step": 16840
+    },
+    {
+      "epoch": 0.269456,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.364516129032259e-05,
+      "loss": 0.1303,
+      "step": 16841
+    },
+    {
+      "epoch": 0.269472,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.364354838709677e-05,
+      "loss": 0.194,
+      "step": 16842
+    },
+    {
+      "epoch": 0.269488,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.364193548387097e-05,
+      "loss": 0.1861,
+      "step": 16843
+    },
+    {
+      "epoch": 0.269504,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.364032258064516e-05,
+      "loss": 0.1694,
+      "step": 16844
+    },
+    {
+      "epoch": 0.26952,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.363870967741936e-05,
+      "loss": 0.1624,
+      "step": 16845
+    },
+    {
+      "epoch": 0.269536,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.363709677419354e-05,
+      "loss": 0.1784,
+      "step": 16846
+    },
+    {
+      "epoch": 0.269552,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.363548387096774e-05,
+      "loss": 0.1231,
+      "step": 16847
+    },
+    {
+      "epoch": 0.269568,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.363387096774193e-05,
+      "loss": 0.1393,
+      "step": 16848
+    },
+    {
+      "epoch": 0.269584,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.363225806451613e-05,
+      "loss": 0.2085,
+      "step": 16849
+    },
+    {
+      "epoch": 0.2696,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.363064516129033e-05,
+      "loss": 0.1703,
+      "step": 16850
+    },
+    {
+      "epoch": 0.269616,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.362903225806453e-05,
+      "loss": 0.1744,
+      "step": 16851
+    },
+    {
+      "epoch": 0.269632,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.362741935483871e-05,
+      "loss": 0.2112,
+      "step": 16852
+    },
+    {
+      "epoch": 0.269648,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.362580645161291e-05,
+      "loss": 0.1762,
+      "step": 16853
+    },
+    {
+      "epoch": 0.269664,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.36241935483871e-05,
+      "loss": 0.1659,
+      "step": 16854
+    },
+    {
+      "epoch": 0.26968,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.36225806451613e-05,
+      "loss": 0.152,
+      "step": 16855
+    },
+    {
+      "epoch": 0.269696,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.362096774193549e-05,
+      "loss": 0.1735,
+      "step": 16856
+    },
+    {
+      "epoch": 0.269712,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.361935483870967e-05,
+      "loss": 0.1718,
+      "step": 16857
+    },
+    {
+      "epoch": 0.269728,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.361774193548387e-05,
+      "loss": 0.2436,
+      "step": 16858
+    },
+    {
+      "epoch": 0.269744,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.361612903225806e-05,
+      "loss": 0.165,
+      "step": 16859
+    },
+    {
+      "epoch": 0.26976,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.361451612903226e-05,
+      "loss": 0.1909,
+      "step": 16860
+    },
+    {
+      "epoch": 0.269776,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.361290322580646e-05,
+      "loss": 0.1849,
+      "step": 16861
+    },
+    {
+      "epoch": 0.269792,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.361129032258066e-05,
+      "loss": 0.198,
+      "step": 16862
+    },
+    {
+      "epoch": 0.269808,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.360967741935484e-05,
+      "loss": 0.1576,
+      "step": 16863
+    },
+    {
+      "epoch": 0.269824,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.360806451612904e-05,
+      "loss": 0.1516,
+      "step": 16864
+    },
+    {
+      "epoch": 0.26984,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.360645161290323e-05,
+      "loss": 0.2225,
+      "step": 16865
+    },
+    {
+      "epoch": 0.269856,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.360483870967743e-05,
+      "loss": 0.1431,
+      "step": 16866
+    },
+    {
+      "epoch": 0.269872,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.360322580645161e-05,
+      "loss": 0.1861,
+      "step": 16867
+    },
+    {
+      "epoch": 0.269888,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.360161290322581e-05,
+      "loss": 0.1659,
+      "step": 16868
+    },
+    {
+      "epoch": 0.269904,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.36e-05,
+      "loss": 0.188,
+      "step": 16869
+    },
+    {
+      "epoch": 0.26992,
+      "grad_norm": 1.4765625,
+      "learning_rate": 7.35983870967742e-05,
+      "loss": 0.1578,
+      "step": 16870
+    },
+    {
+      "epoch": 0.269936,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.359677419354839e-05,
+      "loss": 0.1541,
+      "step": 16871
+    },
+    {
+      "epoch": 0.269952,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.359516129032259e-05,
+      "loss": 0.1625,
+      "step": 16872
+    },
+    {
+      "epoch": 0.269968,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.359354838709677e-05,
+      "loss": 0.1921,
+      "step": 16873
+    },
+    {
+      "epoch": 0.269984,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.359193548387097e-05,
+      "loss": 0.1592,
+      "step": 16874
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.359032258064517e-05,
+      "loss": 0.2101,
+      "step": 16875
+    },
+    {
+      "epoch": 0.270016,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.358870967741936e-05,
+      "loss": 0.1418,
+      "step": 16876
+    },
+    {
+      "epoch": 0.270032,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.358709677419356e-05,
+      "loss": 0.192,
+      "step": 16877
+    },
+    {
+      "epoch": 0.270048,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.358548387096774e-05,
+      "loss": 0.2179,
+      "step": 16878
+    },
+    {
+      "epoch": 0.270064,
+      "grad_norm": 1.296875,
+      "learning_rate": 7.358387096774194e-05,
+      "loss": 0.1922,
+      "step": 16879
+    },
+    {
+      "epoch": 0.27008,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.358225806451613e-05,
+      "loss": 0.1409,
+      "step": 16880
+    },
+    {
+      "epoch": 0.270096,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.358064516129033e-05,
+      "loss": 0.1607,
+      "step": 16881
+    },
+    {
+      "epoch": 0.270112,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.357903225806451e-05,
+      "loss": 0.1512,
+      "step": 16882
+    },
+    {
+      "epoch": 0.270128,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.357741935483871e-05,
+      "loss": 0.1717,
+      "step": 16883
+    },
+    {
+      "epoch": 0.270144,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.35758064516129e-05,
+      "loss": 0.1528,
+      "step": 16884
+    },
+    {
+      "epoch": 0.27016,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.35741935483871e-05,
+      "loss": 0.1558,
+      "step": 16885
+    },
+    {
+      "epoch": 0.270176,
+      "grad_norm": 1.515625,
+      "learning_rate": 7.35725806451613e-05,
+      "loss": 0.2197,
+      "step": 16886
+    },
+    {
+      "epoch": 0.270192,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.35709677419355e-05,
+      "loss": 0.156,
+      "step": 16887
+    },
+    {
+      "epoch": 0.270208,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.356935483870968e-05,
+      "loss": 0.1592,
+      "step": 16888
+    },
+    {
+      "epoch": 0.270224,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.356774193548387e-05,
+      "loss": 0.1403,
+      "step": 16889
+    },
+    {
+      "epoch": 0.27024,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.356612903225807e-05,
+      "loss": 0.1736,
+      "step": 16890
+    },
+    {
+      "epoch": 0.270256,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.356451612903226e-05,
+      "loss": 0.1668,
+      "step": 16891
+    },
+    {
+      "epoch": 0.270272,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.356290322580646e-05,
+      "loss": 0.1827,
+      "step": 16892
+    },
+    {
+      "epoch": 0.270288,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.356129032258064e-05,
+      "loss": 0.1854,
+      "step": 16893
+    },
+    {
+      "epoch": 0.270304,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.355967741935484e-05,
+      "loss": 0.1845,
+      "step": 16894
+    },
+    {
+      "epoch": 0.27032,
+      "grad_norm": 1.2578125,
+      "learning_rate": 7.355806451612903e-05,
+      "loss": 0.1937,
+      "step": 16895
+    },
+    {
+      "epoch": 0.270336,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.355645161290323e-05,
+      "loss": 0.1714,
+      "step": 16896
+    },
+    {
+      "epoch": 0.270352,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.355483870967743e-05,
+      "loss": 0.2303,
+      "step": 16897
+    },
+    {
+      "epoch": 0.270368,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.355322580645163e-05,
+      "loss": 0.1827,
+      "step": 16898
+    },
+    {
+      "epoch": 0.270384,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.355161290322581e-05,
+      "loss": 0.1721,
+      "step": 16899
+    },
+    {
+      "epoch": 0.2704,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.355000000000001e-05,
+      "loss": 0.1746,
+      "step": 16900
+    },
+    {
+      "epoch": 0.270416,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.35483870967742e-05,
+      "loss": 0.2326,
+      "step": 16901
+    },
+    {
+      "epoch": 0.270432,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.35467741935484e-05,
+      "loss": 0.172,
+      "step": 16902
+    },
+    {
+      "epoch": 0.270448,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.354516129032258e-05,
+      "loss": 0.1767,
+      "step": 16903
+    },
+    {
+      "epoch": 0.270464,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.354354838709677e-05,
+      "loss": 0.1634,
+      "step": 16904
+    },
+    {
+      "epoch": 0.27048,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.354193548387097e-05,
+      "loss": 0.1712,
+      "step": 16905
+    },
+    {
+      "epoch": 0.270496,
+      "grad_norm": 1.4921875,
+      "learning_rate": 7.354032258064516e-05,
+      "loss": 0.1785,
+      "step": 16906
+    },
+    {
+      "epoch": 0.270512,
+      "grad_norm": 1.453125,
+      "learning_rate": 7.353870967741936e-05,
+      "loss": 0.1599,
+      "step": 16907
+    },
+    {
+      "epoch": 0.270528,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.353709677419354e-05,
+      "loss": 0.1474,
+      "step": 16908
+    },
+    {
+      "epoch": 0.270544,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.353548387096774e-05,
+      "loss": 0.1465,
+      "step": 16909
+    },
+    {
+      "epoch": 0.27056,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.353387096774194e-05,
+      "loss": 0.1773,
+      "step": 16910
+    },
+    {
+      "epoch": 0.270576,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.353225806451614e-05,
+      "loss": 0.1341,
+      "step": 16911
+    },
+    {
+      "epoch": 0.270592,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.353064516129033e-05,
+      "loss": 0.1738,
+      "step": 16912
+    },
+    {
+      "epoch": 0.270608,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.352903225806453e-05,
+      "loss": 0.1458,
+      "step": 16913
+    },
+    {
+      "epoch": 0.270624,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.352741935483871e-05,
+      "loss": 0.2418,
+      "step": 16914
+    },
+    {
+      "epoch": 0.27064,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.352580645161291e-05,
+      "loss": 0.1633,
+      "step": 16915
+    },
+    {
+      "epoch": 0.270656,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.35241935483871e-05,
+      "loss": 0.1528,
+      "step": 16916
+    },
+    {
+      "epoch": 0.270672,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.35225806451613e-05,
+      "loss": 0.2086,
+      "step": 16917
+    },
+    {
+      "epoch": 0.270688,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.352096774193548e-05,
+      "loss": 0.1675,
+      "step": 16918
+    },
+    {
+      "epoch": 0.270704,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.351935483870968e-05,
+      "loss": 0.1659,
+      "step": 16919
+    },
+    {
+      "epoch": 0.27072,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.351774193548387e-05,
+      "loss": 0.1639,
+      "step": 16920
+    },
+    {
+      "epoch": 0.270736,
+      "grad_norm": 0.49609375,
+      "learning_rate": 7.351612903225807e-05,
+      "loss": 0.1619,
+      "step": 16921
+    },
+    {
+      "epoch": 0.270752,
+      "grad_norm": 1.8828125,
+      "learning_rate": 7.351451612903227e-05,
+      "loss": 0.1669,
+      "step": 16922
+    },
+    {
+      "epoch": 0.270768,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.351290322580645e-05,
+      "loss": 0.1818,
+      "step": 16923
+    },
+    {
+      "epoch": 0.270784,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.351129032258065e-05,
+      "loss": 0.1535,
+      "step": 16924
+    },
+    {
+      "epoch": 0.2708,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.350967741935484e-05,
+      "loss": 0.1746,
+      "step": 16925
+    },
+    {
+      "epoch": 0.270816,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.350806451612904e-05,
+      "loss": 0.1962,
+      "step": 16926
+    },
+    {
+      "epoch": 0.270832,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.350645161290323e-05,
+      "loss": 0.1532,
+      "step": 16927
+    },
+    {
+      "epoch": 0.270848,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.350483870967743e-05,
+      "loss": 0.157,
+      "step": 16928
+    },
+    {
+      "epoch": 0.270864,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.350322580645161e-05,
+      "loss": 0.1721,
+      "step": 16929
+    },
+    {
+      "epoch": 0.27088,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.350161290322581e-05,
+      "loss": 0.1632,
+      "step": 16930
+    },
+    {
+      "epoch": 0.270896,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.35e-05,
+      "loss": 0.1668,
+      "step": 16931
+    },
+    {
+      "epoch": 0.270912,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.34983870967742e-05,
+      "loss": 0.1711,
+      "step": 16932
+    },
+    {
+      "epoch": 0.270928,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.349677419354838e-05,
+      "loss": 0.1468,
+      "step": 16933
+    },
+    {
+      "epoch": 0.270944,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.349516129032258e-05,
+      "loss": 0.2347,
+      "step": 16934
+    },
+    {
+      "epoch": 0.27096,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.349354838709678e-05,
+      "loss": 0.1473,
+      "step": 16935
+    },
+    {
+      "epoch": 0.270976,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.349193548387097e-05,
+      "loss": 0.1205,
+      "step": 16936
+    },
+    {
+      "epoch": 0.270992,
+      "grad_norm": 1.53125,
+      "learning_rate": 7.349032258064517e-05,
+      "loss": 0.1779,
+      "step": 16937
+    },
+    {
+      "epoch": 0.271008,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.348870967741935e-05,
+      "loss": 0.1577,
+      "step": 16938
+    },
+    {
+      "epoch": 0.271024,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.348709677419355e-05,
+      "loss": 0.158,
+      "step": 16939
+    },
+    {
+      "epoch": 0.27104,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.348548387096774e-05,
+      "loss": 0.1904,
+      "step": 16940
+    },
+    {
+      "epoch": 0.271056,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.348387096774194e-05,
+      "loss": 0.1479,
+      "step": 16941
+    },
+    {
+      "epoch": 0.271072,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.348225806451613e-05,
+      "loss": 0.1942,
+      "step": 16942
+    },
+    {
+      "epoch": 0.271088,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.348064516129033e-05,
+      "loss": 0.1613,
+      "step": 16943
+    },
+    {
+      "epoch": 0.271104,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.347903225806451e-05,
+      "loss": 0.1825,
+      "step": 16944
+    },
+    {
+      "epoch": 0.27112,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.347741935483871e-05,
+      "loss": 0.1837,
+      "step": 16945
+    },
+    {
+      "epoch": 0.271136,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.347580645161291e-05,
+      "loss": 0.1726,
+      "step": 16946
+    },
+    {
+      "epoch": 0.271152,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.347419354838711e-05,
+      "loss": 0.1714,
+      "step": 16947
+    },
+    {
+      "epoch": 0.271168,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.34725806451613e-05,
+      "loss": 0.1692,
+      "step": 16948
+    },
+    {
+      "epoch": 0.271184,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.34709677419355e-05,
+      "loss": 0.1678,
+      "step": 16949
+    },
+    {
+      "epoch": 0.2712,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.346935483870968e-05,
+      "loss": 0.1809,
+      "step": 16950
+    },
+    {
+      "epoch": 0.271216,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.346774193548387e-05,
+      "loss": 0.1805,
+      "step": 16951
+    },
+    {
+      "epoch": 0.271232,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.346612903225807e-05,
+      "loss": 0.2147,
+      "step": 16952
+    },
+    {
+      "epoch": 0.271248,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.346451612903225e-05,
+      "loss": 0.1568,
+      "step": 16953
+    },
+    {
+      "epoch": 0.271264,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.346290322580645e-05,
+      "loss": 0.2065,
+      "step": 16954
+    },
+    {
+      "epoch": 0.27128,
+      "grad_norm": 0.75,
+      "learning_rate": 7.346129032258064e-05,
+      "loss": 0.1628,
+      "step": 16955
+    },
+    {
+      "epoch": 0.271296,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.345967741935484e-05,
+      "loss": 0.1914,
+      "step": 16956
+    },
+    {
+      "epoch": 0.271312,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.345806451612904e-05,
+      "loss": 0.1829,
+      "step": 16957
+    },
+    {
+      "epoch": 0.271328,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.345645161290324e-05,
+      "loss": 0.213,
+      "step": 16958
+    },
+    {
+      "epoch": 0.271344,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.345483870967742e-05,
+      "loss": 0.1758,
+      "step": 16959
+    },
+    {
+      "epoch": 0.27136,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.345322580645162e-05,
+      "loss": 0.1917,
+      "step": 16960
+    },
+    {
+      "epoch": 0.271376,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.345161290322581e-05,
+      "loss": 0.173,
+      "step": 16961
+    },
+    {
+      "epoch": 0.271392,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.345000000000001e-05,
+      "loss": 0.1592,
+      "step": 16962
+    },
+    {
+      "epoch": 0.271408,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.34483870967742e-05,
+      "loss": 0.1773,
+      "step": 16963
+    },
+    {
+      "epoch": 0.271424,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.34467741935484e-05,
+      "loss": 0.1657,
+      "step": 16964
+    },
+    {
+      "epoch": 0.27144,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.344516129032258e-05,
+      "loss": 0.1771,
+      "step": 16965
+    },
+    {
+      "epoch": 0.271456,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.344354838709677e-05,
+      "loss": 0.1953,
+      "step": 16966
+    },
+    {
+      "epoch": 0.271472,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.344193548387097e-05,
+      "loss": 0.1807,
+      "step": 16967
+    },
+    {
+      "epoch": 0.271488,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.344032258064515e-05,
+      "loss": 0.2006,
+      "step": 16968
+    },
+    {
+      "epoch": 0.271504,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.343870967741935e-05,
+      "loss": 0.1273,
+      "step": 16969
+    },
+    {
+      "epoch": 0.27152,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.343709677419355e-05,
+      "loss": 0.2197,
+      "step": 16970
+    },
+    {
+      "epoch": 0.271536,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.343548387096775e-05,
+      "loss": 0.1526,
+      "step": 16971
+    },
+    {
+      "epoch": 0.271552,
+      "grad_norm": 0.48046875,
+      "learning_rate": 7.343387096774194e-05,
+      "loss": 0.1491,
+      "step": 16972
+    },
+    {
+      "epoch": 0.271568,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.343225806451614e-05,
+      "loss": 0.1617,
+      "step": 16973
+    },
+    {
+      "epoch": 0.271584,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.343064516129032e-05,
+      "loss": 0.1258,
+      "step": 16974
+    },
+    {
+      "epoch": 0.2716,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.342903225806452e-05,
+      "loss": 0.1765,
+      "step": 16975
+    },
+    {
+      "epoch": 0.271616,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.342741935483871e-05,
+      "loss": 0.1638,
+      "step": 16976
+    },
+    {
+      "epoch": 0.271632,
+      "grad_norm": 1.4765625,
+      "learning_rate": 7.342580645161291e-05,
+      "loss": 0.1723,
+      "step": 16977
+    },
+    {
+      "epoch": 0.271648,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.34241935483871e-05,
+      "loss": 0.1588,
+      "step": 16978
+    },
+    {
+      "epoch": 0.271664,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.34225806451613e-05,
+      "loss": 0.179,
+      "step": 16979
+    },
+    {
+      "epoch": 0.27168,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.342096774193548e-05,
+      "loss": 0.1562,
+      "step": 16980
+    },
+    {
+      "epoch": 0.271696,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.341935483870968e-05,
+      "loss": 0.1643,
+      "step": 16981
+    },
+    {
+      "epoch": 0.271712,
+      "grad_norm": 0.75,
+      "learning_rate": 7.341774193548388e-05,
+      "loss": 0.1697,
+      "step": 16982
+    },
+    {
+      "epoch": 0.271728,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.341612903225807e-05,
+      "loss": 0.1852,
+      "step": 16983
+    },
+    {
+      "epoch": 0.271744,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.341451612903227e-05,
+      "loss": 0.1328,
+      "step": 16984
+    },
+    {
+      "epoch": 0.27176,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.341290322580645e-05,
+      "loss": 0.1395,
+      "step": 16985
+    },
+    {
+      "epoch": 0.271776,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.341129032258065e-05,
+      "loss": 0.1588,
+      "step": 16986
+    },
+    {
+      "epoch": 0.271792,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.340967741935484e-05,
+      "loss": 0.1809,
+      "step": 16987
+    },
+    {
+      "epoch": 0.271808,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.340806451612904e-05,
+      "loss": 0.1614,
+      "step": 16988
+    },
+    {
+      "epoch": 0.271824,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.340645161290322e-05,
+      "loss": 0.1778,
+      "step": 16989
+    },
+    {
+      "epoch": 0.27184,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.340483870967742e-05,
+      "loss": 0.1933,
+      "step": 16990
+    },
+    {
+      "epoch": 0.271856,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.340322580645161e-05,
+      "loss": 0.1965,
+      "step": 16991
+    },
+    {
+      "epoch": 0.271872,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.340161290322581e-05,
+      "loss": 0.1634,
+      "step": 16992
+    },
+    {
+      "epoch": 0.271888,
+      "grad_norm": 0.4765625,
+      "learning_rate": 7.340000000000001e-05,
+      "loss": 0.1257,
+      "step": 16993
+    },
+    {
+      "epoch": 0.271904,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.33983870967742e-05,
+      "loss": 0.1358,
+      "step": 16994
+    },
+    {
+      "epoch": 0.27192,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.33967741935484e-05,
+      "loss": 0.1747,
+      "step": 16995
+    },
+    {
+      "epoch": 0.271936,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.33951612903226e-05,
+      "loss": 0.1624,
+      "step": 16996
+    },
+    {
+      "epoch": 0.271952,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.339354838709678e-05,
+      "loss": 0.1743,
+      "step": 16997
+    },
+    {
+      "epoch": 0.271968,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.339193548387097e-05,
+      "loss": 0.1789,
+      "step": 16998
+    },
+    {
+      "epoch": 0.271984,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.339032258064517e-05,
+      "loss": 0.1309,
+      "step": 16999
+    },
+    {
+      "epoch": 0.272,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.338870967741935e-05,
+      "loss": 0.1512,
+      "step": 17000
+    },
+    {
+      "epoch": 0.272016,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.338709677419355e-05,
+      "loss": 0.1784,
+      "step": 17001
+    },
+    {
+      "epoch": 0.272032,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.338548387096774e-05,
+      "loss": 0.1918,
+      "step": 17002
+    },
+    {
+      "epoch": 0.272048,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.338387096774194e-05,
+      "loss": 0.1812,
+      "step": 17003
+    },
+    {
+      "epoch": 0.272064,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.338225806451612e-05,
+      "loss": 0.1574,
+      "step": 17004
+    },
+    {
+      "epoch": 0.27208,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.338064516129032e-05,
+      "loss": 0.1557,
+      "step": 17005
+    },
+    {
+      "epoch": 0.272096,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.337903225806452e-05,
+      "loss": 0.1686,
+      "step": 17006
+    },
+    {
+      "epoch": 0.272112,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.337741935483872e-05,
+      "loss": 0.1502,
+      "step": 17007
+    },
+    {
+      "epoch": 0.272128,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.337580645161291e-05,
+      "loss": 0.1571,
+      "step": 17008
+    },
+    {
+      "epoch": 0.272144,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.337419354838711e-05,
+      "loss": 0.189,
+      "step": 17009
+    },
+    {
+      "epoch": 0.27216,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.33725806451613e-05,
+      "loss": 0.137,
+      "step": 17010
+    },
+    {
+      "epoch": 0.272176,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.33709677419355e-05,
+      "loss": 0.1912,
+      "step": 17011
+    },
+    {
+      "epoch": 0.272192,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.336935483870968e-05,
+      "loss": 0.1614,
+      "step": 17012
+    },
+    {
+      "epoch": 0.272208,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.336774193548387e-05,
+      "loss": 0.145,
+      "step": 17013
+    },
+    {
+      "epoch": 0.272224,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.336612903225807e-05,
+      "loss": 0.1634,
+      "step": 17014
+    },
+    {
+      "epoch": 0.27224,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.336451612903225e-05,
+      "loss": 0.173,
+      "step": 17015
+    },
+    {
+      "epoch": 0.272256,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.336290322580645e-05,
+      "loss": 0.1775,
+      "step": 17016
+    },
+    {
+      "epoch": 0.272272,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.336129032258065e-05,
+      "loss": 0.1695,
+      "step": 17017
+    },
+    {
+      "epoch": 0.272288,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.335967741935485e-05,
+      "loss": 0.1729,
+      "step": 17018
+    },
+    {
+      "epoch": 0.272304,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.335806451612904e-05,
+      "loss": 0.1724,
+      "step": 17019
+    },
+    {
+      "epoch": 0.27232,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.335645161290324e-05,
+      "loss": 0.1673,
+      "step": 17020
+    },
+    {
+      "epoch": 0.272336,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.335483870967742e-05,
+      "loss": 0.1469,
+      "step": 17021
+    },
+    {
+      "epoch": 0.272352,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.335322580645162e-05,
+      "loss": 0.1543,
+      "step": 17022
+    },
+    {
+      "epoch": 0.272368,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.335161290322581e-05,
+      "loss": 0.1828,
+      "step": 17023
+    },
+    {
+      "epoch": 0.272384,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.335000000000001e-05,
+      "loss": 0.1579,
+      "step": 17024
+    },
+    {
+      "epoch": 0.2724,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.33483870967742e-05,
+      "loss": 0.1588,
+      "step": 17025
+    },
+    {
+      "epoch": 0.272416,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.33467741935484e-05,
+      "loss": 0.1589,
+      "step": 17026
+    },
+    {
+      "epoch": 0.272432,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.334516129032258e-05,
+      "loss": 0.1749,
+      "step": 17027
+    },
+    {
+      "epoch": 0.272448,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.334354838709678e-05,
+      "loss": 0.1633,
+      "step": 17028
+    },
+    {
+      "epoch": 0.272464,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.334193548387097e-05,
+      "loss": 0.1341,
+      "step": 17029
+    },
+    {
+      "epoch": 0.27248,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.334032258064516e-05,
+      "loss": 0.1769,
+      "step": 17030
+    },
+    {
+      "epoch": 0.272496,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.333870967741936e-05,
+      "loss": 0.1416,
+      "step": 17031
+    },
+    {
+      "epoch": 0.272512,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.333709677419355e-05,
+      "loss": 0.1941,
+      "step": 17032
+    },
+    {
+      "epoch": 0.272528,
+      "grad_norm": 1.2109375,
+      "learning_rate": 7.333548387096775e-05,
+      "loss": 0.1583,
+      "step": 17033
+    },
+    {
+      "epoch": 0.272544,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.333387096774194e-05,
+      "loss": 0.178,
+      "step": 17034
+    },
+    {
+      "epoch": 0.27256,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.333225806451614e-05,
+      "loss": 0.1788,
+      "step": 17035
+    },
+    {
+      "epoch": 0.272576,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.333064516129032e-05,
+      "loss": 0.1853,
+      "step": 17036
+    },
+    {
+      "epoch": 0.272592,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.332903225806452e-05,
+      "loss": 0.1681,
+      "step": 17037
+    },
+    {
+      "epoch": 0.272608,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.332741935483871e-05,
+      "loss": 0.1855,
+      "step": 17038
+    },
+    {
+      "epoch": 0.272624,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.332580645161291e-05,
+      "loss": 0.2056,
+      "step": 17039
+    },
+    {
+      "epoch": 0.27264,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.332419354838709e-05,
+      "loss": 0.1559,
+      "step": 17040
+    },
+    {
+      "epoch": 0.272656,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.332258064516129e-05,
+      "loss": 0.1669,
+      "step": 17041
+    },
+    {
+      "epoch": 0.272672,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.332096774193549e-05,
+      "loss": 0.1689,
+      "step": 17042
+    },
+    {
+      "epoch": 0.272688,
+      "grad_norm": 0.625,
+      "learning_rate": 7.331935483870969e-05,
+      "loss": 0.1928,
+      "step": 17043
+    },
+    {
+      "epoch": 0.272704,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.331774193548388e-05,
+      "loss": 0.1795,
+      "step": 17044
+    },
+    {
+      "epoch": 0.27272,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.331612903225806e-05,
+      "loss": 0.1579,
+      "step": 17045
+    },
+    {
+      "epoch": 0.272736,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.331451612903226e-05,
+      "loss": 0.1647,
+      "step": 17046
+    },
+    {
+      "epoch": 0.272752,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.331290322580645e-05,
+      "loss": 0.1928,
+      "step": 17047
+    },
+    {
+      "epoch": 0.272768,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.331129032258065e-05,
+      "loss": 0.2484,
+      "step": 17048
+    },
+    {
+      "epoch": 0.272784,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.330967741935484e-05,
+      "loss": 0.129,
+      "step": 17049
+    },
+    {
+      "epoch": 0.2728,
+      "grad_norm": 1.3046875,
+      "learning_rate": 7.330806451612904e-05,
+      "loss": 0.1653,
+      "step": 17050
+    },
+    {
+      "epoch": 0.272816,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.330645161290322e-05,
+      "loss": 0.2099,
+      "step": 17051
+    },
+    {
+      "epoch": 0.272832,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.330483870967742e-05,
+      "loss": 0.1903,
+      "step": 17052
+    },
+    {
+      "epoch": 0.272848,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.330322580645162e-05,
+      "loss": 0.1498,
+      "step": 17053
+    },
+    {
+      "epoch": 0.272864,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.330161290322582e-05,
+      "loss": 0.1252,
+      "step": 17054
+    },
+    {
+      "epoch": 0.27288,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.33e-05,
+      "loss": 0.1669,
+      "step": 17055
+    },
+    {
+      "epoch": 0.272896,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.32983870967742e-05,
+      "loss": 0.1788,
+      "step": 17056
+    },
+    {
+      "epoch": 0.272912,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.329677419354839e-05,
+      "loss": 0.158,
+      "step": 17057
+    },
+    {
+      "epoch": 0.272928,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.329516129032259e-05,
+      "loss": 0.2003,
+      "step": 17058
+    },
+    {
+      "epoch": 0.272944,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.329354838709678e-05,
+      "loss": 0.181,
+      "step": 17059
+    },
+    {
+      "epoch": 0.27296,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.329193548387096e-05,
+      "loss": 0.2012,
+      "step": 17060
+    },
+    {
+      "epoch": 0.272976,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.329032258064516e-05,
+      "loss": 0.1846,
+      "step": 17061
+    },
+    {
+      "epoch": 0.272992,
+      "grad_norm": 1.421875,
+      "learning_rate": 7.328870967741935e-05,
+      "loss": 0.1796,
+      "step": 17062
+    },
+    {
+      "epoch": 0.273008,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.328709677419355e-05,
+      "loss": 0.1642,
+      "step": 17063
+    },
+    {
+      "epoch": 0.273024,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.328548387096774e-05,
+      "loss": 0.2015,
+      "step": 17064
+    },
+    {
+      "epoch": 0.27304,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.328387096774193e-05,
+      "loss": 0.1401,
+      "step": 17065
+    },
+    {
+      "epoch": 0.273056,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.328225806451613e-05,
+      "loss": 0.1852,
+      "step": 17066
+    },
+    {
+      "epoch": 0.273072,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.328064516129033e-05,
+      "loss": 0.1657,
+      "step": 17067
+    },
+    {
+      "epoch": 0.273088,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.327903225806452e-05,
+      "loss": 0.1525,
+      "step": 17068
+    },
+    {
+      "epoch": 0.273104,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.327741935483872e-05,
+      "loss": 0.1809,
+      "step": 17069
+    },
+    {
+      "epoch": 0.27312,
+      "grad_norm": 1.2890625,
+      "learning_rate": 7.32758064516129e-05,
+      "loss": 0.1678,
+      "step": 17070
+    },
+    {
+      "epoch": 0.273136,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.32741935483871e-05,
+      "loss": 0.1465,
+      "step": 17071
+    },
+    {
+      "epoch": 0.273152,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.327258064516129e-05,
+      "loss": 0.1508,
+      "step": 17072
+    },
+    {
+      "epoch": 0.273168,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.327096774193549e-05,
+      "loss": 0.1741,
+      "step": 17073
+    },
+    {
+      "epoch": 0.273184,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.326935483870968e-05,
+      "loss": 0.1832,
+      "step": 17074
+    },
+    {
+      "epoch": 0.2732,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.326774193548386e-05,
+      "loss": 0.1611,
+      "step": 17075
+    },
+    {
+      "epoch": 0.273216,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.326612903225806e-05,
+      "loss": 0.1879,
+      "step": 17076
+    },
+    {
+      "epoch": 0.273232,
+      "grad_norm": 1.2734375,
+      "learning_rate": 7.326451612903226e-05,
+      "loss": 0.1634,
+      "step": 17077
+    },
+    {
+      "epoch": 0.273248,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.326290322580646e-05,
+      "loss": 0.1883,
+      "step": 17078
+    },
+    {
+      "epoch": 0.273264,
+      "grad_norm": 1.0,
+      "learning_rate": 7.326129032258065e-05,
+      "loss": 0.1631,
+      "step": 17079
+    },
+    {
+      "epoch": 0.27328,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.325967741935485e-05,
+      "loss": 0.1647,
+      "step": 17080
+    },
+    {
+      "epoch": 0.273296,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.325806451612903e-05,
+      "loss": 0.1853,
+      "step": 17081
+    },
+    {
+      "epoch": 0.273312,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.325645161290323e-05,
+      "loss": 0.17,
+      "step": 17082
+    },
+    {
+      "epoch": 0.273328,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.325483870967742e-05,
+      "loss": 0.1576,
+      "step": 17083
+    },
+    {
+      "epoch": 0.273344,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.325322580645162e-05,
+      "loss": 0.1747,
+      "step": 17084
+    },
+    {
+      "epoch": 0.27336,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.32516129032258e-05,
+      "loss": 0.1674,
+      "step": 17085
+    },
+    {
+      "epoch": 0.273376,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.325e-05,
+      "loss": 0.1877,
+      "step": 17086
+    },
+    {
+      "epoch": 0.273392,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.324838709677419e-05,
+      "loss": 0.1509,
+      "step": 17087
+    },
+    {
+      "epoch": 0.273408,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.324677419354839e-05,
+      "loss": 0.1651,
+      "step": 17088
+    },
+    {
+      "epoch": 0.273424,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.324516129032259e-05,
+      "loss": 0.1566,
+      "step": 17089
+    },
+    {
+      "epoch": 0.27344,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.324354838709678e-05,
+      "loss": 0.1374,
+      "step": 17090
+    },
+    {
+      "epoch": 0.273456,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.324193548387098e-05,
+      "loss": 0.1574,
+      "step": 17091
+    },
+    {
+      "epoch": 0.273472,
+      "grad_norm": 1.296875,
+      "learning_rate": 7.324032258064516e-05,
+      "loss": 0.2059,
+      "step": 17092
+    },
+    {
+      "epoch": 0.273488,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.323870967741936e-05,
+      "loss": 0.1796,
+      "step": 17093
+    },
+    {
+      "epoch": 0.273504,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.323709677419355e-05,
+      "loss": 0.1674,
+      "step": 17094
+    },
+    {
+      "epoch": 0.27352,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.323548387096775e-05,
+      "loss": 0.1692,
+      "step": 17095
+    },
+    {
+      "epoch": 0.273536,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.323387096774193e-05,
+      "loss": 0.166,
+      "step": 17096
+    },
+    {
+      "epoch": 0.273552,
+      "grad_norm": 1.2578125,
+      "learning_rate": 7.323225806451613e-05,
+      "loss": 0.1718,
+      "step": 17097
+    },
+    {
+      "epoch": 0.273568,
+      "grad_norm": 1.3046875,
+      "learning_rate": 7.323064516129032e-05,
+      "loss": 0.1876,
+      "step": 17098
+    },
+    {
+      "epoch": 0.273584,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.322903225806452e-05,
+      "loss": 0.1593,
+      "step": 17099
+    },
+    {
+      "epoch": 0.2736,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.32274193548387e-05,
+      "loss": 0.1368,
+      "step": 17100
+    },
+    {
+      "epoch": 0.273616,
+      "grad_norm": 0.75,
+      "learning_rate": 7.32258064516129e-05,
+      "loss": 0.1649,
+      "step": 17101
+    },
+    {
+      "epoch": 0.273632,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.32241935483871e-05,
+      "loss": 0.1715,
+      "step": 17102
+    },
+    {
+      "epoch": 0.273648,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.32225806451613e-05,
+      "loss": 0.1668,
+      "step": 17103
+    },
+    {
+      "epoch": 0.273664,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.322096774193549e-05,
+      "loss": 0.1523,
+      "step": 17104
+    },
+    {
+      "epoch": 0.27368,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.321935483870969e-05,
+      "loss": 0.1577,
+      "step": 17105
+    },
+    {
+      "epoch": 0.273696,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.321774193548388e-05,
+      "loss": 0.158,
+      "step": 17106
+    },
+    {
+      "epoch": 0.273712,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.321612903225806e-05,
+      "loss": 0.1545,
+      "step": 17107
+    },
+    {
+      "epoch": 0.273728,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.321451612903226e-05,
+      "loss": 0.1705,
+      "step": 17108
+    },
+    {
+      "epoch": 0.273744,
+      "grad_norm": 0.75,
+      "learning_rate": 7.321290322580645e-05,
+      "loss": 0.1864,
+      "step": 17109
+    },
+    {
+      "epoch": 0.27376,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.321129032258065e-05,
+      "loss": 0.1454,
+      "step": 17110
+    },
+    {
+      "epoch": 0.273776,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.320967741935483e-05,
+      "loss": 0.142,
+      "step": 17111
+    },
+    {
+      "epoch": 0.273792,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.320806451612903e-05,
+      "loss": 0.1859,
+      "step": 17112
+    },
+    {
+      "epoch": 0.273808,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.320645161290323e-05,
+      "loss": 0.14,
+      "step": 17113
+    },
+    {
+      "epoch": 0.273824,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.320483870967743e-05,
+      "loss": 0.1867,
+      "step": 17114
+    },
+    {
+      "epoch": 0.27384,
+      "grad_norm": 1.34375,
+      "learning_rate": 7.320322580645162e-05,
+      "loss": 0.157,
+      "step": 17115
+    },
+    {
+      "epoch": 0.273856,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.320161290322582e-05,
+      "loss": 0.2096,
+      "step": 17116
+    },
+    {
+      "epoch": 0.273872,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.32e-05,
+      "loss": 0.1555,
+      "step": 17117
+    },
+    {
+      "epoch": 0.273888,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.31983870967742e-05,
+      "loss": 0.1727,
+      "step": 17118
+    },
+    {
+      "epoch": 0.273904,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.319677419354839e-05,
+      "loss": 0.1371,
+      "step": 17119
+    },
+    {
+      "epoch": 0.27392,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.319516129032259e-05,
+      "loss": 0.1543,
+      "step": 17120
+    },
+    {
+      "epoch": 0.273936,
+      "grad_norm": 0.75,
+      "learning_rate": 7.319354838709678e-05,
+      "loss": 0.1863,
+      "step": 17121
+    },
+    {
+      "epoch": 0.273952,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.319193548387096e-05,
+      "loss": 0.1695,
+      "step": 17122
+    },
+    {
+      "epoch": 0.273968,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.319032258064516e-05,
+      "loss": 0.1708,
+      "step": 17123
+    },
+    {
+      "epoch": 0.273984,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.318870967741935e-05,
+      "loss": 0.1948,
+      "step": 17124
+    },
+    {
+      "epoch": 0.274,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.318709677419355e-05,
+      "loss": 0.1582,
+      "step": 17125
+    },
+    {
+      "epoch": 0.274016,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.318548387096775e-05,
+      "loss": 0.144,
+      "step": 17126
+    },
+    {
+      "epoch": 0.274032,
+      "grad_norm": 0.875,
+      "learning_rate": 7.318387096774195e-05,
+      "loss": 0.1687,
+      "step": 17127
+    },
+    {
+      "epoch": 0.274048,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.318225806451613e-05,
+      "loss": 0.1417,
+      "step": 17128
+    },
+    {
+      "epoch": 0.274064,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.318064516129033e-05,
+      "loss": 0.1557,
+      "step": 17129
+    },
+    {
+      "epoch": 0.27408,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.317903225806452e-05,
+      "loss": 0.1871,
+      "step": 17130
+    },
+    {
+      "epoch": 0.274096,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.317741935483872e-05,
+      "loss": 0.1619,
+      "step": 17131
+    },
+    {
+      "epoch": 0.274112,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.31758064516129e-05,
+      "loss": 0.1865,
+      "step": 17132
+    },
+    {
+      "epoch": 0.274128,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.31741935483871e-05,
+      "loss": 0.1148,
+      "step": 17133
+    },
+    {
+      "epoch": 0.274144,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.317258064516129e-05,
+      "loss": 0.1614,
+      "step": 17134
+    },
+    {
+      "epoch": 0.27416,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.317096774193549e-05,
+      "loss": 0.1398,
+      "step": 17135
+    },
+    {
+      "epoch": 0.274176,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.316935483870967e-05,
+      "loss": 0.1975,
+      "step": 17136
+    },
+    {
+      "epoch": 0.274192,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.316774193548387e-05,
+      "loss": 0.1958,
+      "step": 17137
+    },
+    {
+      "epoch": 0.274208,
+      "grad_norm": 0.75,
+      "learning_rate": 7.316612903225807e-05,
+      "loss": 0.1891,
+      "step": 17138
+    },
+    {
+      "epoch": 0.274224,
+      "grad_norm": 0.75,
+      "learning_rate": 7.316451612903226e-05,
+      "loss": 0.1441,
+      "step": 17139
+    },
+    {
+      "epoch": 0.27424,
+      "grad_norm": 1.296875,
+      "learning_rate": 7.316290322580646e-05,
+      "loss": 0.1965,
+      "step": 17140
+    },
+    {
+      "epoch": 0.274256,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.316129032258065e-05,
+      "loss": 0.1983,
+      "step": 17141
+    },
+    {
+      "epoch": 0.274272,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.315967741935485e-05,
+      "loss": 0.198,
+      "step": 17142
+    },
+    {
+      "epoch": 0.274288,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.315806451612903e-05,
+      "loss": 0.1502,
+      "step": 17143
+    },
+    {
+      "epoch": 0.274304,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.315645161290323e-05,
+      "loss": 0.1642,
+      "step": 17144
+    },
+    {
+      "epoch": 0.27432,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.315483870967742e-05,
+      "loss": 0.1714,
+      "step": 17145
+    },
+    {
+      "epoch": 0.274336,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.315322580645162e-05,
+      "loss": 0.1831,
+      "step": 17146
+    },
+    {
+      "epoch": 0.274352,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.31516129032258e-05,
+      "loss": 0.1606,
+      "step": 17147
+    },
+    {
+      "epoch": 0.274368,
+      "grad_norm": 1.3828125,
+      "learning_rate": 7.315e-05,
+      "loss": 0.2123,
+      "step": 17148
+    },
+    {
+      "epoch": 0.274384,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.31483870967742e-05,
+      "loss": 0.1789,
+      "step": 17149
+    },
+    {
+      "epoch": 0.2744,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.31467741935484e-05,
+      "loss": 0.2095,
+      "step": 17150
+    },
+    {
+      "epoch": 0.274416,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.314516129032259e-05,
+      "loss": 0.1614,
+      "step": 17151
+    },
+    {
+      "epoch": 0.274432,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.314354838709679e-05,
+      "loss": 0.1511,
+      "step": 17152
+    },
+    {
+      "epoch": 0.274448,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.314193548387097e-05,
+      "loss": 0.1944,
+      "step": 17153
+    },
+    {
+      "epoch": 0.274464,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.314032258064516e-05,
+      "loss": 0.1385,
+      "step": 17154
+    },
+    {
+      "epoch": 0.27448,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.313870967741936e-05,
+      "loss": 0.1659,
+      "step": 17155
+    },
+    {
+      "epoch": 0.274496,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.313709677419355e-05,
+      "loss": 0.178,
+      "step": 17156
+    },
+    {
+      "epoch": 0.274512,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.313548387096775e-05,
+      "loss": 0.1761,
+      "step": 17157
+    },
+    {
+      "epoch": 0.274528,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.313387096774193e-05,
+      "loss": 0.176,
+      "step": 17158
+    },
+    {
+      "epoch": 0.274544,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.313225806451613e-05,
+      "loss": 0.1691,
+      "step": 17159
+    },
+    {
+      "epoch": 0.27456,
+      "grad_norm": 1.4296875,
+      "learning_rate": 7.313064516129032e-05,
+      "loss": 0.1655,
+      "step": 17160
+    },
+    {
+      "epoch": 0.274576,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.312903225806452e-05,
+      "loss": 0.1588,
+      "step": 17161
+    },
+    {
+      "epoch": 0.274592,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.312741935483872e-05,
+      "loss": 0.1675,
+      "step": 17162
+    },
+    {
+      "epoch": 0.274608,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.312580645161292e-05,
+      "loss": 0.1818,
+      "step": 17163
+    },
+    {
+      "epoch": 0.274624,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.31241935483871e-05,
+      "loss": 0.176,
+      "step": 17164
+    },
+    {
+      "epoch": 0.27464,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.31225806451613e-05,
+      "loss": 0.189,
+      "step": 17165
+    },
+    {
+      "epoch": 0.274656,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.312096774193549e-05,
+      "loss": 0.1548,
+      "step": 17166
+    },
+    {
+      "epoch": 0.274672,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.311935483870969e-05,
+      "loss": 0.1487,
+      "step": 17167
+    },
+    {
+      "epoch": 0.274688,
+      "grad_norm": 1.4453125,
+      "learning_rate": 7.311774193548387e-05,
+      "loss": 0.2193,
+      "step": 17168
+    },
+    {
+      "epoch": 0.274704,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.311612903225806e-05,
+      "loss": 0.1747,
+      "step": 17169
+    },
+    {
+      "epoch": 0.27472,
+      "grad_norm": 1.5078125,
+      "learning_rate": 7.311451612903226e-05,
+      "loss": 0.2146,
+      "step": 17170
+    },
+    {
+      "epoch": 0.274736,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.311290322580645e-05,
+      "loss": 0.1749,
+      "step": 17171
+    },
+    {
+      "epoch": 0.274752,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.311129032258064e-05,
+      "loss": 0.1822,
+      "step": 17172
+    },
+    {
+      "epoch": 0.274768,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.310967741935484e-05,
+      "loss": 0.1396,
+      "step": 17173
+    },
+    {
+      "epoch": 0.274784,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.310806451612904e-05,
+      "loss": 0.1621,
+      "step": 17174
+    },
+    {
+      "epoch": 0.2748,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.310645161290323e-05,
+      "loss": 0.1642,
+      "step": 17175
+    },
+    {
+      "epoch": 0.274816,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.310483870967743e-05,
+      "loss": 0.1794,
+      "step": 17176
+    },
+    {
+      "epoch": 0.274832,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.310322580645162e-05,
+      "loss": 0.1626,
+      "step": 17177
+    },
+    {
+      "epoch": 0.274848,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.310161290322582e-05,
+      "loss": 0.1698,
+      "step": 17178
+    },
+    {
+      "epoch": 0.274864,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.31e-05,
+      "loss": 0.1739,
+      "step": 17179
+    },
+    {
+      "epoch": 0.27488,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.30983870967742e-05,
+      "loss": 0.1737,
+      "step": 17180
+    },
+    {
+      "epoch": 0.274896,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.309677419354839e-05,
+      "loss": 0.1781,
+      "step": 17181
+    },
+    {
+      "epoch": 0.274912,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.309516129032259e-05,
+      "loss": 0.2049,
+      "step": 17182
+    },
+    {
+      "epoch": 0.274928,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.309354838709677e-05,
+      "loss": 0.1574,
+      "step": 17183
+    },
+    {
+      "epoch": 0.274944,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.309193548387097e-05,
+      "loss": 0.168,
+      "step": 17184
+    },
+    {
+      "epoch": 0.27496,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.309032258064516e-05,
+      "loss": 0.198,
+      "step": 17185
+    },
+    {
+      "epoch": 0.274976,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.308870967741936e-05,
+      "loss": 0.1486,
+      "step": 17186
+    },
+    {
+      "epoch": 0.274992,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.308709677419356e-05,
+      "loss": 0.2221,
+      "step": 17187
+    },
+    {
+      "epoch": 0.275008,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.308548387096774e-05,
+      "loss": 0.1846,
+      "step": 17188
+    },
+    {
+      "epoch": 0.275024,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.308387096774194e-05,
+      "loss": 0.1855,
+      "step": 17189
+    },
+    {
+      "epoch": 0.27504,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.308225806451613e-05,
+      "loss": 0.1487,
+      "step": 17190
+    },
+    {
+      "epoch": 0.275056,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.308064516129033e-05,
+      "loss": 0.1925,
+      "step": 17191
+    },
+    {
+      "epoch": 0.275072,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.307903225806452e-05,
+      "loss": 0.1771,
+      "step": 17192
+    },
+    {
+      "epoch": 0.275088,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.307741935483871e-05,
+      "loss": 0.2015,
+      "step": 17193
+    },
+    {
+      "epoch": 0.275104,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.30758064516129e-05,
+      "loss": 0.1442,
+      "step": 17194
+    },
+    {
+      "epoch": 0.27512,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.30741935483871e-05,
+      "loss": 0.1761,
+      "step": 17195
+    },
+    {
+      "epoch": 0.275136,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.307258064516129e-05,
+      "loss": 0.1195,
+      "step": 17196
+    },
+    {
+      "epoch": 0.275152,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.307096774193549e-05,
+      "loss": 0.1204,
+      "step": 17197
+    },
+    {
+      "epoch": 0.275168,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.306935483870969e-05,
+      "loss": 0.1437,
+      "step": 17198
+    },
+    {
+      "epoch": 0.275184,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.306774193548389e-05,
+      "loss": 0.1518,
+      "step": 17199
+    },
+    {
+      "epoch": 0.2752,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.306612903225807e-05,
+      "loss": 0.1639,
+      "step": 17200
+    },
+    {
+      "epoch": 0.275216,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.306451612903226e-05,
+      "loss": 0.1758,
+      "step": 17201
+    },
+    {
+      "epoch": 0.275232,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.306290322580646e-05,
+      "loss": 0.1605,
+      "step": 17202
+    },
+    {
+      "epoch": 0.275248,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.306129032258064e-05,
+      "loss": 0.1814,
+      "step": 17203
+    },
+    {
+      "epoch": 0.275264,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.305967741935484e-05,
+      "loss": 0.2131,
+      "step": 17204
+    },
+    {
+      "epoch": 0.27528,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.305806451612903e-05,
+      "loss": 0.1596,
+      "step": 17205
+    },
+    {
+      "epoch": 0.275296,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.305645161290323e-05,
+      "loss": 0.1615,
+      "step": 17206
+    },
+    {
+      "epoch": 0.275312,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.305483870967741e-05,
+      "loss": 0.1727,
+      "step": 17207
+    },
+    {
+      "epoch": 0.275328,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.305322580645161e-05,
+      "loss": 0.1579,
+      "step": 17208
+    },
+    {
+      "epoch": 0.275344,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.305161290322581e-05,
+      "loss": 0.1942,
+      "step": 17209
+    },
+    {
+      "epoch": 0.27536,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.305000000000001e-05,
+      "loss": 0.1579,
+      "step": 17210
+    },
+    {
+      "epoch": 0.275376,
+      "grad_norm": 0.75,
+      "learning_rate": 7.30483870967742e-05,
+      "loss": 0.1347,
+      "step": 17211
+    },
+    {
+      "epoch": 0.275392,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.30467741935484e-05,
+      "loss": 0.1774,
+      "step": 17212
+    },
+    {
+      "epoch": 0.275408,
+      "grad_norm": 0.482421875,
+      "learning_rate": 7.304516129032259e-05,
+      "loss": 0.1562,
+      "step": 17213
+    },
+    {
+      "epoch": 0.275424,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.304354838709679e-05,
+      "loss": 0.1817,
+      "step": 17214
+    },
+    {
+      "epoch": 0.27544,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.304193548387097e-05,
+      "loss": 0.2052,
+      "step": 17215
+    },
+    {
+      "epoch": 0.275456,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.304032258064516e-05,
+      "loss": 0.1795,
+      "step": 17216
+    },
+    {
+      "epoch": 0.275472,
+      "grad_norm": 1.328125,
+      "learning_rate": 7.303870967741936e-05,
+      "loss": 0.1453,
+      "step": 17217
+    },
+    {
+      "epoch": 0.275488,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.303709677419354e-05,
+      "loss": 0.1549,
+      "step": 17218
+    },
+    {
+      "epoch": 0.275504,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.303548387096774e-05,
+      "loss": 0.1938,
+      "step": 17219
+    },
+    {
+      "epoch": 0.27552,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.303387096774193e-05,
+      "loss": 0.154,
+      "step": 17220
+    },
+    {
+      "epoch": 0.275536,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.303225806451613e-05,
+      "loss": 0.1895,
+      "step": 17221
+    },
+    {
+      "epoch": 0.275552,
+      "grad_norm": 1.5,
+      "learning_rate": 7.303064516129033e-05,
+      "loss": 0.204,
+      "step": 17222
+    },
+    {
+      "epoch": 0.275568,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.302903225806453e-05,
+      "loss": 0.1468,
+      "step": 17223
+    },
+    {
+      "epoch": 0.275584,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.302741935483871e-05,
+      "loss": 0.1821,
+      "step": 17224
+    },
+    {
+      "epoch": 0.2756,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.302580645161291e-05,
+      "loss": 0.1774,
+      "step": 17225
+    },
+    {
+      "epoch": 0.275616,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.30241935483871e-05,
+      "loss": 0.1948,
+      "step": 17226
+    },
+    {
+      "epoch": 0.275632,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.30225806451613e-05,
+      "loss": 0.187,
+      "step": 17227
+    },
+    {
+      "epoch": 0.275648,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.302096774193549e-05,
+      "loss": 0.2195,
+      "step": 17228
+    },
+    {
+      "epoch": 0.275664,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.301935483870968e-05,
+      "loss": 0.147,
+      "step": 17229
+    },
+    {
+      "epoch": 0.27568,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.301774193548387e-05,
+      "loss": 0.1516,
+      "step": 17230
+    },
+    {
+      "epoch": 0.275696,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.301612903225806e-05,
+      "loss": 0.1643,
+      "step": 17231
+    },
+    {
+      "epoch": 0.275712,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.301451612903226e-05,
+      "loss": 0.1751,
+      "step": 17232
+    },
+    {
+      "epoch": 0.275728,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.301290322580646e-05,
+      "loss": 0.1774,
+      "step": 17233
+    },
+    {
+      "epoch": 0.275744,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.301129032258066e-05,
+      "loss": 0.184,
+      "step": 17234
+    },
+    {
+      "epoch": 0.27576,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.300967741935484e-05,
+      "loss": 0.1591,
+      "step": 17235
+    },
+    {
+      "epoch": 0.275776,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.300806451612904e-05,
+      "loss": 0.1682,
+      "step": 17236
+    },
+    {
+      "epoch": 0.275792,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.300645161290323e-05,
+      "loss": 0.1298,
+      "step": 17237
+    },
+    {
+      "epoch": 0.275808,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.300483870967743e-05,
+      "loss": 0.2038,
+      "step": 17238
+    },
+    {
+      "epoch": 0.275824,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.300322580645161e-05,
+      "loss": 0.1776,
+      "step": 17239
+    },
+    {
+      "epoch": 0.27584,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.300161290322581e-05,
+      "loss": 0.1485,
+      "step": 17240
+    },
+    {
+      "epoch": 0.275856,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.3e-05,
+      "loss": 0.2076,
+      "step": 17241
+    },
+    {
+      "epoch": 0.275872,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.29983870967742e-05,
+      "loss": 0.2094,
+      "step": 17242
+    },
+    {
+      "epoch": 0.275888,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.299677419354838e-05,
+      "loss": 0.164,
+      "step": 17243
+    },
+    {
+      "epoch": 0.275904,
+      "grad_norm": 1.3125,
+      "learning_rate": 7.299516129032258e-05,
+      "loss": 0.2004,
+      "step": 17244
+    },
+    {
+      "epoch": 0.27592,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.299354838709678e-05,
+      "loss": 0.1594,
+      "step": 17245
+    },
+    {
+      "epoch": 0.275936,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.299193548387097e-05,
+      "loss": 0.2049,
+      "step": 17246
+    },
+    {
+      "epoch": 0.275952,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.299032258064517e-05,
+      "loss": 0.1532,
+      "step": 17247
+    },
+    {
+      "epoch": 0.275968,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.298870967741936e-05,
+      "loss": 0.1668,
+      "step": 17248
+    },
+    {
+      "epoch": 0.275984,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.298709677419356e-05,
+      "loss": 0.2175,
+      "step": 17249
+    },
+    {
+      "epoch": 0.276,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.298548387096774e-05,
+      "loss": 0.1642,
+      "step": 17250
+    },
+    {
+      "epoch": 0.276016,
+      "grad_norm": 1.2734375,
+      "learning_rate": 7.298387096774194e-05,
+      "loss": 0.2048,
+      "step": 17251
+    },
+    {
+      "epoch": 0.276032,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.298225806451613e-05,
+      "loss": 0.1885,
+      "step": 17252
+    },
+    {
+      "epoch": 0.276048,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.298064516129033e-05,
+      "loss": 0.1618,
+      "step": 17253
+    },
+    {
+      "epoch": 0.276064,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.297903225806451e-05,
+      "loss": 0.1973,
+      "step": 17254
+    },
+    {
+      "epoch": 0.27608,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.297741935483871e-05,
+      "loss": 0.1663,
+      "step": 17255
+    },
+    {
+      "epoch": 0.276096,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.29758064516129e-05,
+      "loss": 0.1973,
+      "step": 17256
+    },
+    {
+      "epoch": 0.276112,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.29741935483871e-05,
+      "loss": 0.1873,
+      "step": 17257
+    },
+    {
+      "epoch": 0.276128,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.29725806451613e-05,
+      "loss": 0.1778,
+      "step": 17258
+    },
+    {
+      "epoch": 0.276144,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.29709677419355e-05,
+      "loss": 0.1596,
+      "step": 17259
+    },
+    {
+      "epoch": 0.27616,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.296935483870968e-05,
+      "loss": 0.1793,
+      "step": 17260
+    },
+    {
+      "epoch": 0.276176,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.296774193548388e-05,
+      "loss": 0.1598,
+      "step": 17261
+    },
+    {
+      "epoch": 0.276192,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.296612903225807e-05,
+      "loss": 0.1813,
+      "step": 17262
+    },
+    {
+      "epoch": 0.276208,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.296451612903226e-05,
+      "loss": 0.1619,
+      "step": 17263
+    },
+    {
+      "epoch": 0.276224,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.296290322580645e-05,
+      "loss": 0.136,
+      "step": 17264
+    },
+    {
+      "epoch": 0.27624,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.296129032258064e-05,
+      "loss": 0.1783,
+      "step": 17265
+    },
+    {
+      "epoch": 0.276256,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.295967741935484e-05,
+      "loss": 0.1541,
+      "step": 17266
+    },
+    {
+      "epoch": 0.276272,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.295806451612903e-05,
+      "loss": 0.1756,
+      "step": 17267
+    },
+    {
+      "epoch": 0.276288,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.295645161290323e-05,
+      "loss": 0.1817,
+      "step": 17268
+    },
+    {
+      "epoch": 0.276304,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.295483870967743e-05,
+      "loss": 0.1614,
+      "step": 17269
+    },
+    {
+      "epoch": 0.27632,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.295322580645163e-05,
+      "loss": 0.1985,
+      "step": 17270
+    },
+    {
+      "epoch": 0.276336,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.295161290322581e-05,
+      "loss": 0.2037,
+      "step": 17271
+    },
+    {
+      "epoch": 0.276352,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.295000000000001e-05,
+      "loss": 0.1392,
+      "step": 17272
+    },
+    {
+      "epoch": 0.276368,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.29483870967742e-05,
+      "loss": 0.1732,
+      "step": 17273
+    },
+    {
+      "epoch": 0.276384,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.29467741935484e-05,
+      "loss": 0.1855,
+      "step": 17274
+    },
+    {
+      "epoch": 0.2764,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.294516129032258e-05,
+      "loss": 0.1451,
+      "step": 17275
+    },
+    {
+      "epoch": 0.276416,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.294354838709678e-05,
+      "loss": 0.1732,
+      "step": 17276
+    },
+    {
+      "epoch": 0.276432,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.294193548387097e-05,
+      "loss": 0.171,
+      "step": 17277
+    },
+    {
+      "epoch": 0.276448,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.294032258064515e-05,
+      "loss": 0.1465,
+      "step": 17278
+    },
+    {
+      "epoch": 0.276464,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.293870967741935e-05,
+      "loss": 0.1833,
+      "step": 17279
+    },
+    {
+      "epoch": 0.27648,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.293709677419354e-05,
+      "loss": 0.1536,
+      "step": 17280
+    },
+    {
+      "epoch": 0.276496,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.293548387096774e-05,
+      "loss": 0.1449,
+      "step": 17281
+    },
+    {
+      "epoch": 0.276512,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.293387096774194e-05,
+      "loss": 0.1801,
+      "step": 17282
+    },
+    {
+      "epoch": 0.276528,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.293225806451614e-05,
+      "loss": 0.141,
+      "step": 17283
+    },
+    {
+      "epoch": 0.276544,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.293064516129033e-05,
+      "loss": 0.1791,
+      "step": 17284
+    },
+    {
+      "epoch": 0.27656,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.292903225806453e-05,
+      "loss": 0.1546,
+      "step": 17285
+    },
+    {
+      "epoch": 0.276576,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.292741935483871e-05,
+      "loss": 0.1877,
+      "step": 17286
+    },
+    {
+      "epoch": 0.276592,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.292580645161291e-05,
+      "loss": 0.1439,
+      "step": 17287
+    },
+    {
+      "epoch": 0.276608,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.29241935483871e-05,
+      "loss": 0.1411,
+      "step": 17288
+    },
+    {
+      "epoch": 0.276624,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.29225806451613e-05,
+      "loss": 0.1901,
+      "step": 17289
+    },
+    {
+      "epoch": 0.27664,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.292096774193548e-05,
+      "loss": 0.1589,
+      "step": 17290
+    },
+    {
+      "epoch": 0.276656,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.291935483870968e-05,
+      "loss": 0.1899,
+      "step": 17291
+    },
+    {
+      "epoch": 0.276672,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.291774193548387e-05,
+      "loss": 0.1673,
+      "step": 17292
+    },
+    {
+      "epoch": 0.276688,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.291612903225807e-05,
+      "loss": 0.1934,
+      "step": 17293
+    },
+    {
+      "epoch": 0.276704,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.291451612903227e-05,
+      "loss": 0.1408,
+      "step": 17294
+    },
+    {
+      "epoch": 0.27672,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.291290322580645e-05,
+      "loss": 0.1831,
+      "step": 17295
+    },
+    {
+      "epoch": 0.276736,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.291129032258065e-05,
+      "loss": 0.1517,
+      "step": 17296
+    },
+    {
+      "epoch": 0.276752,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.290967741935484e-05,
+      "loss": 0.1572,
+      "step": 17297
+    },
+    {
+      "epoch": 0.276768,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.290806451612904e-05,
+      "loss": 0.1692,
+      "step": 17298
+    },
+    {
+      "epoch": 0.276784,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.290645161290323e-05,
+      "loss": 0.1724,
+      "step": 17299
+    },
+    {
+      "epoch": 0.2768,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.290483870967742e-05,
+      "loss": 0.1537,
+      "step": 17300
+    },
+    {
+      "epoch": 0.276816,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.290322580645161e-05,
+      "loss": 0.1569,
+      "step": 17301
+    },
+    {
+      "epoch": 0.276832,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.290161290322581e-05,
+      "loss": 0.162,
+      "step": 17302
+    },
+    {
+      "epoch": 0.276848,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.29e-05,
+      "loss": 0.1734,
+      "step": 17303
+    },
+    {
+      "epoch": 0.276864,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.28983870967742e-05,
+      "loss": 0.1401,
+      "step": 17304
+    },
+    {
+      "epoch": 0.27688,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.28967741935484e-05,
+      "loss": 0.1927,
+      "step": 17305
+    },
+    {
+      "epoch": 0.276896,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.28951612903226e-05,
+      "loss": 0.145,
+      "step": 17306
+    },
+    {
+      "epoch": 0.276912,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.289354838709678e-05,
+      "loss": 0.1593,
+      "step": 17307
+    },
+    {
+      "epoch": 0.276928,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.289193548387098e-05,
+      "loss": 0.1444,
+      "step": 17308
+    },
+    {
+      "epoch": 0.276944,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.289032258064517e-05,
+      "loss": 0.1641,
+      "step": 17309
+    },
+    {
+      "epoch": 0.27696,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.288870967741935e-05,
+      "loss": 0.1593,
+      "step": 17310
+    },
+    {
+      "epoch": 0.276976,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.288709677419355e-05,
+      "loss": 0.1462,
+      "step": 17311
+    },
+    {
+      "epoch": 0.276992,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.288548387096774e-05,
+      "loss": 0.1431,
+      "step": 17312
+    },
+    {
+      "epoch": 0.277008,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.288387096774194e-05,
+      "loss": 0.1976,
+      "step": 17313
+    },
+    {
+      "epoch": 0.277024,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.288225806451612e-05,
+      "loss": 0.1851,
+      "step": 17314
+    },
+    {
+      "epoch": 0.27704,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.288064516129032e-05,
+      "loss": 0.1555,
+      "step": 17315
+    },
+    {
+      "epoch": 0.277056,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.287903225806451e-05,
+      "loss": 0.1423,
+      "step": 17316
+    },
+    {
+      "epoch": 0.277072,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.287741935483871e-05,
+      "loss": 0.1728,
+      "step": 17317
+    },
+    {
+      "epoch": 0.277088,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.287580645161291e-05,
+      "loss": 0.182,
+      "step": 17318
+    },
+    {
+      "epoch": 0.277104,
+      "grad_norm": 0.625,
+      "learning_rate": 7.287419354838711e-05,
+      "loss": 0.1739,
+      "step": 17319
+    },
+    {
+      "epoch": 0.27712,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.28725806451613e-05,
+      "loss": 0.1772,
+      "step": 17320
+    },
+    {
+      "epoch": 0.277136,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.28709677419355e-05,
+      "loss": 0.2037,
+      "step": 17321
+    },
+    {
+      "epoch": 0.277152,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.286935483870968e-05,
+      "loss": 0.1923,
+      "step": 17322
+    },
+    {
+      "epoch": 0.277168,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.286774193548388e-05,
+      "loss": 0.1634,
+      "step": 17323
+    },
+    {
+      "epoch": 0.277184,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.286612903225807e-05,
+      "loss": 0.1613,
+      "step": 17324
+    },
+    {
+      "epoch": 0.2772,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.286451612903225e-05,
+      "loss": 0.1587,
+      "step": 17325
+    },
+    {
+      "epoch": 0.277216,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.286290322580645e-05,
+      "loss": 0.1539,
+      "step": 17326
+    },
+    {
+      "epoch": 0.277232,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.286129032258064e-05,
+      "loss": 0.1775,
+      "step": 17327
+    },
+    {
+      "epoch": 0.277248,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.285967741935484e-05,
+      "loss": 0.1636,
+      "step": 17328
+    },
+    {
+      "epoch": 0.277264,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.285806451612904e-05,
+      "loss": 0.1432,
+      "step": 17329
+    },
+    {
+      "epoch": 0.27728,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.285645161290324e-05,
+      "loss": 0.18,
+      "step": 17330
+    },
+    {
+      "epoch": 0.277296,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.285483870967742e-05,
+      "loss": 0.2083,
+      "step": 17331
+    },
+    {
+      "epoch": 0.277312,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.285322580645162e-05,
+      "loss": 0.1916,
+      "step": 17332
+    },
+    {
+      "epoch": 0.277328,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.285161290322581e-05,
+      "loss": 0.1671,
+      "step": 17333
+    },
+    {
+      "epoch": 0.277344,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.285000000000001e-05,
+      "loss": 0.1993,
+      "step": 17334
+    },
+    {
+      "epoch": 0.27736,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.28483870967742e-05,
+      "loss": 0.1874,
+      "step": 17335
+    },
+    {
+      "epoch": 0.277376,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.28467741935484e-05,
+      "loss": 0.186,
+      "step": 17336
+    },
+    {
+      "epoch": 0.277392,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.284516129032258e-05,
+      "loss": 0.1669,
+      "step": 17337
+    },
+    {
+      "epoch": 0.277408,
+      "grad_norm": 1.25,
+      "learning_rate": 7.284354838709678e-05,
+      "loss": 0.224,
+      "step": 17338
+    },
+    {
+      "epoch": 0.277424,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.284193548387097e-05,
+      "loss": 0.1772,
+      "step": 17339
+    },
+    {
+      "epoch": 0.27744,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.284032258064517e-05,
+      "loss": 0.1758,
+      "step": 17340
+    },
+    {
+      "epoch": 0.277456,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.283870967741937e-05,
+      "loss": 0.1648,
+      "step": 17341
+    },
+    {
+      "epoch": 0.277472,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.283709677419355e-05,
+      "loss": 0.1995,
+      "step": 17342
+    },
+    {
+      "epoch": 0.277488,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.283548387096775e-05,
+      "loss": 0.1681,
+      "step": 17343
+    },
+    {
+      "epoch": 0.277504,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.283387096774194e-05,
+      "loss": 0.1694,
+      "step": 17344
+    },
+    {
+      "epoch": 0.27752,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.283225806451614e-05,
+      "loss": 0.2197,
+      "step": 17345
+    },
+    {
+      "epoch": 0.277536,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.283064516129032e-05,
+      "loss": 0.1929,
+      "step": 17346
+    },
+    {
+      "epoch": 0.277552,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.282903225806452e-05,
+      "loss": 0.181,
+      "step": 17347
+    },
+    {
+      "epoch": 0.277568,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.282741935483871e-05,
+      "loss": 0.1527,
+      "step": 17348
+    },
+    {
+      "epoch": 0.277584,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.282580645161291e-05,
+      "loss": 0.1867,
+      "step": 17349
+    },
+    {
+      "epoch": 0.2776,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.28241935483871e-05,
+      "loss": 0.1682,
+      "step": 17350
+    },
+    {
+      "epoch": 0.277616,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.28225806451613e-05,
+      "loss": 0.1644,
+      "step": 17351
+    },
+    {
+      "epoch": 0.277632,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.282096774193548e-05,
+      "loss": 0.2049,
+      "step": 17352
+    },
+    {
+      "epoch": 0.277648,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.281935483870968e-05,
+      "loss": 0.1546,
+      "step": 17353
+    },
+    {
+      "epoch": 0.277664,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.281774193548388e-05,
+      "loss": 0.1623,
+      "step": 17354
+    },
+    {
+      "epoch": 0.27768,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.281612903225807e-05,
+      "loss": 0.156,
+      "step": 17355
+    },
+    {
+      "epoch": 0.277696,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.281451612903227e-05,
+      "loss": 0.1511,
+      "step": 17356
+    },
+    {
+      "epoch": 0.277712,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.281290322580645e-05,
+      "loss": 0.1456,
+      "step": 17357
+    },
+    {
+      "epoch": 0.277728,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.281129032258065e-05,
+      "loss": 0.125,
+      "step": 17358
+    },
+    {
+      "epoch": 0.277744,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.280967741935484e-05,
+      "loss": 0.1555,
+      "step": 17359
+    },
+    {
+      "epoch": 0.27776,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.280806451612904e-05,
+      "loss": 0.168,
+      "step": 17360
+    },
+    {
+      "epoch": 0.277776,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.280645161290322e-05,
+      "loss": 0.1636,
+      "step": 17361
+    },
+    {
+      "epoch": 0.277792,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.280483870967742e-05,
+      "loss": 0.1516,
+      "step": 17362
+    },
+    {
+      "epoch": 0.277808,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.280322580645161e-05,
+      "loss": 0.1774,
+      "step": 17363
+    },
+    {
+      "epoch": 0.277824,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.280161290322581e-05,
+      "loss": 0.1647,
+      "step": 17364
+    },
+    {
+      "epoch": 0.27784,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.280000000000001e-05,
+      "loss": 0.1777,
+      "step": 17365
+    },
+    {
+      "epoch": 0.277856,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.279838709677421e-05,
+      "loss": 0.1557,
+      "step": 17366
+    },
+    {
+      "epoch": 0.277872,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.27967741935484e-05,
+      "loss": 0.1904,
+      "step": 17367
+    },
+    {
+      "epoch": 0.277888,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.279516129032259e-05,
+      "loss": 0.1538,
+      "step": 17368
+    },
+    {
+      "epoch": 0.277904,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.279354838709678e-05,
+      "loss": 0.1829,
+      "step": 17369
+    },
+    {
+      "epoch": 0.27792,
+      "grad_norm": 1.0,
+      "learning_rate": 7.279193548387098e-05,
+      "loss": 0.1609,
+      "step": 17370
+    },
+    {
+      "epoch": 0.277936,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.279032258064516e-05,
+      "loss": 0.1843,
+      "step": 17371
+    },
+    {
+      "epoch": 0.277952,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.278870967741935e-05,
+      "loss": 0.1656,
+      "step": 17372
+    },
+    {
+      "epoch": 0.277968,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.278709677419355e-05,
+      "loss": 0.1952,
+      "step": 17373
+    },
+    {
+      "epoch": 0.277984,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.278548387096774e-05,
+      "loss": 0.1804,
+      "step": 17374
+    },
+    {
+      "epoch": 0.278,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.278387096774194e-05,
+      "loss": 0.1812,
+      "step": 17375
+    },
+    {
+      "epoch": 0.278016,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.278225806451612e-05,
+      "loss": 0.1996,
+      "step": 17376
+    },
+    {
+      "epoch": 0.278032,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.278064516129032e-05,
+      "loss": 0.2021,
+      "step": 17377
+    },
+    {
+      "epoch": 0.278048,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.277903225806452e-05,
+      "loss": 0.148,
+      "step": 17378
+    },
+    {
+      "epoch": 0.278064,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.277741935483872e-05,
+      "loss": 0.1784,
+      "step": 17379
+    },
+    {
+      "epoch": 0.27808,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.277580645161291e-05,
+      "loss": 0.1587,
+      "step": 17380
+    },
+    {
+      "epoch": 0.278096,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.277419354838711e-05,
+      "loss": 0.2162,
+      "step": 17381
+    },
+    {
+      "epoch": 0.278112,
+      "grad_norm": 1.296875,
+      "learning_rate": 7.277258064516129e-05,
+      "loss": 0.1846,
+      "step": 17382
+    },
+    {
+      "epoch": 0.278128,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.277096774193549e-05,
+      "loss": 0.1783,
+      "step": 17383
+    },
+    {
+      "epoch": 0.278144,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.276935483870968e-05,
+      "loss": 0.1538,
+      "step": 17384
+    },
+    {
+      "epoch": 0.27816,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.276774193548388e-05,
+      "loss": 0.1654,
+      "step": 17385
+    },
+    {
+      "epoch": 0.278176,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.276612903225806e-05,
+      "loss": 0.1585,
+      "step": 17386
+    },
+    {
+      "epoch": 0.278192,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.276451612903225e-05,
+      "loss": 0.1872,
+      "step": 17387
+    },
+    {
+      "epoch": 0.278208,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.276290322580645e-05,
+      "loss": 0.169,
+      "step": 17388
+    },
+    {
+      "epoch": 0.278224,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.276129032258065e-05,
+      "loss": 0.19,
+      "step": 17389
+    },
+    {
+      "epoch": 0.27824,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.275967741935485e-05,
+      "loss": 0.1791,
+      "step": 17390
+    },
+    {
+      "epoch": 0.278256,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.275806451612904e-05,
+      "loss": 0.1437,
+      "step": 17391
+    },
+    {
+      "epoch": 0.278272,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.275645161290324e-05,
+      "loss": 0.1644,
+      "step": 17392
+    },
+    {
+      "epoch": 0.278288,
+      "grad_norm": 0.875,
+      "learning_rate": 7.275483870967742e-05,
+      "loss": 0.1585,
+      "step": 17393
+    },
+    {
+      "epoch": 0.278304,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.275322580645162e-05,
+      "loss": 0.1732,
+      "step": 17394
+    },
+    {
+      "epoch": 0.27832,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.275161290322581e-05,
+      "loss": 0.1216,
+      "step": 17395
+    },
+    {
+      "epoch": 0.278336,
+      "grad_norm": 0.5078125,
+      "learning_rate": 7.275e-05,
+      "loss": 0.1398,
+      "step": 17396
+    },
+    {
+      "epoch": 0.278352,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.274838709677419e-05,
+      "loss": 0.1201,
+      "step": 17397
+    },
+    {
+      "epoch": 0.278368,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.274677419354839e-05,
+      "loss": 0.1759,
+      "step": 17398
+    },
+    {
+      "epoch": 0.278384,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.274516129032258e-05,
+      "loss": 0.1726,
+      "step": 17399
+    },
+    {
+      "epoch": 0.2784,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.274354838709678e-05,
+      "loss": 0.1436,
+      "step": 17400
+    },
+    {
+      "epoch": 0.278416,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.274193548387098e-05,
+      "loss": 0.2321,
+      "step": 17401
+    },
+    {
+      "epoch": 0.278432,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.274032258064516e-05,
+      "loss": 0.1969,
+      "step": 17402
+    },
+    {
+      "epoch": 0.278448,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.273870967741936e-05,
+      "loss": 0.1332,
+      "step": 17403
+    },
+    {
+      "epoch": 0.278464,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.273709677419355e-05,
+      "loss": 0.2605,
+      "step": 17404
+    },
+    {
+      "epoch": 0.27848,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.273548387096775e-05,
+      "loss": 0.1469,
+      "step": 17405
+    },
+    {
+      "epoch": 0.278496,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.273387096774194e-05,
+      "loss": 0.1831,
+      "step": 17406
+    },
+    {
+      "epoch": 0.278512,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.273225806451613e-05,
+      "loss": 0.1835,
+      "step": 17407
+    },
+    {
+      "epoch": 0.278528,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.273064516129032e-05,
+      "loss": 0.1623,
+      "step": 17408
+    },
+    {
+      "epoch": 0.278544,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.272903225806452e-05,
+      "loss": 0.1572,
+      "step": 17409
+    },
+    {
+      "epoch": 0.27856,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.27274193548387e-05,
+      "loss": 0.1359,
+      "step": 17410
+    },
+    {
+      "epoch": 0.278576,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.27258064516129e-05,
+      "loss": 0.0962,
+      "step": 17411
+    },
+    {
+      "epoch": 0.278592,
+      "grad_norm": 1.2734375,
+      "learning_rate": 7.272419354838709e-05,
+      "loss": 0.2104,
+      "step": 17412
+    },
+    {
+      "epoch": 0.278608,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.272258064516129e-05,
+      "loss": 0.1666,
+      "step": 17413
+    },
+    {
+      "epoch": 0.278624,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.272096774193549e-05,
+      "loss": 0.1551,
+      "step": 17414
+    },
+    {
+      "epoch": 0.27864,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.271935483870969e-05,
+      "loss": 0.2027,
+      "step": 17415
+    },
+    {
+      "epoch": 0.278656,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.271774193548388e-05,
+      "loss": 0.2039,
+      "step": 17416
+    },
+    {
+      "epoch": 0.278672,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.271612903225808e-05,
+      "loss": 0.1594,
+      "step": 17417
+    },
+    {
+      "epoch": 0.278688,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.271451612903226e-05,
+      "loss": 0.1684,
+      "step": 17418
+    },
+    {
+      "epoch": 0.278704,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.271290322580645e-05,
+      "loss": 0.1913,
+      "step": 17419
+    },
+    {
+      "epoch": 0.27872,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.271129032258065e-05,
+      "loss": 0.1487,
+      "step": 17420
+    },
+    {
+      "epoch": 0.278736,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.270967741935483e-05,
+      "loss": 0.1905,
+      "step": 17421
+    },
+    {
+      "epoch": 0.278752,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.270806451612903e-05,
+      "loss": 0.1402,
+      "step": 17422
+    },
+    {
+      "epoch": 0.278768,
+      "grad_norm": 1.3046875,
+      "learning_rate": 7.270645161290322e-05,
+      "loss": 0.23,
+      "step": 17423
+    },
+    {
+      "epoch": 0.278784,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.270483870967742e-05,
+      "loss": 0.1699,
+      "step": 17424
+    },
+    {
+      "epoch": 0.2788,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.270322580645162e-05,
+      "loss": 0.1664,
+      "step": 17425
+    },
+    {
+      "epoch": 0.278816,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.270161290322582e-05,
+      "loss": 0.1416,
+      "step": 17426
+    },
+    {
+      "epoch": 0.278832,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.27e-05,
+      "loss": 0.1409,
+      "step": 17427
+    },
+    {
+      "epoch": 0.278848,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.26983870967742e-05,
+      "loss": 0.1931,
+      "step": 17428
+    },
+    {
+      "epoch": 0.278864,
+      "grad_norm": 0.75,
+      "learning_rate": 7.269677419354839e-05,
+      "loss": 0.136,
+      "step": 17429
+    },
+    {
+      "epoch": 0.27888,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.269516129032259e-05,
+      "loss": 0.1893,
+      "step": 17430
+    },
+    {
+      "epoch": 0.278896,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.269354838709678e-05,
+      "loss": 0.1647,
+      "step": 17431
+    },
+    {
+      "epoch": 0.278912,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.269193548387098e-05,
+      "loss": 0.154,
+      "step": 17432
+    },
+    {
+      "epoch": 0.278928,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.269032258064516e-05,
+      "loss": 0.1726,
+      "step": 17433
+    },
+    {
+      "epoch": 0.278944,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.268870967741935e-05,
+      "loss": 0.1871,
+      "step": 17434
+    },
+    {
+      "epoch": 0.27896,
+      "grad_norm": 1.5234375,
+      "learning_rate": 7.268709677419355e-05,
+      "loss": 0.1508,
+      "step": 17435
+    },
+    {
+      "epoch": 0.278976,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.268548387096775e-05,
+      "loss": 0.1409,
+      "step": 17436
+    },
+    {
+      "epoch": 0.278992,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.268387096774193e-05,
+      "loss": 0.1716,
+      "step": 17437
+    },
+    {
+      "epoch": 0.279008,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.268225806451613e-05,
+      "loss": 0.1404,
+      "step": 17438
+    },
+    {
+      "epoch": 0.279024,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.268064516129033e-05,
+      "loss": 0.1743,
+      "step": 17439
+    },
+    {
+      "epoch": 0.27904,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.267903225806452e-05,
+      "loss": 0.1615,
+      "step": 17440
+    },
+    {
+      "epoch": 0.279056,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.267741935483872e-05,
+      "loss": 0.1463,
+      "step": 17441
+    },
+    {
+      "epoch": 0.279072,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.26758064516129e-05,
+      "loss": 0.1652,
+      "step": 17442
+    },
+    {
+      "epoch": 0.279088,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.26741935483871e-05,
+      "loss": 0.1701,
+      "step": 17443
+    },
+    {
+      "epoch": 0.279104,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.267258064516129e-05,
+      "loss": 0.1685,
+      "step": 17444
+    },
+    {
+      "epoch": 0.27912,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.267096774193549e-05,
+      "loss": 0.1659,
+      "step": 17445
+    },
+    {
+      "epoch": 0.279136,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.266935483870968e-05,
+      "loss": 0.1534,
+      "step": 17446
+    },
+    {
+      "epoch": 0.279152,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.266774193548388e-05,
+      "loss": 0.1466,
+      "step": 17447
+    },
+    {
+      "epoch": 0.279168,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.266612903225806e-05,
+      "loss": 0.1387,
+      "step": 17448
+    },
+    {
+      "epoch": 0.279184,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.266451612903226e-05,
+      "loss": 0.1782,
+      "step": 17449
+    },
+    {
+      "epoch": 0.2792,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.266290322580646e-05,
+      "loss": 0.1683,
+      "step": 17450
+    },
+    {
+      "epoch": 0.279216,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.266129032258065e-05,
+      "loss": 0.1894,
+      "step": 17451
+    },
+    {
+      "epoch": 0.279232,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.265967741935485e-05,
+      "loss": 0.1921,
+      "step": 17452
+    },
+    {
+      "epoch": 0.279248,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.265806451612903e-05,
+      "loss": 0.1414,
+      "step": 17453
+    },
+    {
+      "epoch": 0.279264,
+      "grad_norm": 1.7734375,
+      "learning_rate": 7.265645161290323e-05,
+      "loss": 0.1969,
+      "step": 17454
+    },
+    {
+      "epoch": 0.27928,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.265483870967742e-05,
+      "loss": 0.164,
+      "step": 17455
+    },
+    {
+      "epoch": 0.279296,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.265322580645162e-05,
+      "loss": 0.1534,
+      "step": 17456
+    },
+    {
+      "epoch": 0.279312,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.26516129032258e-05,
+      "loss": 0.1365,
+      "step": 17457
+    },
+    {
+      "epoch": 0.279328,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.265e-05,
+      "loss": 0.161,
+      "step": 17458
+    },
+    {
+      "epoch": 0.279344,
+      "grad_norm": 0.625,
+      "learning_rate": 7.264838709677419e-05,
+      "loss": 0.1527,
+      "step": 17459
+    },
+    {
+      "epoch": 0.27936,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.264677419354839e-05,
+      "loss": 0.1462,
+      "step": 17460
+    },
+    {
+      "epoch": 0.279376,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.264516129032259e-05,
+      "loss": 0.1756,
+      "step": 17461
+    },
+    {
+      "epoch": 0.279392,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.264354838709679e-05,
+      "loss": 0.1848,
+      "step": 17462
+    },
+    {
+      "epoch": 0.279408,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.264193548387098e-05,
+      "loss": 0.1915,
+      "step": 17463
+    },
+    {
+      "epoch": 0.279424,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.264032258064516e-05,
+      "loss": 0.169,
+      "step": 17464
+    },
+    {
+      "epoch": 0.27944,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.263870967741936e-05,
+      "loss": 0.1882,
+      "step": 17465
+    },
+    {
+      "epoch": 0.279456,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.263709677419355e-05,
+      "loss": 0.1409,
+      "step": 17466
+    },
+    {
+      "epoch": 0.279472,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.263548387096775e-05,
+      "loss": 0.1854,
+      "step": 17467
+    },
+    {
+      "epoch": 0.279488,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.263387096774193e-05,
+      "loss": 0.1706,
+      "step": 17468
+    },
+    {
+      "epoch": 0.279504,
+      "grad_norm": 1.0,
+      "learning_rate": 7.263225806451613e-05,
+      "loss": 0.1582,
+      "step": 17469
+    },
+    {
+      "epoch": 0.27952,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.263064516129032e-05,
+      "loss": 0.13,
+      "step": 17470
+    },
+    {
+      "epoch": 0.279536,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.262903225806452e-05,
+      "loss": 0.1621,
+      "step": 17471
+    },
+    {
+      "epoch": 0.279552,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.26274193548387e-05,
+      "loss": 0.131,
+      "step": 17472
+    },
+    {
+      "epoch": 0.279568,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.26258064516129e-05,
+      "loss": 0.1892,
+      "step": 17473
+    },
+    {
+      "epoch": 0.279584,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.26241935483871e-05,
+      "loss": 0.2122,
+      "step": 17474
+    },
+    {
+      "epoch": 0.2796,
+      "grad_norm": 1.2421875,
+      "learning_rate": 7.26225806451613e-05,
+      "loss": 0.1742,
+      "step": 17475
+    },
+    {
+      "epoch": 0.279616,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.262096774193549e-05,
+      "loss": 0.1962,
+      "step": 17476
+    },
+    {
+      "epoch": 0.279632,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.261935483870969e-05,
+      "loss": 0.2189,
+      "step": 17477
+    },
+    {
+      "epoch": 0.279648,
+      "grad_norm": 1.3671875,
+      "learning_rate": 7.261774193548387e-05,
+      "loss": 0.162,
+      "step": 17478
+    },
+    {
+      "epoch": 0.279664,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.261612903225807e-05,
+      "loss": 0.1509,
+      "step": 17479
+    },
+    {
+      "epoch": 0.27968,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.261451612903226e-05,
+      "loss": 0.2175,
+      "step": 17480
+    },
+    {
+      "epoch": 0.279696,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.261290322580645e-05,
+      "loss": 0.175,
+      "step": 17481
+    },
+    {
+      "epoch": 0.279712,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.261129032258065e-05,
+      "loss": 0.1608,
+      "step": 17482
+    },
+    {
+      "epoch": 0.279728,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.260967741935483e-05,
+      "loss": 0.1615,
+      "step": 17483
+    },
+    {
+      "epoch": 0.279744,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.260806451612903e-05,
+      "loss": 0.1651,
+      "step": 17484
+    },
+    {
+      "epoch": 0.27976,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.260645161290323e-05,
+      "loss": 0.1702,
+      "step": 17485
+    },
+    {
+      "epoch": 0.279776,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.260483870967743e-05,
+      "loss": 0.1643,
+      "step": 17486
+    },
+    {
+      "epoch": 0.279792,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.260322580645162e-05,
+      "loss": 0.1547,
+      "step": 17487
+    },
+    {
+      "epoch": 0.279808,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.260161290322582e-05,
+      "loss": 0.1497,
+      "step": 17488
+    },
+    {
+      "epoch": 0.279824,
+      "grad_norm": 1.28125,
+      "learning_rate": 7.26e-05,
+      "loss": 0.2584,
+      "step": 17489
+    },
+    {
+      "epoch": 0.27984,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.25983870967742e-05,
+      "loss": 0.2104,
+      "step": 17490
+    },
+    {
+      "epoch": 0.279856,
+      "grad_norm": 0.625,
+      "learning_rate": 7.259677419354839e-05,
+      "loss": 0.1807,
+      "step": 17491
+    },
+    {
+      "epoch": 0.279872,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.259516129032259e-05,
+      "loss": 0.1622,
+      "step": 17492
+    },
+    {
+      "epoch": 0.279888,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.259354838709677e-05,
+      "loss": 0.1355,
+      "step": 17493
+    },
+    {
+      "epoch": 0.279904,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.259193548387097e-05,
+      "loss": 0.1534,
+      "step": 17494
+    },
+    {
+      "epoch": 0.27992,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.259032258064516e-05,
+      "loss": 0.2233,
+      "step": 17495
+    },
+    {
+      "epoch": 0.279936,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.258870967741936e-05,
+      "loss": 0.1672,
+      "step": 17496
+    },
+    {
+      "epoch": 0.279952,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.258709677419356e-05,
+      "loss": 0.1478,
+      "step": 17497
+    },
+    {
+      "epoch": 0.279968,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.258548387096775e-05,
+      "loss": 0.1585,
+      "step": 17498
+    },
+    {
+      "epoch": 0.279984,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.258387096774194e-05,
+      "loss": 0.1606,
+      "step": 17499
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.258225806451613e-05,
+      "loss": 0.1421,
+      "step": 17500
+    },
+    {
+      "epoch": 0.280016,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.258064516129033e-05,
+      "loss": 0.1668,
+      "step": 17501
+    },
+    {
+      "epoch": 0.280032,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.257903225806452e-05,
+      "loss": 0.1744,
+      "step": 17502
+    },
+    {
+      "epoch": 0.280048,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.257741935483872e-05,
+      "loss": 0.1723,
+      "step": 17503
+    },
+    {
+      "epoch": 0.280064,
+      "grad_norm": 1.125,
+      "learning_rate": 7.25758064516129e-05,
+      "loss": 0.1866,
+      "step": 17504
+    },
+    {
+      "epoch": 0.28008,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.25741935483871e-05,
+      "loss": 0.1734,
+      "step": 17505
+    },
+    {
+      "epoch": 0.280096,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.257258064516129e-05,
+      "loss": 0.1622,
+      "step": 17506
+    },
+    {
+      "epoch": 0.280112,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.257096774193549e-05,
+      "loss": 0.134,
+      "step": 17507
+    },
+    {
+      "epoch": 0.280128,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.256935483870967e-05,
+      "loss": 0.195,
+      "step": 17508
+    },
+    {
+      "epoch": 0.280144,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.256774193548387e-05,
+      "loss": 0.1669,
+      "step": 17509
+    },
+    {
+      "epoch": 0.28016,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.256612903225807e-05,
+      "loss": 0.1866,
+      "step": 17510
+    },
+    {
+      "epoch": 0.280176,
+      "grad_norm": 1.3125,
+      "learning_rate": 7.256451612903226e-05,
+      "loss": 0.1365,
+      "step": 17511
+    },
+    {
+      "epoch": 0.280192,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.256290322580646e-05,
+      "loss": 0.1894,
+      "step": 17512
+    },
+    {
+      "epoch": 0.280208,
+      "grad_norm": 1.0,
+      "learning_rate": 7.256129032258064e-05,
+      "loss": 0.1936,
+      "step": 17513
+    },
+    {
+      "epoch": 0.280224,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.255967741935484e-05,
+      "loss": 0.1723,
+      "step": 17514
+    },
+    {
+      "epoch": 0.28024,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.255806451612903e-05,
+      "loss": 0.1502,
+      "step": 17515
+    },
+    {
+      "epoch": 0.280256,
+      "grad_norm": 0.75,
+      "learning_rate": 7.255645161290323e-05,
+      "loss": 0.161,
+      "step": 17516
+    },
+    {
+      "epoch": 0.280272,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.255483870967742e-05,
+      "loss": 0.1653,
+      "step": 17517
+    },
+    {
+      "epoch": 0.280288,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.255322580645162e-05,
+      "loss": 0.1894,
+      "step": 17518
+    },
+    {
+      "epoch": 0.280304,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.25516129032258e-05,
+      "loss": 0.1664,
+      "step": 17519
+    },
+    {
+      "epoch": 0.28032,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.255e-05,
+      "loss": 0.1439,
+      "step": 17520
+    },
+    {
+      "epoch": 0.280336,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.25483870967742e-05,
+      "loss": 0.1928,
+      "step": 17521
+    },
+    {
+      "epoch": 0.280352,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.25467741935484e-05,
+      "loss": 0.1991,
+      "step": 17522
+    },
+    {
+      "epoch": 0.280368,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.254516129032259e-05,
+      "loss": 0.1816,
+      "step": 17523
+    },
+    {
+      "epoch": 0.280384,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.254354838709679e-05,
+      "loss": 0.172,
+      "step": 17524
+    },
+    {
+      "epoch": 0.2804,
+      "grad_norm": 1.265625,
+      "learning_rate": 7.254193548387097e-05,
+      "loss": 0.16,
+      "step": 17525
+    },
+    {
+      "epoch": 0.280416,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.254032258064517e-05,
+      "loss": 0.1711,
+      "step": 17526
+    },
+    {
+      "epoch": 0.280432,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.253870967741936e-05,
+      "loss": 0.1825,
+      "step": 17527
+    },
+    {
+      "epoch": 0.280448,
+      "grad_norm": 1.328125,
+      "learning_rate": 7.253709677419354e-05,
+      "loss": 0.2051,
+      "step": 17528
+    },
+    {
+      "epoch": 0.280464,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.253548387096774e-05,
+      "loss": 0.1589,
+      "step": 17529
+    },
+    {
+      "epoch": 0.28048,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.253387096774193e-05,
+      "loss": 0.1676,
+      "step": 17530
+    },
+    {
+      "epoch": 0.280496,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.253225806451613e-05,
+      "loss": 0.1875,
+      "step": 17531
+    },
+    {
+      "epoch": 0.280512,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.253064516129032e-05,
+      "loss": 0.1577,
+      "step": 17532
+    },
+    {
+      "epoch": 0.280528,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.252903225806452e-05,
+      "loss": 0.2189,
+      "step": 17533
+    },
+    {
+      "epoch": 0.280544,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.252741935483872e-05,
+      "loss": 0.1664,
+      "step": 17534
+    },
+    {
+      "epoch": 0.28056,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.252580645161291e-05,
+      "loss": 0.1797,
+      "step": 17535
+    },
+    {
+      "epoch": 0.280576,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.25241935483871e-05,
+      "loss": 0.1935,
+      "step": 17536
+    },
+    {
+      "epoch": 0.280592,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.25225806451613e-05,
+      "loss": 0.1637,
+      "step": 17537
+    },
+    {
+      "epoch": 0.280608,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.252096774193549e-05,
+      "loss": 0.1996,
+      "step": 17538
+    },
+    {
+      "epoch": 0.280624,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.251935483870969e-05,
+      "loss": 0.1821,
+      "step": 17539
+    },
+    {
+      "epoch": 0.28064,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.251774193548387e-05,
+      "loss": 0.1783,
+      "step": 17540
+    },
+    {
+      "epoch": 0.280656,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.251612903225807e-05,
+      "loss": 0.2041,
+      "step": 17541
+    },
+    {
+      "epoch": 0.280672,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.251451612903226e-05,
+      "loss": 0.1678,
+      "step": 17542
+    },
+    {
+      "epoch": 0.280688,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.251290322580644e-05,
+      "loss": 0.1892,
+      "step": 17543
+    },
+    {
+      "epoch": 0.280704,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.251129032258064e-05,
+      "loss": 0.1827,
+      "step": 17544
+    },
+    {
+      "epoch": 0.28072,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.250967741935484e-05,
+      "loss": 0.1793,
+      "step": 17545
+    },
+    {
+      "epoch": 0.280736,
+      "grad_norm": 0.75,
+      "learning_rate": 7.250806451612904e-05,
+      "loss": 0.1639,
+      "step": 17546
+    },
+    {
+      "epoch": 0.280752,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.250645161290323e-05,
+      "loss": 0.1516,
+      "step": 17547
+    },
+    {
+      "epoch": 0.280768,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.250483870967743e-05,
+      "loss": 0.1706,
+      "step": 17548
+    },
+    {
+      "epoch": 0.280784,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.250322580645161e-05,
+      "loss": 0.1677,
+      "step": 17549
+    },
+    {
+      "epoch": 0.2808,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.250161290322581e-05,
+      "loss": 0.1685,
+      "step": 17550
+    },
+    {
+      "epoch": 0.280816,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.25e-05,
+      "loss": 0.1815,
+      "step": 17551
+    },
+    {
+      "epoch": 0.280832,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.24983870967742e-05,
+      "loss": 0.1335,
+      "step": 17552
+    },
+    {
+      "epoch": 0.280848,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.249677419354839e-05,
+      "loss": 0.1715,
+      "step": 17553
+    },
+    {
+      "epoch": 0.280864,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.249516129032259e-05,
+      "loss": 0.2009,
+      "step": 17554
+    },
+    {
+      "epoch": 0.28088,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.249354838709677e-05,
+      "loss": 0.1351,
+      "step": 17555
+    },
+    {
+      "epoch": 0.280896,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.249193548387097e-05,
+      "loss": 0.1589,
+      "step": 17556
+    },
+    {
+      "epoch": 0.280912,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.249032258064517e-05,
+      "loss": 0.1798,
+      "step": 17557
+    },
+    {
+      "epoch": 0.280928,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.248870967741936e-05,
+      "loss": 0.171,
+      "step": 17558
+    },
+    {
+      "epoch": 0.280944,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.248709677419356e-05,
+      "loss": 0.1477,
+      "step": 17559
+    },
+    {
+      "epoch": 0.28096,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.248548387096774e-05,
+      "loss": 0.1446,
+      "step": 17560
+    },
+    {
+      "epoch": 0.280976,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.248387096774194e-05,
+      "loss": 0.1883,
+      "step": 17561
+    },
+    {
+      "epoch": 0.280992,
+      "grad_norm": 0.455078125,
+      "learning_rate": 7.248225806451613e-05,
+      "loss": 0.1437,
+      "step": 17562
+    },
+    {
+      "epoch": 0.281008,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.248064516129033e-05,
+      "loss": 0.2001,
+      "step": 17563
+    },
+    {
+      "epoch": 0.281024,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.247903225806451e-05,
+      "loss": 0.1545,
+      "step": 17564
+    },
+    {
+      "epoch": 0.28104,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.247741935483871e-05,
+      "loss": 0.1611,
+      "step": 17565
+    },
+    {
+      "epoch": 0.281056,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.24758064516129e-05,
+      "loss": 0.1327,
+      "step": 17566
+    },
+    {
+      "epoch": 0.281072,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.24741935483871e-05,
+      "loss": 0.1615,
+      "step": 17567
+    },
+    {
+      "epoch": 0.281088,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.247258064516129e-05,
+      "loss": 0.1538,
+      "step": 17568
+    },
+    {
+      "epoch": 0.281104,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.247096774193549e-05,
+      "loss": 0.2148,
+      "step": 17569
+    },
+    {
+      "epoch": 0.28112,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.246935483870968e-05,
+      "loss": 0.1522,
+      "step": 17570
+    },
+    {
+      "epoch": 0.281136,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.246774193548388e-05,
+      "loss": 0.1674,
+      "step": 17571
+    },
+    {
+      "epoch": 0.281152,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.246612903225807e-05,
+      "loss": 0.1243,
+      "step": 17572
+    },
+    {
+      "epoch": 0.281168,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.246451612903226e-05,
+      "loss": 0.1715,
+      "step": 17573
+    },
+    {
+      "epoch": 0.281184,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.246290322580646e-05,
+      "loss": 0.17,
+      "step": 17574
+    },
+    {
+      "epoch": 0.2812,
+      "grad_norm": 1.2578125,
+      "learning_rate": 7.246129032258064e-05,
+      "loss": 0.1779,
+      "step": 17575
+    },
+    {
+      "epoch": 0.281216,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.245967741935484e-05,
+      "loss": 0.1507,
+      "step": 17576
+    },
+    {
+      "epoch": 0.281232,
+      "grad_norm": 1.40625,
+      "learning_rate": 7.245806451612903e-05,
+      "loss": 0.184,
+      "step": 17577
+    },
+    {
+      "epoch": 0.281248,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.245645161290323e-05,
+      "loss": 0.1703,
+      "step": 17578
+    },
+    {
+      "epoch": 0.281264,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.245483870967741e-05,
+      "loss": 0.1441,
+      "step": 17579
+    },
+    {
+      "epoch": 0.28128,
+      "grad_norm": 0.75,
+      "learning_rate": 7.245322580645161e-05,
+      "loss": 0.1835,
+      "step": 17580
+    },
+    {
+      "epoch": 0.281296,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.245161290322581e-05,
+      "loss": 0.1482,
+      "step": 17581
+    },
+    {
+      "epoch": 0.281312,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.245000000000001e-05,
+      "loss": 0.162,
+      "step": 17582
+    },
+    {
+      "epoch": 0.281328,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.24483870967742e-05,
+      "loss": 0.1979,
+      "step": 17583
+    },
+    {
+      "epoch": 0.281344,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.24467741935484e-05,
+      "loss": 0.1825,
+      "step": 17584
+    },
+    {
+      "epoch": 0.28136,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.244516129032258e-05,
+      "loss": 0.1954,
+      "step": 17585
+    },
+    {
+      "epoch": 0.281376,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.244354838709678e-05,
+      "loss": 0.1896,
+      "step": 17586
+    },
+    {
+      "epoch": 0.281392,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.244193548387097e-05,
+      "loss": 0.1278,
+      "step": 17587
+    },
+    {
+      "epoch": 0.281408,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.244032258064517e-05,
+      "loss": 0.153,
+      "step": 17588
+    },
+    {
+      "epoch": 0.281424,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.243870967741936e-05,
+      "loss": 0.179,
+      "step": 17589
+    },
+    {
+      "epoch": 0.28144,
+      "grad_norm": 1.3515625,
+      "learning_rate": 7.243709677419354e-05,
+      "loss": 0.2282,
+      "step": 17590
+    },
+    {
+      "epoch": 0.281456,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.243548387096774e-05,
+      "loss": 0.1597,
+      "step": 17591
+    },
+    {
+      "epoch": 0.281472,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.243387096774194e-05,
+      "loss": 0.1791,
+      "step": 17592
+    },
+    {
+      "epoch": 0.281488,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.243225806451614e-05,
+      "loss": 0.2075,
+      "step": 17593
+    },
+    {
+      "epoch": 0.281504,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.243064516129033e-05,
+      "loss": 0.163,
+      "step": 17594
+    },
+    {
+      "epoch": 0.28152,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.242903225806453e-05,
+      "loss": 0.1231,
+      "step": 17595
+    },
+    {
+      "epoch": 0.281536,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.242741935483871e-05,
+      "loss": 0.1784,
+      "step": 17596
+    },
+    {
+      "epoch": 0.281552,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.242580645161291e-05,
+      "loss": 0.225,
+      "step": 17597
+    },
+    {
+      "epoch": 0.281568,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.24241935483871e-05,
+      "loss": 0.1421,
+      "step": 17598
+    },
+    {
+      "epoch": 0.281584,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.24225806451613e-05,
+      "loss": 0.217,
+      "step": 17599
+    },
+    {
+      "epoch": 0.2816,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.242096774193548e-05,
+      "loss": 0.188,
+      "step": 17600
+    },
+    {
+      "epoch": 0.281616,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.241935483870968e-05,
+      "loss": 0.1632,
+      "step": 17601
+    },
+    {
+      "epoch": 0.281632,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.241774193548387e-05,
+      "loss": 0.2109,
+      "step": 17602
+    },
+    {
+      "epoch": 0.281648,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.241612903225807e-05,
+      "loss": 0.1541,
+      "step": 17603
+    },
+    {
+      "epoch": 0.281664,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.241451612903226e-05,
+      "loss": 0.1692,
+      "step": 17604
+    },
+    {
+      "epoch": 0.28168,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.241290322580646e-05,
+      "loss": 0.2225,
+      "step": 17605
+    },
+    {
+      "epoch": 0.281696,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.241129032258065e-05,
+      "loss": 0.1452,
+      "step": 17606
+    },
+    {
+      "epoch": 0.281712,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.240967741935484e-05,
+      "loss": 0.1857,
+      "step": 17607
+    },
+    {
+      "epoch": 0.281728,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.240806451612904e-05,
+      "loss": 0.1676,
+      "step": 17608
+    },
+    {
+      "epoch": 0.281744,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.240645161290323e-05,
+      "loss": 0.2206,
+      "step": 17609
+    },
+    {
+      "epoch": 0.28176,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.240483870967743e-05,
+      "loss": 0.1561,
+      "step": 17610
+    },
+    {
+      "epoch": 0.281776,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.240322580645161e-05,
+      "loss": 0.1693,
+      "step": 17611
+    },
+    {
+      "epoch": 0.281792,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.240161290322581e-05,
+      "loss": 0.1989,
+      "step": 17612
+    },
+    {
+      "epoch": 0.281808,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.24e-05,
+      "loss": 0.1621,
+      "step": 17613
+    },
+    {
+      "epoch": 0.281824,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.23983870967742e-05,
+      "loss": 0.2,
+      "step": 17614
+    },
+    {
+      "epoch": 0.28184,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.239677419354838e-05,
+      "loss": 0.1759,
+      "step": 17615
+    },
+    {
+      "epoch": 0.281856,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.239516129032258e-05,
+      "loss": 0.1574,
+      "step": 17616
+    },
+    {
+      "epoch": 0.281872,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.239354838709678e-05,
+      "loss": 0.2171,
+      "step": 17617
+    },
+    {
+      "epoch": 0.281888,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.239193548387098e-05,
+      "loss": 0.16,
+      "step": 17618
+    },
+    {
+      "epoch": 0.281904,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.239032258064517e-05,
+      "loss": 0.196,
+      "step": 17619
+    },
+    {
+      "epoch": 0.28192,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.238870967741935e-05,
+      "loss": 0.1617,
+      "step": 17620
+    },
+    {
+      "epoch": 0.281936,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.238709677419355e-05,
+      "loss": 0.1843,
+      "step": 17621
+    },
+    {
+      "epoch": 0.281952,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.238548387096774e-05,
+      "loss": 0.1637,
+      "step": 17622
+    },
+    {
+      "epoch": 0.281968,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.238387096774194e-05,
+      "loss": 0.2031,
+      "step": 17623
+    },
+    {
+      "epoch": 0.281984,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.238225806451613e-05,
+      "loss": 0.1677,
+      "step": 17624
+    },
+    {
+      "epoch": 0.282,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.238064516129033e-05,
+      "loss": 0.1584,
+      "step": 17625
+    },
+    {
+      "epoch": 0.282016,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.237903225806451e-05,
+      "loss": 0.1763,
+      "step": 17626
+    },
+    {
+      "epoch": 0.282032,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.237741935483871e-05,
+      "loss": 0.1926,
+      "step": 17627
+    },
+    {
+      "epoch": 0.282048,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.23758064516129e-05,
+      "loss": 0.1859,
+      "step": 17628
+    },
+    {
+      "epoch": 0.282064,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.23741935483871e-05,
+      "loss": 0.1983,
+      "step": 17629
+    },
+    {
+      "epoch": 0.28208,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.23725806451613e-05,
+      "loss": 0.1801,
+      "step": 17630
+    },
+    {
+      "epoch": 0.282096,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.23709677419355e-05,
+      "loss": 0.1513,
+      "step": 17631
+    },
+    {
+      "epoch": 0.282112,
+      "grad_norm": 1.25,
+      "learning_rate": 7.236935483870968e-05,
+      "loss": 0.1515,
+      "step": 17632
+    },
+    {
+      "epoch": 0.282128,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.236774193548388e-05,
+      "loss": 0.2019,
+      "step": 17633
+    },
+    {
+      "epoch": 0.282144,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.236612903225807e-05,
+      "loss": 0.1728,
+      "step": 17634
+    },
+    {
+      "epoch": 0.28216,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.236451612903227e-05,
+      "loss": 0.1482,
+      "step": 17635
+    },
+    {
+      "epoch": 0.282176,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.236290322580645e-05,
+      "loss": 0.1577,
+      "step": 17636
+    },
+    {
+      "epoch": 0.282192,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.236129032258064e-05,
+      "loss": 0.1413,
+      "step": 17637
+    },
+    {
+      "epoch": 0.282208,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.235967741935484e-05,
+      "loss": 0.1738,
+      "step": 17638
+    },
+    {
+      "epoch": 0.282224,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.235806451612903e-05,
+      "loss": 0.1643,
+      "step": 17639
+    },
+    {
+      "epoch": 0.28224,
+      "grad_norm": 0.625,
+      "learning_rate": 7.235645161290323e-05,
+      "loss": 0.1678,
+      "step": 17640
+    },
+    {
+      "epoch": 0.282256,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.235483870967742e-05,
+      "loss": 0.1652,
+      "step": 17641
+    },
+    {
+      "epoch": 0.282272,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.235322580645162e-05,
+      "loss": 0.1698,
+      "step": 17642
+    },
+    {
+      "epoch": 0.282288,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.235161290322581e-05,
+      "loss": 0.1506,
+      "step": 17643
+    },
+    {
+      "epoch": 0.282304,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.235000000000001e-05,
+      "loss": 0.1607,
+      "step": 17644
+    },
+    {
+      "epoch": 0.28232,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.23483870967742e-05,
+      "loss": 0.1763,
+      "step": 17645
+    },
+    {
+      "epoch": 0.282336,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.23467741935484e-05,
+      "loss": 0.1392,
+      "step": 17646
+    },
+    {
+      "epoch": 0.282352,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.234516129032258e-05,
+      "loss": 0.1819,
+      "step": 17647
+    },
+    {
+      "epoch": 0.282368,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.234354838709678e-05,
+      "loss": 0.1926,
+      "step": 17648
+    },
+    {
+      "epoch": 0.282384,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.234193548387097e-05,
+      "loss": 0.167,
+      "step": 17649
+    },
+    {
+      "epoch": 0.2824,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.234032258064517e-05,
+      "loss": 0.2008,
+      "step": 17650
+    },
+    {
+      "epoch": 0.282416,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.233870967741935e-05,
+      "loss": 0.1775,
+      "step": 17651
+    },
+    {
+      "epoch": 0.282432,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.233709677419355e-05,
+      "loss": 0.1523,
+      "step": 17652
+    },
+    {
+      "epoch": 0.282448,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.233548387096775e-05,
+      "loss": 0.1872,
+      "step": 17653
+    },
+    {
+      "epoch": 0.282464,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.233387096774194e-05,
+      "loss": 0.1459,
+      "step": 17654
+    },
+    {
+      "epoch": 0.28248,
+      "grad_norm": 0.75,
+      "learning_rate": 7.233225806451614e-05,
+      "loss": 0.1546,
+      "step": 17655
+    },
+    {
+      "epoch": 0.282496,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.233064516129032e-05,
+      "loss": 0.1607,
+      "step": 17656
+    },
+    {
+      "epoch": 0.282512,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.232903225806452e-05,
+      "loss": 0.1766,
+      "step": 17657
+    },
+    {
+      "epoch": 0.282528,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.232741935483871e-05,
+      "loss": 0.1626,
+      "step": 17658
+    },
+    {
+      "epoch": 0.282544,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.232580645161291e-05,
+      "loss": 0.154,
+      "step": 17659
+    },
+    {
+      "epoch": 0.28256,
+      "grad_norm": 0.419921875,
+      "learning_rate": 7.23241935483871e-05,
+      "loss": 0.1452,
+      "step": 17660
+    },
+    {
+      "epoch": 0.282576,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.23225806451613e-05,
+      "loss": 0.1693,
+      "step": 17661
+    },
+    {
+      "epoch": 0.282592,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.232096774193548e-05,
+      "loss": 0.1787,
+      "step": 17662
+    },
+    {
+      "epoch": 0.282608,
+      "grad_norm": 1.375,
+      "learning_rate": 7.231935483870968e-05,
+      "loss": 0.2303,
+      "step": 17663
+    },
+    {
+      "epoch": 0.282624,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.231774193548387e-05,
+      "loss": 0.1489,
+      "step": 17664
+    },
+    {
+      "epoch": 0.28264,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.231612903225807e-05,
+      "loss": 0.1536,
+      "step": 17665
+    },
+    {
+      "epoch": 0.282656,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.231451612903227e-05,
+      "loss": 0.1856,
+      "step": 17666
+    },
+    {
+      "epoch": 0.282672,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.231290322580645e-05,
+      "loss": 0.1512,
+      "step": 17667
+    },
+    {
+      "epoch": 0.282688,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.231129032258065e-05,
+      "loss": 0.1723,
+      "step": 17668
+    },
+    {
+      "epoch": 0.282704,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.230967741935484e-05,
+      "loss": 0.1849,
+      "step": 17669
+    },
+    {
+      "epoch": 0.28272,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.230806451612904e-05,
+      "loss": 0.1995,
+      "step": 17670
+    },
+    {
+      "epoch": 0.282736,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.230645161290322e-05,
+      "loss": 0.1639,
+      "step": 17671
+    },
+    {
+      "epoch": 0.282752,
+      "grad_norm": 1.125,
+      "learning_rate": 7.230483870967742e-05,
+      "loss": 0.1833,
+      "step": 17672
+    },
+    {
+      "epoch": 0.282768,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.230322580645161e-05,
+      "loss": 0.1854,
+      "step": 17673
+    },
+    {
+      "epoch": 0.282784,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.230161290322581e-05,
+      "loss": 0.2002,
+      "step": 17674
+    },
+    {
+      "epoch": 0.2828,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.23e-05,
+      "loss": 0.1765,
+      "step": 17675
+    },
+    {
+      "epoch": 0.282816,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.22983870967742e-05,
+      "loss": 0.1843,
+      "step": 17676
+    },
+    {
+      "epoch": 0.282832,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.22967741935484e-05,
+      "loss": 0.1797,
+      "step": 17677
+    },
+    {
+      "epoch": 0.282848,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.22951612903226e-05,
+      "loss": 0.144,
+      "step": 17678
+    },
+    {
+      "epoch": 0.282864,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.229354838709678e-05,
+      "loss": 0.1504,
+      "step": 17679
+    },
+    {
+      "epoch": 0.28288,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.229193548387098e-05,
+      "loss": 0.1685,
+      "step": 17680
+    },
+    {
+      "epoch": 0.282896,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.229032258064517e-05,
+      "loss": 0.1695,
+      "step": 17681
+    },
+    {
+      "epoch": 0.282912,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.228870967741935e-05,
+      "loss": 0.2144,
+      "step": 17682
+    },
+    {
+      "epoch": 0.282928,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.228709677419355e-05,
+      "loss": 0.1793,
+      "step": 17683
+    },
+    {
+      "epoch": 0.282944,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.228548387096774e-05,
+      "loss": 0.1875,
+      "step": 17684
+    },
+    {
+      "epoch": 0.28296,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.228387096774194e-05,
+      "loss": 0.1222,
+      "step": 17685
+    },
+    {
+      "epoch": 0.282976,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.228225806451612e-05,
+      "loss": 0.1565,
+      "step": 17686
+    },
+    {
+      "epoch": 0.282992,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.228064516129032e-05,
+      "loss": 0.1723,
+      "step": 17687
+    },
+    {
+      "epoch": 0.283008,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.227903225806452e-05,
+      "loss": 0.1591,
+      "step": 17688
+    },
+    {
+      "epoch": 0.283024,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.227741935483871e-05,
+      "loss": 0.2051,
+      "step": 17689
+    },
+    {
+      "epoch": 0.28304,
+      "grad_norm": 0.625,
+      "learning_rate": 7.227580645161291e-05,
+      "loss": 0.1406,
+      "step": 17690
+    },
+    {
+      "epoch": 0.283056,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.227419354838711e-05,
+      "loss": 0.1869,
+      "step": 17691
+    },
+    {
+      "epoch": 0.283072,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.22725806451613e-05,
+      "loss": 0.1774,
+      "step": 17692
+    },
+    {
+      "epoch": 0.283088,
+      "grad_norm": 0.47265625,
+      "learning_rate": 7.22709677419355e-05,
+      "loss": 0.137,
+      "step": 17693
+    },
+    {
+      "epoch": 0.283104,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.226935483870968e-05,
+      "loss": 0.1465,
+      "step": 17694
+    },
+    {
+      "epoch": 0.28312,
+      "grad_norm": 1.84375,
+      "learning_rate": 7.226774193548388e-05,
+      "loss": 0.1688,
+      "step": 17695
+    },
+    {
+      "epoch": 0.283136,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.226612903225807e-05,
+      "loss": 0.1504,
+      "step": 17696
+    },
+    {
+      "epoch": 0.283152,
+      "grad_norm": 1.2578125,
+      "learning_rate": 7.226451612903227e-05,
+      "loss": 0.1735,
+      "step": 17697
+    },
+    {
+      "epoch": 0.283168,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.226290322580645e-05,
+      "loss": 0.1271,
+      "step": 17698
+    },
+    {
+      "epoch": 0.283184,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.226129032258064e-05,
+      "loss": 0.1265,
+      "step": 17699
+    },
+    {
+      "epoch": 0.2832,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.225967741935484e-05,
+      "loss": 0.1403,
+      "step": 17700
+    },
+    {
+      "epoch": 0.283216,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.225806451612904e-05,
+      "loss": 0.1941,
+      "step": 17701
+    },
+    {
+      "epoch": 0.283232,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.225645161290324e-05,
+      "loss": 0.1942,
+      "step": 17702
+    },
+    {
+      "epoch": 0.283248,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.225483870967742e-05,
+      "loss": 0.1799,
+      "step": 17703
+    },
+    {
+      "epoch": 0.283264,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.225322580645162e-05,
+      "loss": 0.1957,
+      "step": 17704
+    },
+    {
+      "epoch": 0.28328,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.225161290322581e-05,
+      "loss": 0.2172,
+      "step": 17705
+    },
+    {
+      "epoch": 0.283296,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.225000000000001e-05,
+      "loss": 0.1957,
+      "step": 17706
+    },
+    {
+      "epoch": 0.283312,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.22483870967742e-05,
+      "loss": 0.1567,
+      "step": 17707
+    },
+    {
+      "epoch": 0.283328,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.22467741935484e-05,
+      "loss": 0.1382,
+      "step": 17708
+    },
+    {
+      "epoch": 0.283344,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.224516129032258e-05,
+      "loss": 0.1126,
+      "step": 17709
+    },
+    {
+      "epoch": 0.28336,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.224354838709678e-05,
+      "loss": 0.1609,
+      "step": 17710
+    },
+    {
+      "epoch": 0.283376,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.224193548387097e-05,
+      "loss": 0.1686,
+      "step": 17711
+    },
+    {
+      "epoch": 0.283392,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.224032258064517e-05,
+      "loss": 0.1722,
+      "step": 17712
+    },
+    {
+      "epoch": 0.283408,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.223870967741936e-05,
+      "loss": 0.1874,
+      "step": 17713
+    },
+    {
+      "epoch": 0.283424,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.223709677419355e-05,
+      "loss": 0.1898,
+      "step": 17714
+    },
+    {
+      "epoch": 0.28344,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.223548387096775e-05,
+      "loss": 0.1181,
+      "step": 17715
+    },
+    {
+      "epoch": 0.283456,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.223387096774194e-05,
+      "loss": 0.1499,
+      "step": 17716
+    },
+    {
+      "epoch": 0.283472,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.223225806451614e-05,
+      "loss": 0.1888,
+      "step": 17717
+    },
+    {
+      "epoch": 0.283488,
+      "grad_norm": 1.265625,
+      "learning_rate": 7.223064516129032e-05,
+      "loss": 0.1877,
+      "step": 17718
+    },
+    {
+      "epoch": 0.283504,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.222903225806452e-05,
+      "loss": 0.1545,
+      "step": 17719
+    },
+    {
+      "epoch": 0.28352,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.222741935483871e-05,
+      "loss": 0.1852,
+      "step": 17720
+    },
+    {
+      "epoch": 0.283536,
+      "grad_norm": 1.4765625,
+      "learning_rate": 7.222580645161291e-05,
+      "loss": 0.1481,
+      "step": 17721
+    },
+    {
+      "epoch": 0.283552,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.22241935483871e-05,
+      "loss": 0.1526,
+      "step": 17722
+    },
+    {
+      "epoch": 0.283568,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.222258064516129e-05,
+      "loss": 0.1809,
+      "step": 17723
+    },
+    {
+      "epoch": 0.283584,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.222096774193548e-05,
+      "loss": 0.1436,
+      "step": 17724
+    },
+    {
+      "epoch": 0.2836,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.221935483870968e-05,
+      "loss": 0.2084,
+      "step": 17725
+    },
+    {
+      "epoch": 0.283616,
+      "grad_norm": 1.265625,
+      "learning_rate": 7.221774193548388e-05,
+      "loss": 0.1543,
+      "step": 17726
+    },
+    {
+      "epoch": 0.283632,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.221612903225808e-05,
+      "loss": 0.1885,
+      "step": 17727
+    },
+    {
+      "epoch": 0.283648,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.221451612903226e-05,
+      "loss": 0.1525,
+      "step": 17728
+    },
+    {
+      "epoch": 0.283664,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.221290322580645e-05,
+      "loss": 0.1497,
+      "step": 17729
+    },
+    {
+      "epoch": 0.28368,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.221129032258065e-05,
+      "loss": 0.1835,
+      "step": 17730
+    },
+    {
+      "epoch": 0.283696,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.220967741935484e-05,
+      "loss": 0.1221,
+      "step": 17731
+    },
+    {
+      "epoch": 0.283712,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.220806451612904e-05,
+      "loss": 0.1742,
+      "step": 17732
+    },
+    {
+      "epoch": 0.283728,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.220645161290322e-05,
+      "loss": 0.2072,
+      "step": 17733
+    },
+    {
+      "epoch": 0.283744,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.220483870967742e-05,
+      "loss": 0.1824,
+      "step": 17734
+    },
+    {
+      "epoch": 0.28376,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.220322580645161e-05,
+      "loss": 0.1804,
+      "step": 17735
+    },
+    {
+      "epoch": 0.283776,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.220161290322581e-05,
+      "loss": 0.2154,
+      "step": 17736
+    },
+    {
+      "epoch": 0.283792,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.22e-05,
+      "loss": 0.1835,
+      "step": 17737
+    },
+    {
+      "epoch": 0.283808,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.21983870967742e-05,
+      "loss": 0.1469,
+      "step": 17738
+    },
+    {
+      "epoch": 0.283824,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.219677419354839e-05,
+      "loss": 0.1736,
+      "step": 17739
+    },
+    {
+      "epoch": 0.28384,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.219516129032259e-05,
+      "loss": 0.1156,
+      "step": 17740
+    },
+    {
+      "epoch": 0.283856,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.219354838709678e-05,
+      "loss": 0.1891,
+      "step": 17741
+    },
+    {
+      "epoch": 0.283872,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.219193548387098e-05,
+      "loss": 0.165,
+      "step": 17742
+    },
+    {
+      "epoch": 0.283888,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.219032258064516e-05,
+      "loss": 0.1763,
+      "step": 17743
+    },
+    {
+      "epoch": 0.283904,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.218870967741935e-05,
+      "loss": 0.1439,
+      "step": 17744
+    },
+    {
+      "epoch": 0.28392,
+      "grad_norm": 1.34375,
+      "learning_rate": 7.218709677419355e-05,
+      "loss": 0.173,
+      "step": 17745
+    },
+    {
+      "epoch": 0.283936,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.218548387096774e-05,
+      "loss": 0.1814,
+      "step": 17746
+    },
+    {
+      "epoch": 0.283952,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.218387096774194e-05,
+      "loss": 0.1482,
+      "step": 17747
+    },
+    {
+      "epoch": 0.283968,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.218225806451613e-05,
+      "loss": 0.1666,
+      "step": 17748
+    },
+    {
+      "epoch": 0.283984,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.218064516129033e-05,
+      "loss": 0.2102,
+      "step": 17749
+    },
+    {
+      "epoch": 0.284,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.217903225806452e-05,
+      "loss": 0.1965,
+      "step": 17750
+    },
+    {
+      "epoch": 0.284016,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.217741935483872e-05,
+      "loss": 0.1439,
+      "step": 17751
+    },
+    {
+      "epoch": 0.284032,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.21758064516129e-05,
+      "loss": 0.2127,
+      "step": 17752
+    },
+    {
+      "epoch": 0.284048,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.21741935483871e-05,
+      "loss": 0.1862,
+      "step": 17753
+    },
+    {
+      "epoch": 0.284064,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.217258064516129e-05,
+      "loss": 0.1638,
+      "step": 17754
+    },
+    {
+      "epoch": 0.28408,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.217096774193549e-05,
+      "loss": 0.1483,
+      "step": 17755
+    },
+    {
+      "epoch": 0.284096,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.216935483870968e-05,
+      "loss": 0.1637,
+      "step": 17756
+    },
+    {
+      "epoch": 0.284112,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.216774193548388e-05,
+      "loss": 0.1564,
+      "step": 17757
+    },
+    {
+      "epoch": 0.284128,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.216612903225806e-05,
+      "loss": 0.138,
+      "step": 17758
+    },
+    {
+      "epoch": 0.284144,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.216451612903226e-05,
+      "loss": 0.1514,
+      "step": 17759
+    },
+    {
+      "epoch": 0.28416,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.216290322580645e-05,
+      "loss": 0.1806,
+      "step": 17760
+    },
+    {
+      "epoch": 0.284176,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.216129032258065e-05,
+      "loss": 0.1689,
+      "step": 17761
+    },
+    {
+      "epoch": 0.284192,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.215967741935485e-05,
+      "loss": 0.1833,
+      "step": 17762
+    },
+    {
+      "epoch": 0.284208,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.215806451612903e-05,
+      "loss": 0.1562,
+      "step": 17763
+    },
+    {
+      "epoch": 0.284224,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.215645161290323e-05,
+      "loss": 0.1724,
+      "step": 17764
+    },
+    {
+      "epoch": 0.28424,
+      "grad_norm": 1.34375,
+      "learning_rate": 7.215483870967742e-05,
+      "loss": 0.1827,
+      "step": 17765
+    },
+    {
+      "epoch": 0.284256,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.215322580645162e-05,
+      "loss": 0.1552,
+      "step": 17766
+    },
+    {
+      "epoch": 0.284272,
+      "grad_norm": 1.375,
+      "learning_rate": 7.21516129032258e-05,
+      "loss": 0.153,
+      "step": 17767
+    },
+    {
+      "epoch": 0.284288,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.215e-05,
+      "loss": 0.1944,
+      "step": 17768
+    },
+    {
+      "epoch": 0.284304,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.214838709677419e-05,
+      "loss": 0.1994,
+      "step": 17769
+    },
+    {
+      "epoch": 0.28432,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.214677419354839e-05,
+      "loss": 0.1969,
+      "step": 17770
+    },
+    {
+      "epoch": 0.284336,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.214516129032258e-05,
+      "loss": 0.2373,
+      "step": 17771
+    },
+    {
+      "epoch": 0.284352,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.214354838709678e-05,
+      "loss": 0.1652,
+      "step": 17772
+    },
+    {
+      "epoch": 0.284368,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.214193548387098e-05,
+      "loss": 0.1626,
+      "step": 17773
+    },
+    {
+      "epoch": 0.284384,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.214032258064518e-05,
+      "loss": 0.1795,
+      "step": 17774
+    },
+    {
+      "epoch": 0.2844,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.213870967741936e-05,
+      "loss": 0.1625,
+      "step": 17775
+    },
+    {
+      "epoch": 0.284416,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.213709677419355e-05,
+      "loss": 0.1715,
+      "step": 17776
+    },
+    {
+      "epoch": 0.284432,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.213548387096775e-05,
+      "loss": 0.1522,
+      "step": 17777
+    },
+    {
+      "epoch": 0.284448,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.213387096774193e-05,
+      "loss": 0.174,
+      "step": 17778
+    },
+    {
+      "epoch": 0.284464,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.213225806451613e-05,
+      "loss": 0.178,
+      "step": 17779
+    },
+    {
+      "epoch": 0.28448,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.213064516129032e-05,
+      "loss": 0.1857,
+      "step": 17780
+    },
+    {
+      "epoch": 0.284496,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.212903225806452e-05,
+      "loss": 0.1804,
+      "step": 17781
+    },
+    {
+      "epoch": 0.284512,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.21274193548387e-05,
+      "loss": 0.1944,
+      "step": 17782
+    },
+    {
+      "epoch": 0.284528,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.21258064516129e-05,
+      "loss": 0.1724,
+      "step": 17783
+    },
+    {
+      "epoch": 0.284544,
+      "grad_norm": 1.0,
+      "learning_rate": 7.212419354838709e-05,
+      "loss": 0.2168,
+      "step": 17784
+    },
+    {
+      "epoch": 0.28456,
+      "grad_norm": 0.75,
+      "learning_rate": 7.212258064516129e-05,
+      "loss": 0.2059,
+      "step": 17785
+    },
+    {
+      "epoch": 0.284576,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.212096774193549e-05,
+      "loss": 0.1572,
+      "step": 17786
+    },
+    {
+      "epoch": 0.284592,
+      "grad_norm": 1.2109375,
+      "learning_rate": 7.211935483870969e-05,
+      "loss": 0.1547,
+      "step": 17787
+    },
+    {
+      "epoch": 0.284608,
+      "grad_norm": 0.875,
+      "learning_rate": 7.211774193548388e-05,
+      "loss": 0.1626,
+      "step": 17788
+    },
+    {
+      "epoch": 0.284624,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.211612903225808e-05,
+      "loss": 0.1741,
+      "step": 17789
+    },
+    {
+      "epoch": 0.28464,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.211451612903226e-05,
+      "loss": 0.1471,
+      "step": 17790
+    },
+    {
+      "epoch": 0.284656,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.211290322580645e-05,
+      "loss": 0.1675,
+      "step": 17791
+    },
+    {
+      "epoch": 0.284672,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.211129032258065e-05,
+      "loss": 0.1355,
+      "step": 17792
+    },
+    {
+      "epoch": 0.284688,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.210967741935483e-05,
+      "loss": 0.172,
+      "step": 17793
+    },
+    {
+      "epoch": 0.284704,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.210806451612903e-05,
+      "loss": 0.1667,
+      "step": 17794
+    },
+    {
+      "epoch": 0.28472,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.210645161290322e-05,
+      "loss": 0.1294,
+      "step": 17795
+    },
+    {
+      "epoch": 0.284736,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.210483870967742e-05,
+      "loss": 0.1767,
+      "step": 17796
+    },
+    {
+      "epoch": 0.284752,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.210322580645162e-05,
+      "loss": 0.166,
+      "step": 17797
+    },
+    {
+      "epoch": 0.284768,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.210161290322582e-05,
+      "loss": 0.1725,
+      "step": 17798
+    },
+    {
+      "epoch": 0.284784,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.21e-05,
+      "loss": 0.1704,
+      "step": 17799
+    },
+    {
+      "epoch": 0.2848,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.20983870967742e-05,
+      "loss": 0.1157,
+      "step": 17800
+    },
+    {
+      "epoch": 0.284816,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.209677419354839e-05,
+      "loss": 0.18,
+      "step": 17801
+    },
+    {
+      "epoch": 0.284832,
+      "grad_norm": 0.5,
+      "learning_rate": 7.209516129032259e-05,
+      "loss": 0.1689,
+      "step": 17802
+    },
+    {
+      "epoch": 0.284848,
+      "grad_norm": 0.45703125,
+      "learning_rate": 7.209354838709678e-05,
+      "loss": 0.1498,
+      "step": 17803
+    },
+    {
+      "epoch": 0.284864,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.209193548387098e-05,
+      "loss": 0.1938,
+      "step": 17804
+    },
+    {
+      "epoch": 0.28488,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.209032258064516e-05,
+      "loss": 0.1858,
+      "step": 17805
+    },
+    {
+      "epoch": 0.284896,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.208870967741936e-05,
+      "loss": 0.1509,
+      "step": 17806
+    },
+    {
+      "epoch": 0.284912,
+      "grad_norm": 0.486328125,
+      "learning_rate": 7.208709677419355e-05,
+      "loss": 0.1531,
+      "step": 17807
+    },
+    {
+      "epoch": 0.284928,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.208548387096775e-05,
+      "loss": 0.1582,
+      "step": 17808
+    },
+    {
+      "epoch": 0.284944,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.208387096774195e-05,
+      "loss": 0.1434,
+      "step": 17809
+    },
+    {
+      "epoch": 0.28496,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.208225806451613e-05,
+      "loss": 0.1699,
+      "step": 17810
+    },
+    {
+      "epoch": 0.284976,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.208064516129033e-05,
+      "loss": 0.1486,
+      "step": 17811
+    },
+    {
+      "epoch": 0.284992,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.207903225806452e-05,
+      "loss": 0.1759,
+      "step": 17812
+    },
+    {
+      "epoch": 0.285008,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.207741935483872e-05,
+      "loss": 0.1725,
+      "step": 17813
+    },
+    {
+      "epoch": 0.285024,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.20758064516129e-05,
+      "loss": 0.1669,
+      "step": 17814
+    },
+    {
+      "epoch": 0.28504,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.20741935483871e-05,
+      "loss": 0.199,
+      "step": 17815
+    },
+    {
+      "epoch": 0.285056,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.207258064516129e-05,
+      "loss": 0.1446,
+      "step": 17816
+    },
+    {
+      "epoch": 0.285072,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.207096774193549e-05,
+      "loss": 0.1519,
+      "step": 17817
+    },
+    {
+      "epoch": 0.285088,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.206935483870968e-05,
+      "loss": 0.1661,
+      "step": 17818
+    },
+    {
+      "epoch": 0.285104,
+      "grad_norm": 1.3984375,
+      "learning_rate": 7.206774193548387e-05,
+      "loss": 0.165,
+      "step": 17819
+    },
+    {
+      "epoch": 0.28512,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.206612903225806e-05,
+      "loss": 0.1565,
+      "step": 17820
+    },
+    {
+      "epoch": 0.285136,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.206451612903226e-05,
+      "loss": 0.1393,
+      "step": 17821
+    },
+    {
+      "epoch": 0.285152,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.206290322580646e-05,
+      "loss": 0.1572,
+      "step": 17822
+    },
+    {
+      "epoch": 0.285168,
+      "grad_norm": 1.328125,
+      "learning_rate": 7.206129032258065e-05,
+      "loss": 0.1355,
+      "step": 17823
+    },
+    {
+      "epoch": 0.285184,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.205967741935485e-05,
+      "loss": 0.1675,
+      "step": 17824
+    },
+    {
+      "epoch": 0.2852,
+      "grad_norm": 0.875,
+      "learning_rate": 7.205806451612903e-05,
+      "loss": 0.1383,
+      "step": 17825
+    },
+    {
+      "epoch": 0.285216,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.205645161290323e-05,
+      "loss": 0.1686,
+      "step": 17826
+    },
+    {
+      "epoch": 0.285232,
+      "grad_norm": 1.25,
+      "learning_rate": 7.205483870967742e-05,
+      "loss": 0.1708,
+      "step": 17827
+    },
+    {
+      "epoch": 0.285248,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.205322580645162e-05,
+      "loss": 0.1901,
+      "step": 17828
+    },
+    {
+      "epoch": 0.285264,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.20516129032258e-05,
+      "loss": 0.1761,
+      "step": 17829
+    },
+    {
+      "epoch": 0.28528,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.205e-05,
+      "loss": 0.1711,
+      "step": 17830
+    },
+    {
+      "epoch": 0.285296,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.204838709677419e-05,
+      "loss": 0.1572,
+      "step": 17831
+    },
+    {
+      "epoch": 0.285312,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.204677419354839e-05,
+      "loss": 0.2092,
+      "step": 17832
+    },
+    {
+      "epoch": 0.285328,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.204516129032259e-05,
+      "loss": 0.1682,
+      "step": 17833
+    },
+    {
+      "epoch": 0.285344,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.204354838709679e-05,
+      "loss": 0.136,
+      "step": 17834
+    },
+    {
+      "epoch": 0.28536,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.204193548387097e-05,
+      "loss": 0.1544,
+      "step": 17835
+    },
+    {
+      "epoch": 0.285376,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.204032258064517e-05,
+      "loss": 0.17,
+      "step": 17836
+    },
+    {
+      "epoch": 0.285392,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.203870967741936e-05,
+      "loss": 0.1549,
+      "step": 17837
+    },
+    {
+      "epoch": 0.285408,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.203709677419355e-05,
+      "loss": 0.1369,
+      "step": 17838
+    },
+    {
+      "epoch": 0.285424,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.203548387096775e-05,
+      "loss": 0.1552,
+      "step": 17839
+    },
+    {
+      "epoch": 0.28544,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.203387096774193e-05,
+      "loss": 0.1674,
+      "step": 17840
+    },
+    {
+      "epoch": 0.285456,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.203225806451613e-05,
+      "loss": 0.1735,
+      "step": 17841
+    },
+    {
+      "epoch": 0.285472,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.203064516129032e-05,
+      "loss": 0.2385,
+      "step": 17842
+    },
+    {
+      "epoch": 0.285488,
+      "grad_norm": 1.4140625,
+      "learning_rate": 7.202903225806452e-05,
+      "loss": 0.1895,
+      "step": 17843
+    },
+    {
+      "epoch": 0.285504,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.202741935483872e-05,
+      "loss": 0.1604,
+      "step": 17844
+    },
+    {
+      "epoch": 0.28552,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.202580645161292e-05,
+      "loss": 0.1357,
+      "step": 17845
+    },
+    {
+      "epoch": 0.285536,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.20241935483871e-05,
+      "loss": 0.1491,
+      "step": 17846
+    },
+    {
+      "epoch": 0.285552,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.20225806451613e-05,
+      "loss": 0.1753,
+      "step": 17847
+    },
+    {
+      "epoch": 0.285568,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.202096774193549e-05,
+      "loss": 0.1651,
+      "step": 17848
+    },
+    {
+      "epoch": 0.285584,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.201935483870969e-05,
+      "loss": 0.157,
+      "step": 17849
+    },
+    {
+      "epoch": 0.2856,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.201774193548387e-05,
+      "loss": 0.1657,
+      "step": 17850
+    },
+    {
+      "epoch": 0.285616,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.201612903225807e-05,
+      "loss": 0.2223,
+      "step": 17851
+    },
+    {
+      "epoch": 0.285632,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.201451612903226e-05,
+      "loss": 0.1681,
+      "step": 17852
+    },
+    {
+      "epoch": 0.285648,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.201290322580645e-05,
+      "loss": 0.1401,
+      "step": 17853
+    },
+    {
+      "epoch": 0.285664,
+      "grad_norm": 1.375,
+      "learning_rate": 7.201129032258065e-05,
+      "loss": 0.1708,
+      "step": 17854
+    },
+    {
+      "epoch": 0.28568,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.200967741935483e-05,
+      "loss": 0.1663,
+      "step": 17855
+    },
+    {
+      "epoch": 0.285696,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.200806451612903e-05,
+      "loss": 0.1723,
+      "step": 17856
+    },
+    {
+      "epoch": 0.285712,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.200645161290323e-05,
+      "loss": 0.1946,
+      "step": 17857
+    },
+    {
+      "epoch": 0.285728,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.200483870967743e-05,
+      "loss": 0.2052,
+      "step": 17858
+    },
+    {
+      "epoch": 0.285744,
+      "grad_norm": 0.498046875,
+      "learning_rate": 7.200322580645162e-05,
+      "loss": 0.1142,
+      "step": 17859
+    },
+    {
+      "epoch": 0.28576,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.200161290322582e-05,
+      "loss": 0.2164,
+      "step": 17860
+    },
+    {
+      "epoch": 0.285776,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.2e-05,
+      "loss": 0.1804,
+      "step": 17861
+    },
+    {
+      "epoch": 0.285792,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.19983870967742e-05,
+      "loss": 0.1858,
+      "step": 17862
+    },
+    {
+      "epoch": 0.285808,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.199677419354839e-05,
+      "loss": 0.1347,
+      "step": 17863
+    },
+    {
+      "epoch": 0.285824,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.199516129032259e-05,
+      "loss": 0.1746,
+      "step": 17864
+    },
+    {
+      "epoch": 0.28584,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.199354838709677e-05,
+      "loss": 0.2206,
+      "step": 17865
+    },
+    {
+      "epoch": 0.285856,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.199193548387097e-05,
+      "loss": 0.1448,
+      "step": 17866
+    },
+    {
+      "epoch": 0.285872,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.199032258064516e-05,
+      "loss": 0.1721,
+      "step": 17867
+    },
+    {
+      "epoch": 0.285888,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.198870967741936e-05,
+      "loss": 0.1754,
+      "step": 17868
+    },
+    {
+      "epoch": 0.285904,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.198709677419356e-05,
+      "loss": 0.1988,
+      "step": 17869
+    },
+    {
+      "epoch": 0.28592,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.198548387096774e-05,
+      "loss": 0.1898,
+      "step": 17870
+    },
+    {
+      "epoch": 0.285936,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.198387096774194e-05,
+      "loss": 0.1946,
+      "step": 17871
+    },
+    {
+      "epoch": 0.285952,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.198225806451613e-05,
+      "loss": 0.1667,
+      "step": 17872
+    },
+    {
+      "epoch": 0.285968,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.198064516129033e-05,
+      "loss": 0.173,
+      "step": 17873
+    },
+    {
+      "epoch": 0.285984,
+      "grad_norm": 1.46875,
+      "learning_rate": 7.197903225806452e-05,
+      "loss": 0.2145,
+      "step": 17874
+    },
+    {
+      "epoch": 0.286,
+      "grad_norm": 1.2578125,
+      "learning_rate": 7.197741935483872e-05,
+      "loss": 0.1684,
+      "step": 17875
+    },
+    {
+      "epoch": 0.286016,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.19758064516129e-05,
+      "loss": 0.1652,
+      "step": 17876
+    },
+    {
+      "epoch": 0.286032,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.19741935483871e-05,
+      "loss": 0.1478,
+      "step": 17877
+    },
+    {
+      "epoch": 0.286048,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.197258064516129e-05,
+      "loss": 0.2307,
+      "step": 17878
+    },
+    {
+      "epoch": 0.286064,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.197096774193549e-05,
+      "loss": 0.1611,
+      "step": 17879
+    },
+    {
+      "epoch": 0.28608,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.196935483870967e-05,
+      "loss": 0.1758,
+      "step": 17880
+    },
+    {
+      "epoch": 0.286096,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.196774193548387e-05,
+      "loss": 0.1185,
+      "step": 17881
+    },
+    {
+      "epoch": 0.286112,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.196612903225807e-05,
+      "loss": 0.1447,
+      "step": 17882
+    },
+    {
+      "epoch": 0.286128,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.196451612903227e-05,
+      "loss": 0.1703,
+      "step": 17883
+    },
+    {
+      "epoch": 0.286144,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.196290322580646e-05,
+      "loss": 0.1406,
+      "step": 17884
+    },
+    {
+      "epoch": 0.28616,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.196129032258064e-05,
+      "loss": 0.1206,
+      "step": 17885
+    },
+    {
+      "epoch": 0.286176,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.195967741935484e-05,
+      "loss": 0.2129,
+      "step": 17886
+    },
+    {
+      "epoch": 0.286192,
+      "grad_norm": 1.484375,
+      "learning_rate": 7.195806451612903e-05,
+      "loss": 0.1868,
+      "step": 17887
+    },
+    {
+      "epoch": 0.286208,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.195645161290323e-05,
+      "loss": 0.1317,
+      "step": 17888
+    },
+    {
+      "epoch": 0.286224,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.195483870967742e-05,
+      "loss": 0.1671,
+      "step": 17889
+    },
+    {
+      "epoch": 0.28624,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.195322580645161e-05,
+      "loss": 0.1303,
+      "step": 17890
+    },
+    {
+      "epoch": 0.286256,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.19516129032258e-05,
+      "loss": 0.1712,
+      "step": 17891
+    },
+    {
+      "epoch": 0.286272,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.195e-05,
+      "loss": 0.2074,
+      "step": 17892
+    },
+    {
+      "epoch": 0.286288,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.19483870967742e-05,
+      "loss": 0.1865,
+      "step": 17893
+    },
+    {
+      "epoch": 0.286304,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.19467741935484e-05,
+      "loss": 0.1636,
+      "step": 17894
+    },
+    {
+      "epoch": 0.28632,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.194516129032259e-05,
+      "loss": 0.2233,
+      "step": 17895
+    },
+    {
+      "epoch": 0.286336,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.194354838709679e-05,
+      "loss": 0.1453,
+      "step": 17896
+    },
+    {
+      "epoch": 0.286352,
+      "grad_norm": 0.75,
+      "learning_rate": 7.194193548387097e-05,
+      "loss": 0.1741,
+      "step": 17897
+    },
+    {
+      "epoch": 0.286368,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.194032258064517e-05,
+      "loss": 0.1743,
+      "step": 17898
+    },
+    {
+      "epoch": 0.286384,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.193870967741936e-05,
+      "loss": 0.1869,
+      "step": 17899
+    },
+    {
+      "epoch": 0.2864,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.193709677419354e-05,
+      "loss": 0.1735,
+      "step": 17900
+    },
+    {
+      "epoch": 0.286416,
+      "grad_norm": 1.3984375,
+      "learning_rate": 7.193548387096774e-05,
+      "loss": 0.208,
+      "step": 17901
+    },
+    {
+      "epoch": 0.286432,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.193387096774193e-05,
+      "loss": 0.1655,
+      "step": 17902
+    },
+    {
+      "epoch": 0.286448,
+      "grad_norm": 0.625,
+      "learning_rate": 7.193225806451613e-05,
+      "loss": 0.15,
+      "step": 17903
+    },
+    {
+      "epoch": 0.286464,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.193064516129033e-05,
+      "loss": 0.1758,
+      "step": 17904
+    },
+    {
+      "epoch": 0.28648,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.192903225806453e-05,
+      "loss": 0.1447,
+      "step": 17905
+    },
+    {
+      "epoch": 0.286496,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.192741935483871e-05,
+      "loss": 0.2025,
+      "step": 17906
+    },
+    {
+      "epoch": 0.286512,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.192580645161291e-05,
+      "loss": 0.168,
+      "step": 17907
+    },
+    {
+      "epoch": 0.286528,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.19241935483871e-05,
+      "loss": 0.1327,
+      "step": 17908
+    },
+    {
+      "epoch": 0.286544,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.19225806451613e-05,
+      "loss": 0.1955,
+      "step": 17909
+    },
+    {
+      "epoch": 0.28656,
+      "grad_norm": 1.328125,
+      "learning_rate": 7.192096774193549e-05,
+      "loss": 0.202,
+      "step": 17910
+    },
+    {
+      "epoch": 0.286576,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.191935483870969e-05,
+      "loss": 0.1694,
+      "step": 17911
+    },
+    {
+      "epoch": 0.286592,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.191774193548387e-05,
+      "loss": 0.215,
+      "step": 17912
+    },
+    {
+      "epoch": 0.286608,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.191612903225807e-05,
+      "loss": 0.1591,
+      "step": 17913
+    },
+    {
+      "epoch": 0.286624,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.191451612903226e-05,
+      "loss": 0.1406,
+      "step": 17914
+    },
+    {
+      "epoch": 0.28664,
+      "grad_norm": 0.75,
+      "learning_rate": 7.191290322580646e-05,
+      "loss": 0.1594,
+      "step": 17915
+    },
+    {
+      "epoch": 0.286656,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.191129032258064e-05,
+      "loss": 0.1621,
+      "step": 17916
+    },
+    {
+      "epoch": 0.286672,
+      "grad_norm": 1.5234375,
+      "learning_rate": 7.190967741935484e-05,
+      "loss": 0.1783,
+      "step": 17917
+    },
+    {
+      "epoch": 0.286688,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.190806451612904e-05,
+      "loss": 0.1524,
+      "step": 17918
+    },
+    {
+      "epoch": 0.286704,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.190645161290323e-05,
+      "loss": 0.1931,
+      "step": 17919
+    },
+    {
+      "epoch": 0.28672,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.190483870967743e-05,
+      "loss": 0.1554,
+      "step": 17920
+    },
+    {
+      "epoch": 0.286736,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.190322580645161e-05,
+      "loss": 0.1566,
+      "step": 17921
+    },
+    {
+      "epoch": 0.286752,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.190161290322581e-05,
+      "loss": 0.137,
+      "step": 17922
+    },
+    {
+      "epoch": 0.286768,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.19e-05,
+      "loss": 0.1647,
+      "step": 17923
+    },
+    {
+      "epoch": 0.286784,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.18983870967742e-05,
+      "loss": 0.1775,
+      "step": 17924
+    },
+    {
+      "epoch": 0.2868,
+      "grad_norm": 0.875,
+      "learning_rate": 7.189677419354839e-05,
+      "loss": 0.1967,
+      "step": 17925
+    },
+    {
+      "epoch": 0.286816,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.189516129032258e-05,
+      "loss": 0.1554,
+      "step": 17926
+    },
+    {
+      "epoch": 0.286832,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.189354838709677e-05,
+      "loss": 0.1363,
+      "step": 17927
+    },
+    {
+      "epoch": 0.286848,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.189193548387097e-05,
+      "loss": 0.1579,
+      "step": 17928
+    },
+    {
+      "epoch": 0.286864,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.189032258064517e-05,
+      "loss": 0.1865,
+      "step": 17929
+    },
+    {
+      "epoch": 0.28688,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.188870967741937e-05,
+      "loss": 0.1624,
+      "step": 17930
+    },
+    {
+      "epoch": 0.286896,
+      "grad_norm": 0.5,
+      "learning_rate": 7.188709677419356e-05,
+      "loss": 0.1365,
+      "step": 17931
+    },
+    {
+      "epoch": 0.286912,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.188548387096774e-05,
+      "loss": 0.1584,
+      "step": 17932
+    },
+    {
+      "epoch": 0.286928,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.188387096774194e-05,
+      "loss": 0.1372,
+      "step": 17933
+    },
+    {
+      "epoch": 0.286944,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.188225806451613e-05,
+      "loss": 0.2213,
+      "step": 17934
+    },
+    {
+      "epoch": 0.28696,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.188064516129033e-05,
+      "loss": 0.1612,
+      "step": 17935
+    },
+    {
+      "epoch": 0.286976,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.187903225806451e-05,
+      "loss": 0.2026,
+      "step": 17936
+    },
+    {
+      "epoch": 0.286992,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.187741935483871e-05,
+      "loss": 0.1844,
+      "step": 17937
+    },
+    {
+      "epoch": 0.287008,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.18758064516129e-05,
+      "loss": 0.1738,
+      "step": 17938
+    },
+    {
+      "epoch": 0.287024,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.18741935483871e-05,
+      "loss": 0.1637,
+      "step": 17939
+    },
+    {
+      "epoch": 0.28704,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.18725806451613e-05,
+      "loss": 0.1767,
+      "step": 17940
+    },
+    {
+      "epoch": 0.287056,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.187096774193548e-05,
+      "loss": 0.1823,
+      "step": 17941
+    },
+    {
+      "epoch": 0.287072,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.186935483870968e-05,
+      "loss": 0.1656,
+      "step": 17942
+    },
+    {
+      "epoch": 0.287088,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.186774193548388e-05,
+      "loss": 0.1966,
+      "step": 17943
+    },
+    {
+      "epoch": 0.287104,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.186612903225807e-05,
+      "loss": 0.1797,
+      "step": 17944
+    },
+    {
+      "epoch": 0.28712,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.186451612903227e-05,
+      "loss": 0.172,
+      "step": 17945
+    },
+    {
+      "epoch": 0.287136,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.186290322580646e-05,
+      "loss": 0.1779,
+      "step": 17946
+    },
+    {
+      "epoch": 0.287152,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.186129032258064e-05,
+      "loss": 0.2105,
+      "step": 17947
+    },
+    {
+      "epoch": 0.287168,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.185967741935484e-05,
+      "loss": 0.1836,
+      "step": 17948
+    },
+    {
+      "epoch": 0.287184,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.185806451612903e-05,
+      "loss": 0.1505,
+      "step": 17949
+    },
+    {
+      "epoch": 0.2872,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.185645161290323e-05,
+      "loss": 0.153,
+      "step": 17950
+    },
+    {
+      "epoch": 0.287216,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.185483870967741e-05,
+      "loss": 0.2282,
+      "step": 17951
+    },
+    {
+      "epoch": 0.287232,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.185322580645161e-05,
+      "loss": 0.1224,
+      "step": 17952
+    },
+    {
+      "epoch": 0.287248,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.185161290322581e-05,
+      "loss": 0.1529,
+      "step": 17953
+    },
+    {
+      "epoch": 0.287264,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.185000000000001e-05,
+      "loss": 0.1338,
+      "step": 17954
+    },
+    {
+      "epoch": 0.28728,
+      "grad_norm": 1.390625,
+      "learning_rate": 7.18483870967742e-05,
+      "loss": 0.1934,
+      "step": 17955
+    },
+    {
+      "epoch": 0.287296,
+      "grad_norm": 1.4765625,
+      "learning_rate": 7.18467741935484e-05,
+      "loss": 0.2188,
+      "step": 17956
+    },
+    {
+      "epoch": 0.287312,
+      "grad_norm": 1.25,
+      "learning_rate": 7.184516129032258e-05,
+      "loss": 0.1631,
+      "step": 17957
+    },
+    {
+      "epoch": 0.287328,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.184354838709678e-05,
+      "loss": 0.1792,
+      "step": 17958
+    },
+    {
+      "epoch": 0.287344,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.184193548387097e-05,
+      "loss": 0.1741,
+      "step": 17959
+    },
+    {
+      "epoch": 0.28736,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.184032258064517e-05,
+      "loss": 0.1545,
+      "step": 17960
+    },
+    {
+      "epoch": 0.287376,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.183870967741935e-05,
+      "loss": 0.189,
+      "step": 17961
+    },
+    {
+      "epoch": 0.287392,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.183709677419354e-05,
+      "loss": 0.1843,
+      "step": 17962
+    },
+    {
+      "epoch": 0.287408,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.183548387096774e-05,
+      "loss": 0.1326,
+      "step": 17963
+    },
+    {
+      "epoch": 0.287424,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.183387096774194e-05,
+      "loss": 0.1892,
+      "step": 17964
+    },
+    {
+      "epoch": 0.28744,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.183225806451614e-05,
+      "loss": 0.1282,
+      "step": 17965
+    },
+    {
+      "epoch": 0.287456,
+      "grad_norm": 0.494140625,
+      "learning_rate": 7.183064516129033e-05,
+      "loss": 0.1478,
+      "step": 17966
+    },
+    {
+      "epoch": 0.287472,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.182903225806453e-05,
+      "loss": 0.1859,
+      "step": 17967
+    },
+    {
+      "epoch": 0.287488,
+      "grad_norm": 0.875,
+      "learning_rate": 7.182741935483871e-05,
+      "loss": 0.1911,
+      "step": 17968
+    },
+    {
+      "epoch": 0.287504,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.182580645161291e-05,
+      "loss": 0.153,
+      "step": 17969
+    },
+    {
+      "epoch": 0.28752,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.18241935483871e-05,
+      "loss": 0.1783,
+      "step": 17970
+    },
+    {
+      "epoch": 0.287536,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.18225806451613e-05,
+      "loss": 0.1375,
+      "step": 17971
+    },
+    {
+      "epoch": 0.287552,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.182096774193548e-05,
+      "loss": 0.2075,
+      "step": 17972
+    },
+    {
+      "epoch": 0.287568,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.181935483870968e-05,
+      "loss": 0.1405,
+      "step": 17973
+    },
+    {
+      "epoch": 0.287584,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.181774193548387e-05,
+      "loss": 0.163,
+      "step": 17974
+    },
+    {
+      "epoch": 0.2876,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.181612903225807e-05,
+      "loss": 0.1412,
+      "step": 17975
+    },
+    {
+      "epoch": 0.287616,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.181451612903225e-05,
+      "loss": 0.198,
+      "step": 17976
+    },
+    {
+      "epoch": 0.287632,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.181290322580645e-05,
+      "loss": 0.1403,
+      "step": 17977
+    },
+    {
+      "epoch": 0.287648,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.181129032258065e-05,
+      "loss": 0.1646,
+      "step": 17978
+    },
+    {
+      "epoch": 0.287664,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.180967741935484e-05,
+      "loss": 0.1931,
+      "step": 17979
+    },
+    {
+      "epoch": 0.28768,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.180806451612904e-05,
+      "loss": 0.1536,
+      "step": 17980
+    },
+    {
+      "epoch": 0.287696,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.180645161290323e-05,
+      "loss": 0.1999,
+      "step": 17981
+    },
+    {
+      "epoch": 0.287712,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.180483870967743e-05,
+      "loss": 0.1836,
+      "step": 17982
+    },
+    {
+      "epoch": 0.287728,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.180322580645161e-05,
+      "loss": 0.1549,
+      "step": 17983
+    },
+    {
+      "epoch": 0.287744,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.180161290322581e-05,
+      "loss": 0.1678,
+      "step": 17984
+    },
+    {
+      "epoch": 0.28776,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.18e-05,
+      "loss": 0.2062,
+      "step": 17985
+    },
+    {
+      "epoch": 0.287776,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.17983870967742e-05,
+      "loss": 0.1469,
+      "step": 17986
+    },
+    {
+      "epoch": 0.287792,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.179677419354838e-05,
+      "loss": 0.1546,
+      "step": 17987
+    },
+    {
+      "epoch": 0.287808,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.179516129032258e-05,
+      "loss": 0.1553,
+      "step": 17988
+    },
+    {
+      "epoch": 0.287824,
+      "grad_norm": 0.455078125,
+      "learning_rate": 7.179354838709678e-05,
+      "loss": 0.1464,
+      "step": 17989
+    },
+    {
+      "epoch": 0.28784,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.179193548387098e-05,
+      "loss": 0.1818,
+      "step": 17990
+    },
+    {
+      "epoch": 0.287856,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.179032258064517e-05,
+      "loss": 0.1228,
+      "step": 17991
+    },
+    {
+      "epoch": 0.287872,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.178870967741937e-05,
+      "loss": 0.14,
+      "step": 17992
+    },
+    {
+      "epoch": 0.287888,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.178709677419355e-05,
+      "loss": 0.1753,
+      "step": 17993
+    },
+    {
+      "epoch": 0.287904,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.178548387096774e-05,
+      "loss": 0.1555,
+      "step": 17994
+    },
+    {
+      "epoch": 0.28792,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.178387096774194e-05,
+      "loss": 0.1648,
+      "step": 17995
+    },
+    {
+      "epoch": 0.287936,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.178225806451613e-05,
+      "loss": 0.1911,
+      "step": 17996
+    },
+    {
+      "epoch": 0.287952,
+      "grad_norm": 0.625,
+      "learning_rate": 7.178064516129032e-05,
+      "loss": 0.1637,
+      "step": 17997
+    },
+    {
+      "epoch": 0.287968,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.177903225806451e-05,
+      "loss": 0.167,
+      "step": 17998
+    },
+    {
+      "epoch": 0.287984,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.177741935483871e-05,
+      "loss": 0.156,
+      "step": 17999
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.177580645161291e-05,
+      "loss": 0.1754,
+      "step": 18000
+    },
+    {
+      "epoch": 0.288016,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.177419354838711e-05,
+      "loss": 0.1853,
+      "step": 18001
+    },
+    {
+      "epoch": 0.288032,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.17725806451613e-05,
+      "loss": 0.183,
+      "step": 18002
+    },
+    {
+      "epoch": 0.288048,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.17709677419355e-05,
+      "loss": 0.1947,
+      "step": 18003
+    },
+    {
+      "epoch": 0.288064,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.176935483870968e-05,
+      "loss": 0.0988,
+      "step": 18004
+    },
+    {
+      "epoch": 0.28808,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.176774193548388e-05,
+      "loss": 0.1455,
+      "step": 18005
+    },
+    {
+      "epoch": 0.288096,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.176612903225807e-05,
+      "loss": 0.1694,
+      "step": 18006
+    },
+    {
+      "epoch": 0.288112,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.176451612903227e-05,
+      "loss": 0.2039,
+      "step": 18007
+    },
+    {
+      "epoch": 0.288128,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.176290322580645e-05,
+      "loss": 0.1818,
+      "step": 18008
+    },
+    {
+      "epoch": 0.288144,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.176129032258064e-05,
+      "loss": 0.1735,
+      "step": 18009
+    },
+    {
+      "epoch": 0.28816,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.175967741935484e-05,
+      "loss": 0.1475,
+      "step": 18010
+    },
+    {
+      "epoch": 0.288176,
+      "grad_norm": 1.2578125,
+      "learning_rate": 7.175806451612902e-05,
+      "loss": 0.1654,
+      "step": 18011
+    },
+    {
+      "epoch": 0.288192,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.175645161290322e-05,
+      "loss": 0.2004,
+      "step": 18012
+    },
+    {
+      "epoch": 0.288208,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.175483870967742e-05,
+      "loss": 0.1376,
+      "step": 18013
+    },
+    {
+      "epoch": 0.288224,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.175322580645162e-05,
+      "loss": 0.1752,
+      "step": 18014
+    },
+    {
+      "epoch": 0.28824,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.175161290322581e-05,
+      "loss": 0.1442,
+      "step": 18015
+    },
+    {
+      "epoch": 0.288256,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.175000000000001e-05,
+      "loss": 0.1678,
+      "step": 18016
+    },
+    {
+      "epoch": 0.288272,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.17483870967742e-05,
+      "loss": 0.1402,
+      "step": 18017
+    },
+    {
+      "epoch": 0.288288,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.17467741935484e-05,
+      "loss": 0.1961,
+      "step": 18018
+    },
+    {
+      "epoch": 0.288304,
+      "grad_norm": 1.359375,
+      "learning_rate": 7.174516129032258e-05,
+      "loss": 0.2125,
+      "step": 18019
+    },
+    {
+      "epoch": 0.28832,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.174354838709678e-05,
+      "loss": 0.1627,
+      "step": 18020
+    },
+    {
+      "epoch": 0.288336,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.174193548387097e-05,
+      "loss": 0.2053,
+      "step": 18021
+    },
+    {
+      "epoch": 0.288352,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.174032258064517e-05,
+      "loss": 0.1289,
+      "step": 18022
+    },
+    {
+      "epoch": 0.288368,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.173870967741935e-05,
+      "loss": 0.1606,
+      "step": 18023
+    },
+    {
+      "epoch": 0.288384,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.173709677419355e-05,
+      "loss": 0.1785,
+      "step": 18024
+    },
+    {
+      "epoch": 0.2884,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.173548387096775e-05,
+      "loss": 0.1572,
+      "step": 18025
+    },
+    {
+      "epoch": 0.288416,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.173387096774194e-05,
+      "loss": 0.2321,
+      "step": 18026
+    },
+    {
+      "epoch": 0.288432,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.173225806451614e-05,
+      "loss": 0.108,
+      "step": 18027
+    },
+    {
+      "epoch": 0.288448,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.173064516129032e-05,
+      "loss": 0.1781,
+      "step": 18028
+    },
+    {
+      "epoch": 0.288464,
+      "grad_norm": 1.625,
+      "learning_rate": 7.172903225806452e-05,
+      "loss": 0.173,
+      "step": 18029
+    },
+    {
+      "epoch": 0.28848,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.172741935483871e-05,
+      "loss": 0.1357,
+      "step": 18030
+    },
+    {
+      "epoch": 0.288496,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.172580645161291e-05,
+      "loss": 0.1789,
+      "step": 18031
+    },
+    {
+      "epoch": 0.288512,
+      "grad_norm": 0.875,
+      "learning_rate": 7.17241935483871e-05,
+      "loss": 0.1587,
+      "step": 18032
+    },
+    {
+      "epoch": 0.288528,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.17225806451613e-05,
+      "loss": 0.1634,
+      "step": 18033
+    },
+    {
+      "epoch": 0.288544,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.172096774193548e-05,
+      "loss": 0.218,
+      "step": 18034
+    },
+    {
+      "epoch": 0.28856,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.171935483870968e-05,
+      "loss": 0.1628,
+      "step": 18035
+    },
+    {
+      "epoch": 0.288576,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.171774193548387e-05,
+      "loss": 0.1687,
+      "step": 18036
+    },
+    {
+      "epoch": 0.288592,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.171612903225807e-05,
+      "loss": 0.1626,
+      "step": 18037
+    },
+    {
+      "epoch": 0.288608,
+      "grad_norm": 1.421875,
+      "learning_rate": 7.171451612903227e-05,
+      "loss": 0.1279,
+      "step": 18038
+    },
+    {
+      "epoch": 0.288624,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.171290322580647e-05,
+      "loss": 0.1725,
+      "step": 18039
+    },
+    {
+      "epoch": 0.28864,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.171129032258065e-05,
+      "loss": 0.1779,
+      "step": 18040
+    },
+    {
+      "epoch": 0.288656,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.170967741935484e-05,
+      "loss": 0.149,
+      "step": 18041
+    },
+    {
+      "epoch": 0.288672,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.170806451612904e-05,
+      "loss": 0.2173,
+      "step": 18042
+    },
+    {
+      "epoch": 0.288688,
+      "grad_norm": 1.5234375,
+      "learning_rate": 7.170645161290322e-05,
+      "loss": 0.1672,
+      "step": 18043
+    },
+    {
+      "epoch": 0.288704,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.170483870967742e-05,
+      "loss": 0.1855,
+      "step": 18044
+    },
+    {
+      "epoch": 0.28872,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.170322580645161e-05,
+      "loss": 0.1651,
+      "step": 18045
+    },
+    {
+      "epoch": 0.288736,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.170161290322581e-05,
+      "loss": 0.1845,
+      "step": 18046
+    },
+    {
+      "epoch": 0.288752,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.17e-05,
+      "loss": 0.2153,
+      "step": 18047
+    },
+    {
+      "epoch": 0.288768,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.16983870967742e-05,
+      "loss": 0.2138,
+      "step": 18048
+    },
+    {
+      "epoch": 0.288784,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.16967741935484e-05,
+      "loss": 0.1509,
+      "step": 18049
+    },
+    {
+      "epoch": 0.2888,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.16951612903226e-05,
+      "loss": 0.1708,
+      "step": 18050
+    },
+    {
+      "epoch": 0.288816,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.169354838709678e-05,
+      "loss": 0.1864,
+      "step": 18051
+    },
+    {
+      "epoch": 0.288832,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.169193548387098e-05,
+      "loss": 0.1449,
+      "step": 18052
+    },
+    {
+      "epoch": 0.288848,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.169032258064517e-05,
+      "loss": 0.1504,
+      "step": 18053
+    },
+    {
+      "epoch": 0.288864,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.168870967741936e-05,
+      "loss": 0.1479,
+      "step": 18054
+    },
+    {
+      "epoch": 0.28888,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.168709677419355e-05,
+      "loss": 0.1756,
+      "step": 18055
+    },
+    {
+      "epoch": 0.288896,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.168548387096774e-05,
+      "loss": 0.1899,
+      "step": 18056
+    },
+    {
+      "epoch": 0.288912,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.168387096774194e-05,
+      "loss": 0.1615,
+      "step": 18057
+    },
+    {
+      "epoch": 0.288928,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.168225806451612e-05,
+      "loss": 0.1994,
+      "step": 18058
+    },
+    {
+      "epoch": 0.288944,
+      "grad_norm": 1.34375,
+      "learning_rate": 7.168064516129032e-05,
+      "loss": 0.1674,
+      "step": 18059
+    },
+    {
+      "epoch": 0.28896,
+      "grad_norm": 1.28125,
+      "learning_rate": 7.167903225806452e-05,
+      "loss": 0.1711,
+      "step": 18060
+    },
+    {
+      "epoch": 0.288976,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.167741935483872e-05,
+      "loss": 0.1873,
+      "step": 18061
+    },
+    {
+      "epoch": 0.288992,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.167580645161291e-05,
+      "loss": 0.1174,
+      "step": 18062
+    },
+    {
+      "epoch": 0.289008,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.167419354838711e-05,
+      "loss": 0.149,
+      "step": 18063
+    },
+    {
+      "epoch": 0.289024,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.16725806451613e-05,
+      "loss": 0.1845,
+      "step": 18064
+    },
+    {
+      "epoch": 0.28904,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.167096774193549e-05,
+      "loss": 0.1646,
+      "step": 18065
+    },
+    {
+      "epoch": 0.289056,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.166935483870968e-05,
+      "loss": 0.1944,
+      "step": 18066
+    },
+    {
+      "epoch": 0.289072,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.166774193548388e-05,
+      "loss": 0.1801,
+      "step": 18067
+    },
+    {
+      "epoch": 0.289088,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.166612903225806e-05,
+      "loss": 0.1569,
+      "step": 18068
+    },
+    {
+      "epoch": 0.289104,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.166451612903226e-05,
+      "loss": 0.1471,
+      "step": 18069
+    },
+    {
+      "epoch": 0.28912,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.166290322580645e-05,
+      "loss": 0.174,
+      "step": 18070
+    },
+    {
+      "epoch": 0.289136,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.166129032258064e-05,
+      "loss": 0.1722,
+      "step": 18071
+    },
+    {
+      "epoch": 0.289152,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.165967741935484e-05,
+      "loss": 0.1563,
+      "step": 18072
+    },
+    {
+      "epoch": 0.289168,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.165806451612904e-05,
+      "loss": 0.1507,
+      "step": 18073
+    },
+    {
+      "epoch": 0.289184,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.165645161290324e-05,
+      "loss": 0.1904,
+      "step": 18074
+    },
+    {
+      "epoch": 0.2892,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.165483870967742e-05,
+      "loss": 0.157,
+      "step": 18075
+    },
+    {
+      "epoch": 0.289216,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.165322580645162e-05,
+      "loss": 0.126,
+      "step": 18076
+    },
+    {
+      "epoch": 0.289232,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.165161290322581e-05,
+      "loss": 0.1961,
+      "step": 18077
+    },
+    {
+      "epoch": 0.289248,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.165000000000001e-05,
+      "loss": 0.1683,
+      "step": 18078
+    },
+    {
+      "epoch": 0.289264,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.164838709677419e-05,
+      "loss": 0.1487,
+      "step": 18079
+    },
+    {
+      "epoch": 0.28928,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.164677419354839e-05,
+      "loss": 0.1494,
+      "step": 18080
+    },
+    {
+      "epoch": 0.289296,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.164516129032258e-05,
+      "loss": 0.1452,
+      "step": 18081
+    },
+    {
+      "epoch": 0.289312,
+      "grad_norm": 1.3203125,
+      "learning_rate": 7.164354838709678e-05,
+      "loss": 0.1811,
+      "step": 18082
+    },
+    {
+      "epoch": 0.289328,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.164193548387096e-05,
+      "loss": 0.18,
+      "step": 18083
+    },
+    {
+      "epoch": 0.289344,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.164032258064516e-05,
+      "loss": 0.1928,
+      "step": 18084
+    },
+    {
+      "epoch": 0.28936,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.163870967741936e-05,
+      "loss": 0.1791,
+      "step": 18085
+    },
+    {
+      "epoch": 0.289376,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.163709677419356e-05,
+      "loss": 0.1465,
+      "step": 18086
+    },
+    {
+      "epoch": 0.289392,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.163548387096775e-05,
+      "loss": 0.1957,
+      "step": 18087
+    },
+    {
+      "epoch": 0.289408,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.163387096774194e-05,
+      "loss": 0.1828,
+      "step": 18088
+    },
+    {
+      "epoch": 0.289424,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.163225806451614e-05,
+      "loss": 0.151,
+      "step": 18089
+    },
+    {
+      "epoch": 0.28944,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.163064516129032e-05,
+      "loss": 0.185,
+      "step": 18090
+    },
+    {
+      "epoch": 0.289456,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.162903225806452e-05,
+      "loss": 0.1629,
+      "step": 18091
+    },
+    {
+      "epoch": 0.289472,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.162741935483871e-05,
+      "loss": 0.1529,
+      "step": 18092
+    },
+    {
+      "epoch": 0.289488,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.16258064516129e-05,
+      "loss": 0.2062,
+      "step": 18093
+    },
+    {
+      "epoch": 0.289504,
+      "grad_norm": 0.875,
+      "learning_rate": 7.162419354838709e-05,
+      "loss": 0.1373,
+      "step": 18094
+    },
+    {
+      "epoch": 0.28952,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.162258064516129e-05,
+      "loss": 0.2122,
+      "step": 18095
+    },
+    {
+      "epoch": 0.289536,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.162096774193549e-05,
+      "loss": 0.1486,
+      "step": 18096
+    },
+    {
+      "epoch": 0.289552,
+      "grad_norm": 1.125,
+      "learning_rate": 7.161935483870969e-05,
+      "loss": 0.1428,
+      "step": 18097
+    },
+    {
+      "epoch": 0.289568,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.161774193548388e-05,
+      "loss": 0.1742,
+      "step": 18098
+    },
+    {
+      "epoch": 0.289584,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.161612903225808e-05,
+      "loss": 0.1862,
+      "step": 18099
+    },
+    {
+      "epoch": 0.2896,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.161451612903226e-05,
+      "loss": 0.1433,
+      "step": 18100
+    },
+    {
+      "epoch": 0.289616,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.161290322580646e-05,
+      "loss": 0.1724,
+      "step": 18101
+    },
+    {
+      "epoch": 0.289632,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.161129032258065e-05,
+      "loss": 0.1487,
+      "step": 18102
+    },
+    {
+      "epoch": 0.289648,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.160967741935483e-05,
+      "loss": 0.1319,
+      "step": 18103
+    },
+    {
+      "epoch": 0.289664,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.160806451612903e-05,
+      "loss": 0.2024,
+      "step": 18104
+    },
+    {
+      "epoch": 0.28968,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.160645161290322e-05,
+      "loss": 0.1649,
+      "step": 18105
+    },
+    {
+      "epoch": 0.289696,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.160483870967742e-05,
+      "loss": 0.1508,
+      "step": 18106
+    },
+    {
+      "epoch": 0.289712,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.16032258064516e-05,
+      "loss": 0.1495,
+      "step": 18107
+    },
+    {
+      "epoch": 0.289728,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.16016129032258e-05,
+      "loss": 0.1665,
+      "step": 18108
+    },
+    {
+      "epoch": 0.289744,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.16e-05,
+      "loss": 0.1712,
+      "step": 18109
+    },
+    {
+      "epoch": 0.28976,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.15983870967742e-05,
+      "loss": 0.1767,
+      "step": 18110
+    },
+    {
+      "epoch": 0.289776,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.159677419354839e-05,
+      "loss": 0.1593,
+      "step": 18111
+    },
+    {
+      "epoch": 0.289792,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.159516129032259e-05,
+      "loss": 0.1625,
+      "step": 18112
+    },
+    {
+      "epoch": 0.289808,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.159354838709678e-05,
+      "loss": 0.1765,
+      "step": 18113
+    },
+    {
+      "epoch": 0.289824,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.159193548387098e-05,
+      "loss": 0.1357,
+      "step": 18114
+    },
+    {
+      "epoch": 0.28984,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.159032258064516e-05,
+      "loss": 0.1657,
+      "step": 18115
+    },
+    {
+      "epoch": 0.289856,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.158870967741936e-05,
+      "loss": 0.1658,
+      "step": 18116
+    },
+    {
+      "epoch": 0.289872,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.158709677419355e-05,
+      "loss": 0.182,
+      "step": 18117
+    },
+    {
+      "epoch": 0.289888,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.158548387096773e-05,
+      "loss": 0.1186,
+      "step": 18118
+    },
+    {
+      "epoch": 0.289904,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.158387096774193e-05,
+      "loss": 0.1788,
+      "step": 18119
+    },
+    {
+      "epoch": 0.28992,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.158225806451613e-05,
+      "loss": 0.1508,
+      "step": 18120
+    },
+    {
+      "epoch": 0.289936,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.158064516129033e-05,
+      "loss": 0.1702,
+      "step": 18121
+    },
+    {
+      "epoch": 0.289952,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.157903225806452e-05,
+      "loss": 0.1621,
+      "step": 18122
+    },
+    {
+      "epoch": 0.289968,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.157741935483872e-05,
+      "loss": 0.1949,
+      "step": 18123
+    },
+    {
+      "epoch": 0.289984,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.15758064516129e-05,
+      "loss": 0.1618,
+      "step": 18124
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3125,
+      "learning_rate": 7.15741935483871e-05,
+      "loss": 0.1641,
+      "step": 18125
+    },
+    {
+      "epoch": 0.290016,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.157258064516129e-05,
+      "loss": 0.1498,
+      "step": 18126
+    },
+    {
+      "epoch": 0.290032,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.157096774193549e-05,
+      "loss": 0.1813,
+      "step": 18127
+    },
+    {
+      "epoch": 0.290048,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.156935483870968e-05,
+      "loss": 0.1944,
+      "step": 18128
+    },
+    {
+      "epoch": 0.290064,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.156774193548388e-05,
+      "loss": 0.1635,
+      "step": 18129
+    },
+    {
+      "epoch": 0.29008,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.156612903225806e-05,
+      "loss": 0.1925,
+      "step": 18130
+    },
+    {
+      "epoch": 0.290096,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.156451612903226e-05,
+      "loss": 0.1706,
+      "step": 18131
+    },
+    {
+      "epoch": 0.290112,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.156290322580645e-05,
+      "loss": 0.1711,
+      "step": 18132
+    },
+    {
+      "epoch": 0.290128,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.156129032258065e-05,
+      "loss": 0.1693,
+      "step": 18133
+    },
+    {
+      "epoch": 0.290144,
+      "grad_norm": 1.3359375,
+      "learning_rate": 7.155967741935485e-05,
+      "loss": 0.1945,
+      "step": 18134
+    },
+    {
+      "epoch": 0.29016,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.155806451612903e-05,
+      "loss": 0.1516,
+      "step": 18135
+    },
+    {
+      "epoch": 0.290176,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.155645161290323e-05,
+      "loss": 0.1865,
+      "step": 18136
+    },
+    {
+      "epoch": 0.290192,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.155483870967742e-05,
+      "loss": 0.1841,
+      "step": 18137
+    },
+    {
+      "epoch": 0.290208,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.155322580645162e-05,
+      "loss": 0.1757,
+      "step": 18138
+    },
+    {
+      "epoch": 0.290224,
+      "grad_norm": 1.546875,
+      "learning_rate": 7.15516129032258e-05,
+      "loss": 0.1937,
+      "step": 18139
+    },
+    {
+      "epoch": 0.29024,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.155e-05,
+      "loss": 0.1608,
+      "step": 18140
+    },
+    {
+      "epoch": 0.290256,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.154838709677419e-05,
+      "loss": 0.2011,
+      "step": 18141
+    },
+    {
+      "epoch": 0.290272,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.154677419354839e-05,
+      "loss": 0.1491,
+      "step": 18142
+    },
+    {
+      "epoch": 0.290288,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.154516129032258e-05,
+      "loss": 0.1671,
+      "step": 18143
+    },
+    {
+      "epoch": 0.290304,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.154354838709678e-05,
+      "loss": 0.1631,
+      "step": 18144
+    },
+    {
+      "epoch": 0.29032,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.154193548387098e-05,
+      "loss": 0.1602,
+      "step": 18145
+    },
+    {
+      "epoch": 0.290336,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.154032258064518e-05,
+      "loss": 0.1801,
+      "step": 18146
+    },
+    {
+      "epoch": 0.290352,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.153870967741936e-05,
+      "loss": 0.1872,
+      "step": 18147
+    },
+    {
+      "epoch": 0.290368,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.153709677419356e-05,
+      "loss": 0.1531,
+      "step": 18148
+    },
+    {
+      "epoch": 0.290384,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.153548387096775e-05,
+      "loss": 0.1997,
+      "step": 18149
+    },
+    {
+      "epoch": 0.2904,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.153387096774193e-05,
+      "loss": 0.1948,
+      "step": 18150
+    },
+    {
+      "epoch": 0.290416,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.153225806451613e-05,
+      "loss": 0.1909,
+      "step": 18151
+    },
+    {
+      "epoch": 0.290432,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.153064516129032e-05,
+      "loss": 0.1563,
+      "step": 18152
+    },
+    {
+      "epoch": 0.290448,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.152903225806452e-05,
+      "loss": 0.1587,
+      "step": 18153
+    },
+    {
+      "epoch": 0.290464,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.15274193548387e-05,
+      "loss": 0.1712,
+      "step": 18154
+    },
+    {
+      "epoch": 0.29048,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.15258064516129e-05,
+      "loss": 0.1868,
+      "step": 18155
+    },
+    {
+      "epoch": 0.290496,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.15241935483871e-05,
+      "loss": 0.1882,
+      "step": 18156
+    },
+    {
+      "epoch": 0.290512,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.15225806451613e-05,
+      "loss": 0.1793,
+      "step": 18157
+    },
+    {
+      "epoch": 0.290528,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.152096774193549e-05,
+      "loss": 0.1629,
+      "step": 18158
+    },
+    {
+      "epoch": 0.290544,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.151935483870969e-05,
+      "loss": 0.1504,
+      "step": 18159
+    },
+    {
+      "epoch": 0.29056,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.151774193548388e-05,
+      "loss": 0.1852,
+      "step": 18160
+    },
+    {
+      "epoch": 0.290576,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.151612903225807e-05,
+      "loss": 0.1651,
+      "step": 18161
+    },
+    {
+      "epoch": 0.290592,
+      "grad_norm": 1.2421875,
+      "learning_rate": 7.151451612903226e-05,
+      "loss": 0.1778,
+      "step": 18162
+    },
+    {
+      "epoch": 0.290608,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.151290322580646e-05,
+      "loss": 0.1593,
+      "step": 18163
+    },
+    {
+      "epoch": 0.290624,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.151129032258065e-05,
+      "loss": 0.1679,
+      "step": 18164
+    },
+    {
+      "epoch": 0.29064,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.150967741935483e-05,
+      "loss": 0.1794,
+      "step": 18165
+    },
+    {
+      "epoch": 0.290656,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.150806451612903e-05,
+      "loss": 0.1615,
+      "step": 18166
+    },
+    {
+      "epoch": 0.290672,
+      "grad_norm": 1.4765625,
+      "learning_rate": 7.150645161290322e-05,
+      "loss": 0.2258,
+      "step": 18167
+    },
+    {
+      "epoch": 0.290688,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.150483870967742e-05,
+      "loss": 0.1544,
+      "step": 18168
+    },
+    {
+      "epoch": 0.290704,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.150322580645162e-05,
+      "loss": 0.2007,
+      "step": 18169
+    },
+    {
+      "epoch": 0.29072,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.150161290322582e-05,
+      "loss": 0.1583,
+      "step": 18170
+    },
+    {
+      "epoch": 0.290736,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.15e-05,
+      "loss": 0.1727,
+      "step": 18171
+    },
+    {
+      "epoch": 0.290752,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.14983870967742e-05,
+      "loss": 0.1513,
+      "step": 18172
+    },
+    {
+      "epoch": 0.290768,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.149677419354839e-05,
+      "loss": 0.2191,
+      "step": 18173
+    },
+    {
+      "epoch": 0.290784,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.149516129032259e-05,
+      "loss": 0.163,
+      "step": 18174
+    },
+    {
+      "epoch": 0.2908,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.149354838709677e-05,
+      "loss": 0.1605,
+      "step": 18175
+    },
+    {
+      "epoch": 0.290816,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.149193548387097e-05,
+      "loss": 0.1903,
+      "step": 18176
+    },
+    {
+      "epoch": 0.290832,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.149032258064516e-05,
+      "loss": 0.1887,
+      "step": 18177
+    },
+    {
+      "epoch": 0.290848,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.148870967741936e-05,
+      "loss": 0.165,
+      "step": 18178
+    },
+    {
+      "epoch": 0.290864,
+      "grad_norm": 0.75,
+      "learning_rate": 7.148709677419355e-05,
+      "loss": 0.1598,
+      "step": 18179
+    },
+    {
+      "epoch": 0.29088,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.148548387096775e-05,
+      "loss": 0.1712,
+      "step": 18180
+    },
+    {
+      "epoch": 0.290896,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.148387096774195e-05,
+      "loss": 0.133,
+      "step": 18181
+    },
+    {
+      "epoch": 0.290912,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.148225806451613e-05,
+      "loss": 0.18,
+      "step": 18182
+    },
+    {
+      "epoch": 0.290928,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.148064516129033e-05,
+      "loss": 0.1819,
+      "step": 18183
+    },
+    {
+      "epoch": 0.290944,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.147903225806452e-05,
+      "loss": 0.1864,
+      "step": 18184
+    },
+    {
+      "epoch": 0.29096,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.147741935483872e-05,
+      "loss": 0.2048,
+      "step": 18185
+    },
+    {
+      "epoch": 0.290976,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.14758064516129e-05,
+      "loss": 0.1733,
+      "step": 18186
+    },
+    {
+      "epoch": 0.290992,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.14741935483871e-05,
+      "loss": 0.1445,
+      "step": 18187
+    },
+    {
+      "epoch": 0.291008,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.147258064516129e-05,
+      "loss": 0.1512,
+      "step": 18188
+    },
+    {
+      "epoch": 0.291024,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.147096774193549e-05,
+      "loss": 0.1669,
+      "step": 18189
+    },
+    {
+      "epoch": 0.29104,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.146935483870967e-05,
+      "loss": 0.1695,
+      "step": 18190
+    },
+    {
+      "epoch": 0.291056,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.146774193548387e-05,
+      "loss": 0.1699,
+      "step": 18191
+    },
+    {
+      "epoch": 0.291072,
+      "grad_norm": 1.4296875,
+      "learning_rate": 7.146612903225807e-05,
+      "loss": 0.2167,
+      "step": 18192
+    },
+    {
+      "epoch": 0.291088,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.146451612903226e-05,
+      "loss": 0.156,
+      "step": 18193
+    },
+    {
+      "epoch": 0.291104,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.146290322580646e-05,
+      "loss": 0.1573,
+      "step": 18194
+    },
+    {
+      "epoch": 0.29112,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.146129032258066e-05,
+      "loss": 0.1785,
+      "step": 18195
+    },
+    {
+      "epoch": 0.291136,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.145967741935484e-05,
+      "loss": 0.1479,
+      "step": 18196
+    },
+    {
+      "epoch": 0.291152,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.145806451612903e-05,
+      "loss": 0.1682,
+      "step": 18197
+    },
+    {
+      "epoch": 0.291168,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.145645161290323e-05,
+      "loss": 0.1794,
+      "step": 18198
+    },
+    {
+      "epoch": 0.291184,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.145483870967742e-05,
+      "loss": 0.1524,
+      "step": 18199
+    },
+    {
+      "epoch": 0.2912,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.145322580645162e-05,
+      "loss": 0.1791,
+      "step": 18200
+    },
+    {
+      "epoch": 0.291216,
+      "grad_norm": 0.625,
+      "learning_rate": 7.14516129032258e-05,
+      "loss": 0.1994,
+      "step": 18201
+    },
+    {
+      "epoch": 0.291232,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.145e-05,
+      "loss": 0.1488,
+      "step": 18202
+    },
+    {
+      "epoch": 0.291248,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.144838709677419e-05,
+      "loss": 0.146,
+      "step": 18203
+    },
+    {
+      "epoch": 0.291264,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.144677419354839e-05,
+      "loss": 0.1737,
+      "step": 18204
+    },
+    {
+      "epoch": 0.29128,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.144516129032259e-05,
+      "loss": 0.1397,
+      "step": 18205
+    },
+    {
+      "epoch": 0.291296,
+      "grad_norm": 1.4609375,
+      "learning_rate": 7.144354838709679e-05,
+      "loss": 0.2341,
+      "step": 18206
+    },
+    {
+      "epoch": 0.291312,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.144193548387097e-05,
+      "loss": 0.1947,
+      "step": 18207
+    },
+    {
+      "epoch": 0.291328,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.144032258064517e-05,
+      "loss": 0.151,
+      "step": 18208
+    },
+    {
+      "epoch": 0.291344,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.143870967741936e-05,
+      "loss": 0.1495,
+      "step": 18209
+    },
+    {
+      "epoch": 0.29136,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.143709677419356e-05,
+      "loss": 0.1884,
+      "step": 18210
+    },
+    {
+      "epoch": 0.291376,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.143548387096774e-05,
+      "loss": 0.1737,
+      "step": 18211
+    },
+    {
+      "epoch": 0.291392,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.143387096774193e-05,
+      "loss": 0.1936,
+      "step": 18212
+    },
+    {
+      "epoch": 0.291408,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.143225806451613e-05,
+      "loss": 0.1327,
+      "step": 18213
+    },
+    {
+      "epoch": 0.291424,
+      "grad_norm": 1.875,
+      "learning_rate": 7.143064516129032e-05,
+      "loss": 0.1544,
+      "step": 18214
+    },
+    {
+      "epoch": 0.29144,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.142903225806452e-05,
+      "loss": 0.1937,
+      "step": 18215
+    },
+    {
+      "epoch": 0.291456,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.142741935483872e-05,
+      "loss": 0.1317,
+      "step": 18216
+    },
+    {
+      "epoch": 0.291472,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.142580645161292e-05,
+      "loss": 0.1313,
+      "step": 18217
+    },
+    {
+      "epoch": 0.291488,
+      "grad_norm": 1.25,
+      "learning_rate": 7.14241935483871e-05,
+      "loss": 0.2308,
+      "step": 18218
+    },
+    {
+      "epoch": 0.291504,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.14225806451613e-05,
+      "loss": 0.1419,
+      "step": 18219
+    },
+    {
+      "epoch": 0.29152,
+      "grad_norm": 0.75,
+      "learning_rate": 7.142096774193549e-05,
+      "loss": 0.1216,
+      "step": 18220
+    },
+    {
+      "epoch": 0.291536,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.141935483870969e-05,
+      "loss": 0.1435,
+      "step": 18221
+    },
+    {
+      "epoch": 0.291552,
+      "grad_norm": 1.3359375,
+      "learning_rate": 7.141774193548387e-05,
+      "loss": 0.1545,
+      "step": 18222
+    },
+    {
+      "epoch": 0.291568,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.141612903225807e-05,
+      "loss": 0.1674,
+      "step": 18223
+    },
+    {
+      "epoch": 0.291584,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.141451612903226e-05,
+      "loss": 0.1613,
+      "step": 18224
+    },
+    {
+      "epoch": 0.2916,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.141290322580646e-05,
+      "loss": 0.1409,
+      "step": 18225
+    },
+    {
+      "epoch": 0.291616,
+      "grad_norm": 1.2109375,
+      "learning_rate": 7.141129032258064e-05,
+      "loss": 0.1749,
+      "step": 18226
+    },
+    {
+      "epoch": 0.291632,
+      "grad_norm": 1.2890625,
+      "learning_rate": 7.140967741935483e-05,
+      "loss": 0.2008,
+      "step": 18227
+    },
+    {
+      "epoch": 0.291648,
+      "grad_norm": 1.7265625,
+      "learning_rate": 7.140806451612903e-05,
+      "loss": 0.1564,
+      "step": 18228
+    },
+    {
+      "epoch": 0.291664,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.140645161290323e-05,
+      "loss": 0.1423,
+      "step": 18229
+    },
+    {
+      "epoch": 0.29168,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.140483870967743e-05,
+      "loss": 0.2241,
+      "step": 18230
+    },
+    {
+      "epoch": 0.291696,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.140322580645162e-05,
+      "loss": 0.1534,
+      "step": 18231
+    },
+    {
+      "epoch": 0.291712,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.140161290322581e-05,
+      "loss": 0.1959,
+      "step": 18232
+    },
+    {
+      "epoch": 0.291728,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.14e-05,
+      "loss": 0.1709,
+      "step": 18233
+    },
+    {
+      "epoch": 0.291744,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.13983870967742e-05,
+      "loss": 0.1668,
+      "step": 18234
+    },
+    {
+      "epoch": 0.29176,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.139677419354839e-05,
+      "loss": 0.1615,
+      "step": 18235
+    },
+    {
+      "epoch": 0.291776,
+      "grad_norm": 1.625,
+      "learning_rate": 7.139516129032259e-05,
+      "loss": 0.1799,
+      "step": 18236
+    },
+    {
+      "epoch": 0.291792,
+      "grad_norm": 1.390625,
+      "learning_rate": 7.139354838709677e-05,
+      "loss": 0.1782,
+      "step": 18237
+    },
+    {
+      "epoch": 0.291808,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.139193548387097e-05,
+      "loss": 0.172,
+      "step": 18238
+    },
+    {
+      "epoch": 0.291824,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.139032258064516e-05,
+      "loss": 0.172,
+      "step": 18239
+    },
+    {
+      "epoch": 0.29184,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.138870967741936e-05,
+      "loss": 0.1554,
+      "step": 18240
+    },
+    {
+      "epoch": 0.291856,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.138709677419356e-05,
+      "loss": 0.2273,
+      "step": 18241
+    },
+    {
+      "epoch": 0.291872,
+      "grad_norm": 1.796875,
+      "learning_rate": 7.138548387096774e-05,
+      "loss": 0.2072,
+      "step": 18242
+    },
+    {
+      "epoch": 0.291888,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.138387096774194e-05,
+      "loss": 0.1694,
+      "step": 18243
+    },
+    {
+      "epoch": 0.291904,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.138225806451613e-05,
+      "loss": 0.1645,
+      "step": 18244
+    },
+    {
+      "epoch": 0.29192,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.138064516129033e-05,
+      "loss": 0.1632,
+      "step": 18245
+    },
+    {
+      "epoch": 0.291936,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.137903225806451e-05,
+      "loss": 0.1727,
+      "step": 18246
+    },
+    {
+      "epoch": 0.291952,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.137741935483871e-05,
+      "loss": 0.1998,
+      "step": 18247
+    },
+    {
+      "epoch": 0.291968,
+      "grad_norm": 1.359375,
+      "learning_rate": 7.13758064516129e-05,
+      "loss": 0.1463,
+      "step": 18248
+    },
+    {
+      "epoch": 0.291984,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.13741935483871e-05,
+      "loss": 0.1932,
+      "step": 18249
+    },
+    {
+      "epoch": 0.292,
+      "grad_norm": 1.3359375,
+      "learning_rate": 7.137258064516129e-05,
+      "loss": 0.2043,
+      "step": 18250
+    },
+    {
+      "epoch": 0.292016,
+      "grad_norm": 0.75,
+      "learning_rate": 7.137096774193549e-05,
+      "loss": 0.1768,
+      "step": 18251
+    },
+    {
+      "epoch": 0.292032,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.136935483870969e-05,
+      "loss": 0.1863,
+      "step": 18252
+    },
+    {
+      "epoch": 0.292048,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.136774193548388e-05,
+      "loss": 0.1679,
+      "step": 18253
+    },
+    {
+      "epoch": 0.292064,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.136612903225807e-05,
+      "loss": 0.1366,
+      "step": 18254
+    },
+    {
+      "epoch": 0.29208,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.136451612903227e-05,
+      "loss": 0.1809,
+      "step": 18255
+    },
+    {
+      "epoch": 0.292096,
+      "grad_norm": 0.427734375,
+      "learning_rate": 7.136290322580646e-05,
+      "loss": 0.1351,
+      "step": 18256
+    },
+    {
+      "epoch": 0.292112,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.136129032258066e-05,
+      "loss": 0.1919,
+      "step": 18257
+    },
+    {
+      "epoch": 0.292128,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.135967741935484e-05,
+      "loss": 0.1741,
+      "step": 18258
+    },
+    {
+      "epoch": 0.292144,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.135806451612903e-05,
+      "loss": 0.1844,
+      "step": 18259
+    },
+    {
+      "epoch": 0.29216,
+      "grad_norm": 1.2734375,
+      "learning_rate": 7.135645161290323e-05,
+      "loss": 0.1481,
+      "step": 18260
+    },
+    {
+      "epoch": 0.292176,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.135483870967741e-05,
+      "loss": 0.1466,
+      "step": 18261
+    },
+    {
+      "epoch": 0.292192,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.135322580645161e-05,
+      "loss": 0.1392,
+      "step": 18262
+    },
+    {
+      "epoch": 0.292208,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.13516129032258e-05,
+      "loss": 0.2081,
+      "step": 18263
+    },
+    {
+      "epoch": 0.292224,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.135e-05,
+      "loss": 0.1954,
+      "step": 18264
+    },
+    {
+      "epoch": 0.29224,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.13483870967742e-05,
+      "loss": 0.1317,
+      "step": 18265
+    },
+    {
+      "epoch": 0.292256,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.13467741935484e-05,
+      "loss": 0.1953,
+      "step": 18266
+    },
+    {
+      "epoch": 0.292272,
+      "grad_norm": 1.421875,
+      "learning_rate": 7.134516129032258e-05,
+      "loss": 0.2016,
+      "step": 18267
+    },
+    {
+      "epoch": 0.292288,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.134354838709678e-05,
+      "loss": 0.1989,
+      "step": 18268
+    },
+    {
+      "epoch": 0.292304,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.134193548387097e-05,
+      "loss": 0.2095,
+      "step": 18269
+    },
+    {
+      "epoch": 0.29232,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.134032258064517e-05,
+      "loss": 0.181,
+      "step": 18270
+    },
+    {
+      "epoch": 0.292336,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.133870967741936e-05,
+      "loss": 0.1672,
+      "step": 18271
+    },
+    {
+      "epoch": 0.292352,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.133709677419356e-05,
+      "loss": 0.1641,
+      "step": 18272
+    },
+    {
+      "epoch": 0.292368,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.133548387096774e-05,
+      "loss": 0.1595,
+      "step": 18273
+    },
+    {
+      "epoch": 0.292384,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.133387096774193e-05,
+      "loss": 0.1404,
+      "step": 18274
+    },
+    {
+      "epoch": 0.2924,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.133225806451613e-05,
+      "loss": 0.162,
+      "step": 18275
+    },
+    {
+      "epoch": 0.292416,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.133064516129033e-05,
+      "loss": 0.156,
+      "step": 18276
+    },
+    {
+      "epoch": 0.292432,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.132903225806453e-05,
+      "loss": 0.1357,
+      "step": 18277
+    },
+    {
+      "epoch": 0.292448,
+      "grad_norm": 1.5625,
+      "learning_rate": 7.132741935483871e-05,
+      "loss": 0.1971,
+      "step": 18278
+    },
+    {
+      "epoch": 0.292464,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.132580645161291e-05,
+      "loss": 0.1952,
+      "step": 18279
+    },
+    {
+      "epoch": 0.29248,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.13241935483871e-05,
+      "loss": 0.1699,
+      "step": 18280
+    },
+    {
+      "epoch": 0.292496,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.13225806451613e-05,
+      "loss": 0.1697,
+      "step": 18281
+    },
+    {
+      "epoch": 0.292512,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.132096774193548e-05,
+      "loss": 0.1595,
+      "step": 18282
+    },
+    {
+      "epoch": 0.292528,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.131935483870968e-05,
+      "loss": 0.1721,
+      "step": 18283
+    },
+    {
+      "epoch": 0.292544,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.131774193548387e-05,
+      "loss": 0.1674,
+      "step": 18284
+    },
+    {
+      "epoch": 0.29256,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.131612903225807e-05,
+      "loss": 0.1705,
+      "step": 18285
+    },
+    {
+      "epoch": 0.292576,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.131451612903226e-05,
+      "loss": 0.1714,
+      "step": 18286
+    },
+    {
+      "epoch": 0.292592,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.131290322580646e-05,
+      "loss": 0.1537,
+      "step": 18287
+    },
+    {
+      "epoch": 0.292608,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.131129032258064e-05,
+      "loss": 0.1981,
+      "step": 18288
+    },
+    {
+      "epoch": 0.292624,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.130967741935484e-05,
+      "loss": 0.1505,
+      "step": 18289
+    },
+    {
+      "epoch": 0.29264,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.130806451612904e-05,
+      "loss": 0.1639,
+      "step": 18290
+    },
+    {
+      "epoch": 0.292656,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.130645161290323e-05,
+      "loss": 0.1793,
+      "step": 18291
+    },
+    {
+      "epoch": 0.292672,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.130483870967743e-05,
+      "loss": 0.1411,
+      "step": 18292
+    },
+    {
+      "epoch": 0.292688,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.130322580645161e-05,
+      "loss": 0.1648,
+      "step": 18293
+    },
+    {
+      "epoch": 0.292704,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.130161290322581e-05,
+      "loss": 0.1535,
+      "step": 18294
+    },
+    {
+      "epoch": 0.29272,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.13e-05,
+      "loss": 0.2218,
+      "step": 18295
+    },
+    {
+      "epoch": 0.292736,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.12983870967742e-05,
+      "loss": 0.1766,
+      "step": 18296
+    },
+    {
+      "epoch": 0.292752,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.129677419354838e-05,
+      "loss": 0.1929,
+      "step": 18297
+    },
+    {
+      "epoch": 0.292768,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.129516129032258e-05,
+      "loss": 0.1566,
+      "step": 18298
+    },
+    {
+      "epoch": 0.292784,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.129354838709677e-05,
+      "loss": 0.2205,
+      "step": 18299
+    },
+    {
+      "epoch": 0.2928,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.129193548387097e-05,
+      "loss": 0.1763,
+      "step": 18300
+    },
+    {
+      "epoch": 0.292816,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.129032258064517e-05,
+      "loss": 0.1623,
+      "step": 18301
+    },
+    {
+      "epoch": 0.292832,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.128870967741937e-05,
+      "loss": 0.2241,
+      "step": 18302
+    },
+    {
+      "epoch": 0.292848,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.128709677419355e-05,
+      "loss": 0.1908,
+      "step": 18303
+    },
+    {
+      "epoch": 0.292864,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.128548387096775e-05,
+      "loss": 0.1597,
+      "step": 18304
+    },
+    {
+      "epoch": 0.29288,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.128387096774194e-05,
+      "loss": 0.1787,
+      "step": 18305
+    },
+    {
+      "epoch": 0.292896,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.128225806451613e-05,
+      "loss": 0.1806,
+      "step": 18306
+    },
+    {
+      "epoch": 0.292912,
+      "grad_norm": 0.875,
+      "learning_rate": 7.128064516129033e-05,
+      "loss": 0.1528,
+      "step": 18307
+    },
+    {
+      "epoch": 0.292928,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.127903225806451e-05,
+      "loss": 0.1752,
+      "step": 18308
+    },
+    {
+      "epoch": 0.292944,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.127741935483871e-05,
+      "loss": 0.2048,
+      "step": 18309
+    },
+    {
+      "epoch": 0.29296,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.12758064516129e-05,
+      "loss": 0.1846,
+      "step": 18310
+    },
+    {
+      "epoch": 0.292976,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.12741935483871e-05,
+      "loss": 0.1872,
+      "step": 18311
+    },
+    {
+      "epoch": 0.292992,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.12725806451613e-05,
+      "loss": 0.1797,
+      "step": 18312
+    },
+    {
+      "epoch": 0.293008,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.12709677419355e-05,
+      "loss": 0.2054,
+      "step": 18313
+    },
+    {
+      "epoch": 0.293024,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.126935483870968e-05,
+      "loss": 0.1495,
+      "step": 18314
+    },
+    {
+      "epoch": 0.29304,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.126774193548388e-05,
+      "loss": 0.2059,
+      "step": 18315
+    },
+    {
+      "epoch": 0.293056,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.126612903225807e-05,
+      "loss": 0.172,
+      "step": 18316
+    },
+    {
+      "epoch": 0.293072,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.126451612903227e-05,
+      "loss": 0.1304,
+      "step": 18317
+    },
+    {
+      "epoch": 0.293088,
+      "grad_norm": 1.3046875,
+      "learning_rate": 7.126290322580645e-05,
+      "loss": 0.1944,
+      "step": 18318
+    },
+    {
+      "epoch": 0.293104,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.126129032258065e-05,
+      "loss": 0.1764,
+      "step": 18319
+    },
+    {
+      "epoch": 0.29312,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.125967741935484e-05,
+      "loss": 0.1931,
+      "step": 18320
+    },
+    {
+      "epoch": 0.293136,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.125806451612903e-05,
+      "loss": 0.1357,
+      "step": 18321
+    },
+    {
+      "epoch": 0.293152,
+      "grad_norm": 1.3046875,
+      "learning_rate": 7.125645161290323e-05,
+      "loss": 0.1848,
+      "step": 18322
+    },
+    {
+      "epoch": 0.293168,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.125483870967741e-05,
+      "loss": 0.182,
+      "step": 18323
+    },
+    {
+      "epoch": 0.293184,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.125322580645161e-05,
+      "loss": 0.205,
+      "step": 18324
+    },
+    {
+      "epoch": 0.2932,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.125161290322581e-05,
+      "loss": 0.1314,
+      "step": 18325
+    },
+    {
+      "epoch": 0.293216,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.125000000000001e-05,
+      "loss": 0.1755,
+      "step": 18326
+    },
+    {
+      "epoch": 0.293232,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.12483870967742e-05,
+      "loss": 0.1819,
+      "step": 18327
+    },
+    {
+      "epoch": 0.293248,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.12467741935484e-05,
+      "loss": 0.1734,
+      "step": 18328
+    },
+    {
+      "epoch": 0.293264,
+      "grad_norm": 1.0,
+      "learning_rate": 7.124516129032258e-05,
+      "loss": 0.1672,
+      "step": 18329
+    },
+    {
+      "epoch": 0.29328,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.124354838709678e-05,
+      "loss": 0.1786,
+      "step": 18330
+    },
+    {
+      "epoch": 0.293296,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.124193548387097e-05,
+      "loss": 0.1769,
+      "step": 18331
+    },
+    {
+      "epoch": 0.293312,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.124032258064517e-05,
+      "loss": 0.2191,
+      "step": 18332
+    },
+    {
+      "epoch": 0.293328,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.123870967741935e-05,
+      "loss": 0.1863,
+      "step": 18333
+    },
+    {
+      "epoch": 0.293344,
+      "grad_norm": 1.28125,
+      "learning_rate": 7.123709677419355e-05,
+      "loss": 0.1731,
+      "step": 18334
+    },
+    {
+      "epoch": 0.29336,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.123548387096774e-05,
+      "loss": 0.1599,
+      "step": 18335
+    },
+    {
+      "epoch": 0.293376,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.123387096774194e-05,
+      "loss": 0.1913,
+      "step": 18336
+    },
+    {
+      "epoch": 0.293392,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.123225806451614e-05,
+      "loss": 0.1537,
+      "step": 18337
+    },
+    {
+      "epoch": 0.293408,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.123064516129032e-05,
+      "loss": 0.1357,
+      "step": 18338
+    },
+    {
+      "epoch": 0.293424,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.122903225806452e-05,
+      "loss": 0.189,
+      "step": 18339
+    },
+    {
+      "epoch": 0.29344,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.122741935483871e-05,
+      "loss": 0.1782,
+      "step": 18340
+    },
+    {
+      "epoch": 0.293456,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.122580645161291e-05,
+      "loss": 0.1795,
+      "step": 18341
+    },
+    {
+      "epoch": 0.293472,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.12241935483871e-05,
+      "loss": 0.1927,
+      "step": 18342
+    },
+    {
+      "epoch": 0.293488,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.12225806451613e-05,
+      "loss": 0.1698,
+      "step": 18343
+    },
+    {
+      "epoch": 0.293504,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.122096774193548e-05,
+      "loss": 0.1916,
+      "step": 18344
+    },
+    {
+      "epoch": 0.29352,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.121935483870968e-05,
+      "loss": 0.1875,
+      "step": 18345
+    },
+    {
+      "epoch": 0.293536,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.121774193548387e-05,
+      "loss": 0.1742,
+      "step": 18346
+    },
+    {
+      "epoch": 0.293552,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.121612903225807e-05,
+      "loss": 0.1543,
+      "step": 18347
+    },
+    {
+      "epoch": 0.293568,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.121451612903227e-05,
+      "loss": 0.1394,
+      "step": 18348
+    },
+    {
+      "epoch": 0.293584,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.121290322580647e-05,
+      "loss": 0.151,
+      "step": 18349
+    },
+    {
+      "epoch": 0.2936,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.121129032258065e-05,
+      "loss": 0.204,
+      "step": 18350
+    },
+    {
+      "epoch": 0.293616,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.120967741935484e-05,
+      "loss": 0.1535,
+      "step": 18351
+    },
+    {
+      "epoch": 0.293632,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.120806451612904e-05,
+      "loss": 0.1671,
+      "step": 18352
+    },
+    {
+      "epoch": 0.293648,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.120645161290322e-05,
+      "loss": 0.1567,
+      "step": 18353
+    },
+    {
+      "epoch": 0.293664,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.120483870967742e-05,
+      "loss": 0.1329,
+      "step": 18354
+    },
+    {
+      "epoch": 0.29368,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.120322580645161e-05,
+      "loss": 0.1855,
+      "step": 18355
+    },
+    {
+      "epoch": 0.293696,
+      "grad_norm": 1.2109375,
+      "learning_rate": 7.120161290322581e-05,
+      "loss": 0.1429,
+      "step": 18356
+    },
+    {
+      "epoch": 0.293712,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.12e-05,
+      "loss": 0.1794,
+      "step": 18357
+    },
+    {
+      "epoch": 0.293728,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.11983870967742e-05,
+      "loss": 0.2047,
+      "step": 18358
+    },
+    {
+      "epoch": 0.293744,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.119677419354838e-05,
+      "loss": 0.1586,
+      "step": 18359
+    },
+    {
+      "epoch": 0.29376,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.119516129032258e-05,
+      "loss": 0.1539,
+      "step": 18360
+    },
+    {
+      "epoch": 0.293776,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.119354838709678e-05,
+      "loss": 0.1645,
+      "step": 18361
+    },
+    {
+      "epoch": 0.293792,
+      "grad_norm": 0.625,
+      "learning_rate": 7.119193548387098e-05,
+      "loss": 0.1902,
+      "step": 18362
+    },
+    {
+      "epoch": 0.293808,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.119032258064517e-05,
+      "loss": 0.1923,
+      "step": 18363
+    },
+    {
+      "epoch": 0.293824,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.118870967741937e-05,
+      "loss": 0.1686,
+      "step": 18364
+    },
+    {
+      "epoch": 0.29384,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.118709677419355e-05,
+      "loss": 0.158,
+      "step": 18365
+    },
+    {
+      "epoch": 0.293856,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.118548387096775e-05,
+      "loss": 0.1624,
+      "step": 18366
+    },
+    {
+      "epoch": 0.293872,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.118387096774194e-05,
+      "loss": 0.1412,
+      "step": 18367
+    },
+    {
+      "epoch": 0.293888,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.118225806451612e-05,
+      "loss": 0.1453,
+      "step": 18368
+    },
+    {
+      "epoch": 0.293904,
+      "grad_norm": 1.0,
+      "learning_rate": 7.118064516129032e-05,
+      "loss": 0.2033,
+      "step": 18369
+    },
+    {
+      "epoch": 0.29392,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.117903225806451e-05,
+      "loss": 0.1504,
+      "step": 18370
+    },
+    {
+      "epoch": 0.293936,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.117741935483871e-05,
+      "loss": 0.1851,
+      "step": 18371
+    },
+    {
+      "epoch": 0.293952,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.117580645161291e-05,
+      "loss": 0.1309,
+      "step": 18372
+    },
+    {
+      "epoch": 0.293968,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.117419354838711e-05,
+      "loss": 0.1931,
+      "step": 18373
+    },
+    {
+      "epoch": 0.293984,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.11725806451613e-05,
+      "loss": 0.1978,
+      "step": 18374
+    },
+    {
+      "epoch": 0.294,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.11709677419355e-05,
+      "loss": 0.149,
+      "step": 18375
+    },
+    {
+      "epoch": 0.294016,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.116935483870968e-05,
+      "loss": 0.1497,
+      "step": 18376
+    },
+    {
+      "epoch": 0.294032,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.116774193548388e-05,
+      "loss": 0.1685,
+      "step": 18377
+    },
+    {
+      "epoch": 0.294048,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.116612903225807e-05,
+      "loss": 0.1543,
+      "step": 18378
+    },
+    {
+      "epoch": 0.294064,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.116451612903227e-05,
+      "loss": 0.175,
+      "step": 18379
+    },
+    {
+      "epoch": 0.29408,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.116290322580645e-05,
+      "loss": 0.1399,
+      "step": 18380
+    },
+    {
+      "epoch": 0.294096,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.116129032258065e-05,
+      "loss": 0.1834,
+      "step": 18381
+    },
+    {
+      "epoch": 0.294112,
+      "grad_norm": 1.25,
+      "learning_rate": 7.115967741935484e-05,
+      "loss": 0.19,
+      "step": 18382
+    },
+    {
+      "epoch": 0.294128,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.115806451612904e-05,
+      "loss": 0.185,
+      "step": 18383
+    },
+    {
+      "epoch": 0.294144,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.115645161290322e-05,
+      "loss": 0.1947,
+      "step": 18384
+    },
+    {
+      "epoch": 0.29416,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.115483870967742e-05,
+      "loss": 0.2206,
+      "step": 18385
+    },
+    {
+      "epoch": 0.294176,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.115322580645162e-05,
+      "loss": 0.158,
+      "step": 18386
+    },
+    {
+      "epoch": 0.294192,
+      "grad_norm": 1.359375,
+      "learning_rate": 7.115161290322581e-05,
+      "loss": 0.1743,
+      "step": 18387
+    },
+    {
+      "epoch": 0.294208,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.115000000000001e-05,
+      "loss": 0.1709,
+      "step": 18388
+    },
+    {
+      "epoch": 0.294224,
+      "grad_norm": 1.3359375,
+      "learning_rate": 7.11483870967742e-05,
+      "loss": 0.2089,
+      "step": 18389
+    },
+    {
+      "epoch": 0.29424,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.11467741935484e-05,
+      "loss": 0.1441,
+      "step": 18390
+    },
+    {
+      "epoch": 0.294256,
+      "grad_norm": 0.470703125,
+      "learning_rate": 7.114516129032258e-05,
+      "loss": 0.1425,
+      "step": 18391
+    },
+    {
+      "epoch": 0.294272,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.114354838709678e-05,
+      "loss": 0.1625,
+      "step": 18392
+    },
+    {
+      "epoch": 0.294288,
+      "grad_norm": 0.75,
+      "learning_rate": 7.114193548387097e-05,
+      "loss": 0.1622,
+      "step": 18393
+    },
+    {
+      "epoch": 0.294304,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.114032258064517e-05,
+      "loss": 0.209,
+      "step": 18394
+    },
+    {
+      "epoch": 0.29432,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.113870967741935e-05,
+      "loss": 0.1566,
+      "step": 18395
+    },
+    {
+      "epoch": 0.294336,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.113709677419355e-05,
+      "loss": 0.1699,
+      "step": 18396
+    },
+    {
+      "epoch": 0.294352,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.113548387096775e-05,
+      "loss": 0.1823,
+      "step": 18397
+    },
+    {
+      "epoch": 0.294368,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.113387096774194e-05,
+      "loss": 0.1728,
+      "step": 18398
+    },
+    {
+      "epoch": 0.294384,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.113225806451614e-05,
+      "loss": 0.1846,
+      "step": 18399
+    },
+    {
+      "epoch": 0.2944,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.113064516129032e-05,
+      "loss": 0.1978,
+      "step": 18400
+    },
+    {
+      "epoch": 0.294416,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.112903225806452e-05,
+      "loss": 0.1285,
+      "step": 18401
+    },
+    {
+      "epoch": 0.294432,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.112741935483871e-05,
+      "loss": 0.1482,
+      "step": 18402
+    },
+    {
+      "epoch": 0.294448,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.112580645161291e-05,
+      "loss": 0.1851,
+      "step": 18403
+    },
+    {
+      "epoch": 0.294464,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.11241935483871e-05,
+      "loss": 0.152,
+      "step": 18404
+    },
+    {
+      "epoch": 0.29448,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.11225806451613e-05,
+      "loss": 0.1613,
+      "step": 18405
+    },
+    {
+      "epoch": 0.294496,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.112096774193548e-05,
+      "loss": 0.1551,
+      "step": 18406
+    },
+    {
+      "epoch": 0.294512,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.111935483870968e-05,
+      "loss": 0.1423,
+      "step": 18407
+    },
+    {
+      "epoch": 0.294528,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.111774193548388e-05,
+      "loss": 0.1425,
+      "step": 18408
+    },
+    {
+      "epoch": 0.294544,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.111612903225808e-05,
+      "loss": 0.1416,
+      "step": 18409
+    },
+    {
+      "epoch": 0.29456,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.111451612903226e-05,
+      "loss": 0.1486,
+      "step": 18410
+    },
+    {
+      "epoch": 0.294576,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.111290322580646e-05,
+      "loss": 0.1955,
+      "step": 18411
+    },
+    {
+      "epoch": 0.294592,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.111129032258065e-05,
+      "loss": 0.1313,
+      "step": 18412
+    },
+    {
+      "epoch": 0.294608,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.110967741935485e-05,
+      "loss": 0.1533,
+      "step": 18413
+    },
+    {
+      "epoch": 0.294624,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.110806451612904e-05,
+      "loss": 0.1613,
+      "step": 18414
+    },
+    {
+      "epoch": 0.29464,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.110645161290322e-05,
+      "loss": 0.1939,
+      "step": 18415
+    },
+    {
+      "epoch": 0.294656,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.110483870967742e-05,
+      "loss": 0.1703,
+      "step": 18416
+    },
+    {
+      "epoch": 0.294672,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.110322580645161e-05,
+      "loss": 0.1891,
+      "step": 18417
+    },
+    {
+      "epoch": 0.294688,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.110161290322581e-05,
+      "loss": 0.1863,
+      "step": 18418
+    },
+    {
+      "epoch": 0.294704,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.11e-05,
+      "loss": 0.1348,
+      "step": 18419
+    },
+    {
+      "epoch": 0.29472,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.109838709677419e-05,
+      "loss": 0.1972,
+      "step": 18420
+    },
+    {
+      "epoch": 0.294736,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.109677419354839e-05,
+      "loss": 0.1626,
+      "step": 18421
+    },
+    {
+      "epoch": 0.294752,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.109516129032259e-05,
+      "loss": 0.1782,
+      "step": 18422
+    },
+    {
+      "epoch": 0.294768,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.109354838709678e-05,
+      "loss": 0.1477,
+      "step": 18423
+    },
+    {
+      "epoch": 0.294784,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.109193548387098e-05,
+      "loss": 0.198,
+      "step": 18424
+    },
+    {
+      "epoch": 0.2948,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.109032258064516e-05,
+      "loss": 0.1805,
+      "step": 18425
+    },
+    {
+      "epoch": 0.294816,
+      "grad_norm": 0.875,
+      "learning_rate": 7.108870967741936e-05,
+      "loss": 0.14,
+      "step": 18426
+    },
+    {
+      "epoch": 0.294832,
+      "grad_norm": 1.4296875,
+      "learning_rate": 7.108709677419355e-05,
+      "loss": 0.2291,
+      "step": 18427
+    },
+    {
+      "epoch": 0.294848,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.108548387096775e-05,
+      "loss": 0.1678,
+      "step": 18428
+    },
+    {
+      "epoch": 0.294864,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.108387096774194e-05,
+      "loss": 0.1965,
+      "step": 18429
+    },
+    {
+      "epoch": 0.29488,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.108225806451612e-05,
+      "loss": 0.1775,
+      "step": 18430
+    },
+    {
+      "epoch": 0.294896,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.108064516129032e-05,
+      "loss": 0.1942,
+      "step": 18431
+    },
+    {
+      "epoch": 0.294912,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.107903225806452e-05,
+      "loss": 0.1638,
+      "step": 18432
+    },
+    {
+      "epoch": 0.294928,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.107741935483872e-05,
+      "loss": 0.1797,
+      "step": 18433
+    },
+    {
+      "epoch": 0.294944,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.10758064516129e-05,
+      "loss": 0.1737,
+      "step": 18434
+    },
+    {
+      "epoch": 0.29496,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.10741935483871e-05,
+      "loss": 0.1614,
+      "step": 18435
+    },
+    {
+      "epoch": 0.294976,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.107258064516129e-05,
+      "loss": 0.1975,
+      "step": 18436
+    },
+    {
+      "epoch": 0.294992,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.107096774193549e-05,
+      "loss": 0.1249,
+      "step": 18437
+    },
+    {
+      "epoch": 0.295008,
+      "grad_norm": 0.625,
+      "learning_rate": 7.106935483870968e-05,
+      "loss": 0.1649,
+      "step": 18438
+    },
+    {
+      "epoch": 0.295024,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.106774193548388e-05,
+      "loss": 0.1727,
+      "step": 18439
+    },
+    {
+      "epoch": 0.29504,
+      "grad_norm": 0.875,
+      "learning_rate": 7.106612903225806e-05,
+      "loss": 0.1713,
+      "step": 18440
+    },
+    {
+      "epoch": 0.295056,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.106451612903226e-05,
+      "loss": 0.1883,
+      "step": 18441
+    },
+    {
+      "epoch": 0.295072,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.106290322580645e-05,
+      "loss": 0.1655,
+      "step": 18442
+    },
+    {
+      "epoch": 0.295088,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.106129032258065e-05,
+      "loss": 0.139,
+      "step": 18443
+    },
+    {
+      "epoch": 0.295104,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.105967741935485e-05,
+      "loss": 0.2245,
+      "step": 18444
+    },
+    {
+      "epoch": 0.29512,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.105806451612903e-05,
+      "loss": 0.2088,
+      "step": 18445
+    },
+    {
+      "epoch": 0.295136,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.105645161290323e-05,
+      "loss": 0.1885,
+      "step": 18446
+    },
+    {
+      "epoch": 0.295152,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.105483870967742e-05,
+      "loss": 0.1535,
+      "step": 18447
+    },
+    {
+      "epoch": 0.295168,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.105322580645162e-05,
+      "loss": 0.1378,
+      "step": 18448
+    },
+    {
+      "epoch": 0.295184,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.10516129032258e-05,
+      "loss": 0.1405,
+      "step": 18449
+    },
+    {
+      "epoch": 0.2952,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.105e-05,
+      "loss": 0.1925,
+      "step": 18450
+    },
+    {
+      "epoch": 0.295216,
+      "grad_norm": 0.625,
+      "learning_rate": 7.104838709677419e-05,
+      "loss": 0.1364,
+      "step": 18451
+    },
+    {
+      "epoch": 0.295232,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.104677419354839e-05,
+      "loss": 0.1782,
+      "step": 18452
+    },
+    {
+      "epoch": 0.295248,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.104516129032258e-05,
+      "loss": 0.1724,
+      "step": 18453
+    },
+    {
+      "epoch": 0.295264,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.104354838709678e-05,
+      "loss": 0.2078,
+      "step": 18454
+    },
+    {
+      "epoch": 0.29528,
+      "grad_norm": 0.875,
+      "learning_rate": 7.104193548387096e-05,
+      "loss": 0.1968,
+      "step": 18455
+    },
+    {
+      "epoch": 0.295296,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.104032258064516e-05,
+      "loss": 0.1727,
+      "step": 18456
+    },
+    {
+      "epoch": 0.295312,
+      "grad_norm": 1.375,
+      "learning_rate": 7.103870967741936e-05,
+      "loss": 0.2045,
+      "step": 18457
+    },
+    {
+      "epoch": 0.295328,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.103709677419356e-05,
+      "loss": 0.1472,
+      "step": 18458
+    },
+    {
+      "epoch": 0.295344,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.103548387096775e-05,
+      "loss": 0.1423,
+      "step": 18459
+    },
+    {
+      "epoch": 0.29536,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.103387096774193e-05,
+      "loss": 0.1656,
+      "step": 18460
+    },
+    {
+      "epoch": 0.295376,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.103225806451613e-05,
+      "loss": 0.1858,
+      "step": 18461
+    },
+    {
+      "epoch": 0.295392,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.103064516129032e-05,
+      "loss": 0.1792,
+      "step": 18462
+    },
+    {
+      "epoch": 0.295408,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.102903225806452e-05,
+      "loss": 0.1662,
+      "step": 18463
+    },
+    {
+      "epoch": 0.295424,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.10274193548387e-05,
+      "loss": 0.1819,
+      "step": 18464
+    },
+    {
+      "epoch": 0.29544,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.10258064516129e-05,
+      "loss": 0.1496,
+      "step": 18465
+    },
+    {
+      "epoch": 0.295456,
+      "grad_norm": 1.3828125,
+      "learning_rate": 7.102419354838709e-05,
+      "loss": 0.1808,
+      "step": 18466
+    },
+    {
+      "epoch": 0.295472,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.102258064516129e-05,
+      "loss": 0.1765,
+      "step": 18467
+    },
+    {
+      "epoch": 0.295488,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.102096774193549e-05,
+      "loss": 0.1572,
+      "step": 18468
+    },
+    {
+      "epoch": 0.295504,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.101935483870969e-05,
+      "loss": 0.1612,
+      "step": 18469
+    },
+    {
+      "epoch": 0.29552,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.101774193548388e-05,
+      "loss": 0.158,
+      "step": 18470
+    },
+    {
+      "epoch": 0.295536,
+      "grad_norm": 1.2109375,
+      "learning_rate": 7.101612903225808e-05,
+      "loss": 0.1794,
+      "step": 18471
+    },
+    {
+      "epoch": 0.295552,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.101451612903226e-05,
+      "loss": 0.1643,
+      "step": 18472
+    },
+    {
+      "epoch": 0.295568,
+      "grad_norm": 1.0,
+      "learning_rate": 7.101290322580646e-05,
+      "loss": 0.1653,
+      "step": 18473
+    },
+    {
+      "epoch": 0.295584,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.101129032258065e-05,
+      "loss": 0.1782,
+      "step": 18474
+    },
+    {
+      "epoch": 0.2956,
+      "grad_norm": 0.625,
+      "learning_rate": 7.100967741935485e-05,
+      "loss": 0.1654,
+      "step": 18475
+    },
+    {
+      "epoch": 0.295616,
+      "grad_norm": 1.2734375,
+      "learning_rate": 7.100806451612903e-05,
+      "loss": 0.1873,
+      "step": 18476
+    },
+    {
+      "epoch": 0.295632,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.100645161290322e-05,
+      "loss": 0.1892,
+      "step": 18477
+    },
+    {
+      "epoch": 0.295648,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.100483870967742e-05,
+      "loss": 0.1457,
+      "step": 18478
+    },
+    {
+      "epoch": 0.295664,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.10032258064516e-05,
+      "loss": 0.1777,
+      "step": 18479
+    },
+    {
+      "epoch": 0.29568,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.10016129032258e-05,
+      "loss": 0.1242,
+      "step": 18480
+    },
+    {
+      "epoch": 0.295696,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.1e-05,
+      "loss": 0.1628,
+      "step": 18481
+    },
+    {
+      "epoch": 0.295712,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.09983870967742e-05,
+      "loss": 0.1696,
+      "step": 18482
+    },
+    {
+      "epoch": 0.295728,
+      "grad_norm": 1.4453125,
+      "learning_rate": 7.099677419354839e-05,
+      "loss": 0.2041,
+      "step": 18483
+    },
+    {
+      "epoch": 0.295744,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.099516129032259e-05,
+      "loss": 0.1246,
+      "step": 18484
+    },
+    {
+      "epoch": 0.29576,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.099354838709678e-05,
+      "loss": 0.2313,
+      "step": 18485
+    },
+    {
+      "epoch": 0.295776,
+      "grad_norm": 1.359375,
+      "learning_rate": 7.099193548387098e-05,
+      "loss": 0.2025,
+      "step": 18486
+    },
+    {
+      "epoch": 0.295792,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.099032258064516e-05,
+      "loss": 0.1445,
+      "step": 18487
+    },
+    {
+      "epoch": 0.295808,
+      "grad_norm": 0.46875,
+      "learning_rate": 7.098870967741936e-05,
+      "loss": 0.1511,
+      "step": 18488
+    },
+    {
+      "epoch": 0.295824,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.098709677419355e-05,
+      "loss": 0.1766,
+      "step": 18489
+    },
+    {
+      "epoch": 0.29584,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.098548387096775e-05,
+      "loss": 0.1453,
+      "step": 18490
+    },
+    {
+      "epoch": 0.295856,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.098387096774193e-05,
+      "loss": 0.1779,
+      "step": 18491
+    },
+    {
+      "epoch": 0.295872,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.098225806451613e-05,
+      "loss": 0.1481,
+      "step": 18492
+    },
+    {
+      "epoch": 0.295888,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.098064516129033e-05,
+      "loss": 0.1429,
+      "step": 18493
+    },
+    {
+      "epoch": 0.295904,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.097903225806452e-05,
+      "loss": 0.1743,
+      "step": 18494
+    },
+    {
+      "epoch": 0.29592,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.097741935483872e-05,
+      "loss": 0.1953,
+      "step": 18495
+    },
+    {
+      "epoch": 0.295936,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.09758064516129e-05,
+      "loss": 0.1623,
+      "step": 18496
+    },
+    {
+      "epoch": 0.295952,
+      "grad_norm": 1.3046875,
+      "learning_rate": 7.09741935483871e-05,
+      "loss": 0.1947,
+      "step": 18497
+    },
+    {
+      "epoch": 0.295968,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.097258064516129e-05,
+      "loss": 0.1566,
+      "step": 18498
+    },
+    {
+      "epoch": 0.295984,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.097096774193549e-05,
+      "loss": 0.1749,
+      "step": 18499
+    },
+    {
+      "epoch": 0.296,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.096935483870968e-05,
+      "loss": 0.165,
+      "step": 18500
+    },
+    {
+      "epoch": 0.296016,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.096774193548388e-05,
+      "loss": 0.1671,
+      "step": 18501
+    },
+    {
+      "epoch": 0.296032,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.096612903225806e-05,
+      "loss": 0.2053,
+      "step": 18502
+    },
+    {
+      "epoch": 0.296048,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.096451612903226e-05,
+      "loss": 0.2048,
+      "step": 18503
+    },
+    {
+      "epoch": 0.296064,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.096290322580646e-05,
+      "loss": 0.1551,
+      "step": 18504
+    },
+    {
+      "epoch": 0.29608,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.096129032258066e-05,
+      "loss": 0.1596,
+      "step": 18505
+    },
+    {
+      "epoch": 0.296096,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.095967741935485e-05,
+      "loss": 0.1858,
+      "step": 18506
+    },
+    {
+      "epoch": 0.296112,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.095806451612903e-05,
+      "loss": 0.1581,
+      "step": 18507
+    },
+    {
+      "epoch": 0.296128,
+      "grad_norm": 1.0,
+      "learning_rate": 7.095645161290323e-05,
+      "loss": 0.1899,
+      "step": 18508
+    },
+    {
+      "epoch": 0.296144,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.095483870967742e-05,
+      "loss": 0.1998,
+      "step": 18509
+    },
+    {
+      "epoch": 0.29616,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.095322580645162e-05,
+      "loss": 0.1553,
+      "step": 18510
+    },
+    {
+      "epoch": 0.296176,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.09516129032258e-05,
+      "loss": 0.1914,
+      "step": 18511
+    },
+    {
+      "epoch": 0.296192,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.095e-05,
+      "loss": 0.1605,
+      "step": 18512
+    },
+    {
+      "epoch": 0.296208,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.094838709677419e-05,
+      "loss": 0.17,
+      "step": 18513
+    },
+    {
+      "epoch": 0.296224,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.094677419354839e-05,
+      "loss": 0.183,
+      "step": 18514
+    },
+    {
+      "epoch": 0.29624,
+      "grad_norm": 0.75,
+      "learning_rate": 7.094516129032258e-05,
+      "loss": 0.1479,
+      "step": 18515
+    },
+    {
+      "epoch": 0.296256,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.094354838709677e-05,
+      "loss": 0.1636,
+      "step": 18516
+    },
+    {
+      "epoch": 0.296272,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.094193548387097e-05,
+      "loss": 0.1948,
+      "step": 18517
+    },
+    {
+      "epoch": 0.296288,
+      "grad_norm": 1.34375,
+      "learning_rate": 7.094032258064517e-05,
+      "loss": 0.1513,
+      "step": 18518
+    },
+    {
+      "epoch": 0.296304,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.093870967741936e-05,
+      "loss": 0.1249,
+      "step": 18519
+    },
+    {
+      "epoch": 0.29632,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.093709677419356e-05,
+      "loss": 0.1423,
+      "step": 18520
+    },
+    {
+      "epoch": 0.296336,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.093548387096775e-05,
+      "loss": 0.1925,
+      "step": 18521
+    },
+    {
+      "epoch": 0.296352,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.093387096774195e-05,
+      "loss": 0.1733,
+      "step": 18522
+    },
+    {
+      "epoch": 0.296368,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.093225806451613e-05,
+      "loss": 0.1431,
+      "step": 18523
+    },
+    {
+      "epoch": 0.296384,
+      "grad_norm": 0.453125,
+      "learning_rate": 7.093064516129032e-05,
+      "loss": 0.149,
+      "step": 18524
+    },
+    {
+      "epoch": 0.2964,
+      "grad_norm": 1.2578125,
+      "learning_rate": 7.092903225806452e-05,
+      "loss": 0.1716,
+      "step": 18525
+    },
+    {
+      "epoch": 0.296416,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.09274193548387e-05,
+      "loss": 0.1849,
+      "step": 18526
+    },
+    {
+      "epoch": 0.296432,
+      "grad_norm": 1.125,
+      "learning_rate": 7.09258064516129e-05,
+      "loss": 0.2074,
+      "step": 18527
+    },
+    {
+      "epoch": 0.296448,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.09241935483871e-05,
+      "loss": 0.158,
+      "step": 18528
+    },
+    {
+      "epoch": 0.296464,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.09225806451613e-05,
+      "loss": 0.1747,
+      "step": 18529
+    },
+    {
+      "epoch": 0.29648,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.092096774193549e-05,
+      "loss": 0.1704,
+      "step": 18530
+    },
+    {
+      "epoch": 0.296496,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.091935483870969e-05,
+      "loss": 0.1643,
+      "step": 18531
+    },
+    {
+      "epoch": 0.296512,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.091774193548387e-05,
+      "loss": 0.2001,
+      "step": 18532
+    },
+    {
+      "epoch": 0.296528,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.091612903225807e-05,
+      "loss": 0.1721,
+      "step": 18533
+    },
+    {
+      "epoch": 0.296544,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.091451612903226e-05,
+      "loss": 0.2122,
+      "step": 18534
+    },
+    {
+      "epoch": 0.29656,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.091290322580646e-05,
+      "loss": 0.1581,
+      "step": 18535
+    },
+    {
+      "epoch": 0.296576,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.091129032258065e-05,
+      "loss": 0.2062,
+      "step": 18536
+    },
+    {
+      "epoch": 0.296592,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.090967741935485e-05,
+      "loss": 0.1845,
+      "step": 18537
+    },
+    {
+      "epoch": 0.296608,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.090806451612903e-05,
+      "loss": 0.1676,
+      "step": 18538
+    },
+    {
+      "epoch": 0.296624,
+      "grad_norm": 1.578125,
+      "learning_rate": 7.090645161290323e-05,
+      "loss": 0.1951,
+      "step": 18539
+    },
+    {
+      "epoch": 0.29664,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.090483870967742e-05,
+      "loss": 0.2138,
+      "step": 18540
+    },
+    {
+      "epoch": 0.296656,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.090322580645162e-05,
+      "loss": 0.1696,
+      "step": 18541
+    },
+    {
+      "epoch": 0.296672,
+      "grad_norm": 1.0,
+      "learning_rate": 7.090161290322582e-05,
+      "loss": 0.1473,
+      "step": 18542
+    },
+    {
+      "epoch": 0.296688,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.09e-05,
+      "loss": 0.2249,
+      "step": 18543
+    },
+    {
+      "epoch": 0.296704,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.08983870967742e-05,
+      "loss": 0.1509,
+      "step": 18544
+    },
+    {
+      "epoch": 0.29672,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.089677419354839e-05,
+      "loss": 0.1465,
+      "step": 18545
+    },
+    {
+      "epoch": 0.296736,
+      "grad_norm": 0.625,
+      "learning_rate": 7.089516129032259e-05,
+      "loss": 0.1561,
+      "step": 18546
+    },
+    {
+      "epoch": 0.296752,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.089354838709677e-05,
+      "loss": 0.188,
+      "step": 18547
+    },
+    {
+      "epoch": 0.296768,
+      "grad_norm": 0.625,
+      "learning_rate": 7.089193548387097e-05,
+      "loss": 0.1406,
+      "step": 18548
+    },
+    {
+      "epoch": 0.296784,
+      "grad_norm": 0.625,
+      "learning_rate": 7.089032258064516e-05,
+      "loss": 0.1576,
+      "step": 18549
+    },
+    {
+      "epoch": 0.2968,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.088870967741936e-05,
+      "loss": 0.1477,
+      "step": 18550
+    },
+    {
+      "epoch": 0.296816,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.088709677419355e-05,
+      "loss": 0.1837,
+      "step": 18551
+    },
+    {
+      "epoch": 0.296832,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.088548387096774e-05,
+      "loss": 0.2196,
+      "step": 18552
+    },
+    {
+      "epoch": 0.296848,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.088387096774194e-05,
+      "loss": 0.1809,
+      "step": 18553
+    },
+    {
+      "epoch": 0.296864,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.088225806451613e-05,
+      "loss": 0.1442,
+      "step": 18554
+    },
+    {
+      "epoch": 0.29688,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.088064516129033e-05,
+      "loss": 0.2151,
+      "step": 18555
+    },
+    {
+      "epoch": 0.296896,
+      "grad_norm": 0.451171875,
+      "learning_rate": 7.087903225806452e-05,
+      "loss": 0.1382,
+      "step": 18556
+    },
+    {
+      "epoch": 0.296912,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.087741935483872e-05,
+      "loss": 0.1757,
+      "step": 18557
+    },
+    {
+      "epoch": 0.296928,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.08758064516129e-05,
+      "loss": 0.1816,
+      "step": 18558
+    },
+    {
+      "epoch": 0.296944,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.08741935483871e-05,
+      "loss": 0.1649,
+      "step": 18559
+    },
+    {
+      "epoch": 0.29696,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.087258064516129e-05,
+      "loss": 0.1553,
+      "step": 18560
+    },
+    {
+      "epoch": 0.296976,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.087096774193549e-05,
+      "loss": 0.1641,
+      "step": 18561
+    },
+    {
+      "epoch": 0.296992,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.086935483870967e-05,
+      "loss": 0.1559,
+      "step": 18562
+    },
+    {
+      "epoch": 0.297008,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.086774193548387e-05,
+      "loss": 0.1785,
+      "step": 18563
+    },
+    {
+      "epoch": 0.297024,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.086612903225807e-05,
+      "loss": 0.2025,
+      "step": 18564
+    },
+    {
+      "epoch": 0.29704,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.086451612903227e-05,
+      "loss": 0.148,
+      "step": 18565
+    },
+    {
+      "epoch": 0.297056,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.086290322580646e-05,
+      "loss": 0.1941,
+      "step": 18566
+    },
+    {
+      "epoch": 0.297072,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.086129032258066e-05,
+      "loss": 0.1987,
+      "step": 18567
+    },
+    {
+      "epoch": 0.297088,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.085967741935484e-05,
+      "loss": 0.1825,
+      "step": 18568
+    },
+    {
+      "epoch": 0.297104,
+      "grad_norm": 0.75,
+      "learning_rate": 7.085806451612903e-05,
+      "loss": 0.1491,
+      "step": 18569
+    },
+    {
+      "epoch": 0.29712,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.085645161290323e-05,
+      "loss": 0.2205,
+      "step": 18570
+    },
+    {
+      "epoch": 0.297136,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.085483870967742e-05,
+      "loss": 0.1619,
+      "step": 18571
+    },
+    {
+      "epoch": 0.297152,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.085322580645162e-05,
+      "loss": 0.1617,
+      "step": 18572
+    },
+    {
+      "epoch": 0.297168,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.08516129032258e-05,
+      "loss": 0.149,
+      "step": 18573
+    },
+    {
+      "epoch": 0.297184,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.085e-05,
+      "loss": 0.178,
+      "step": 18574
+    },
+    {
+      "epoch": 0.2972,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.084838709677419e-05,
+      "loss": 0.1887,
+      "step": 18575
+    },
+    {
+      "epoch": 0.297216,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.084677419354839e-05,
+      "loss": 0.1639,
+      "step": 18576
+    },
+    {
+      "epoch": 0.297232,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.084516129032259e-05,
+      "loss": 0.185,
+      "step": 18577
+    },
+    {
+      "epoch": 0.297248,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.084354838709679e-05,
+      "loss": 0.1405,
+      "step": 18578
+    },
+    {
+      "epoch": 0.297264,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.084193548387097e-05,
+      "loss": 0.1441,
+      "step": 18579
+    },
+    {
+      "epoch": 0.29728,
+      "grad_norm": 0.75,
+      "learning_rate": 7.084032258064517e-05,
+      "loss": 0.1412,
+      "step": 18580
+    },
+    {
+      "epoch": 0.297296,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.083870967741936e-05,
+      "loss": 0.1127,
+      "step": 18581
+    },
+    {
+      "epoch": 0.297312,
+      "grad_norm": 0.46875,
+      "learning_rate": 7.083709677419356e-05,
+      "loss": 0.151,
+      "step": 18582
+    },
+    {
+      "epoch": 0.297328,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.083548387096774e-05,
+      "loss": 0.1202,
+      "step": 18583
+    },
+    {
+      "epoch": 0.297344,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.083387096774194e-05,
+      "loss": 0.1482,
+      "step": 18584
+    },
+    {
+      "epoch": 0.29736,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.083225806451613e-05,
+      "loss": 0.1164,
+      "step": 18585
+    },
+    {
+      "epoch": 0.297376,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.083064516129032e-05,
+      "loss": 0.169,
+      "step": 18586
+    },
+    {
+      "epoch": 0.297392,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.082903225806451e-05,
+      "loss": 0.1362,
+      "step": 18587
+    },
+    {
+      "epoch": 0.297408,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.082741935483871e-05,
+      "loss": 0.2746,
+      "step": 18588
+    },
+    {
+      "epoch": 0.297424,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.082580645161291e-05,
+      "loss": 0.1467,
+      "step": 18589
+    },
+    {
+      "epoch": 0.29744,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.08241935483871e-05,
+      "loss": 0.1555,
+      "step": 18590
+    },
+    {
+      "epoch": 0.297456,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.08225806451613e-05,
+      "loss": 0.1896,
+      "step": 18591
+    },
+    {
+      "epoch": 0.297472,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.082096774193549e-05,
+      "loss": 0.142,
+      "step": 18592
+    },
+    {
+      "epoch": 0.297488,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.081935483870969e-05,
+      "loss": 0.1564,
+      "step": 18593
+    },
+    {
+      "epoch": 0.297504,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.081774193548387e-05,
+      "loss": 0.108,
+      "step": 18594
+    },
+    {
+      "epoch": 0.29752,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.081612903225807e-05,
+      "loss": 0.1499,
+      "step": 18595
+    },
+    {
+      "epoch": 0.297536,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.081451612903226e-05,
+      "loss": 0.1547,
+      "step": 18596
+    },
+    {
+      "epoch": 0.297552,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.081290322580646e-05,
+      "loss": 0.1444,
+      "step": 18597
+    },
+    {
+      "epoch": 0.297568,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.081129032258064e-05,
+      "loss": 0.1637,
+      "step": 18598
+    },
+    {
+      "epoch": 0.297584,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.080967741935484e-05,
+      "loss": 0.1437,
+      "step": 18599
+    },
+    {
+      "epoch": 0.2976,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.080806451612904e-05,
+      "loss": 0.1486,
+      "step": 18600
+    },
+    {
+      "epoch": 0.297616,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.080645161290323e-05,
+      "loss": 0.1921,
+      "step": 18601
+    },
+    {
+      "epoch": 0.297632,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.080483870967743e-05,
+      "loss": 0.1703,
+      "step": 18602
+    },
+    {
+      "epoch": 0.297648,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.080322580645161e-05,
+      "loss": 0.1597,
+      "step": 18603
+    },
+    {
+      "epoch": 0.297664,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.080161290322581e-05,
+      "loss": 0.1598,
+      "step": 18604
+    },
+    {
+      "epoch": 0.29768,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.08e-05,
+      "loss": 0.1709,
+      "step": 18605
+    },
+    {
+      "epoch": 0.297696,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.07983870967742e-05,
+      "loss": 0.1728,
+      "step": 18606
+    },
+    {
+      "epoch": 0.297712,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.079677419354839e-05,
+      "loss": 0.1823,
+      "step": 18607
+    },
+    {
+      "epoch": 0.297728,
+      "grad_norm": 1.4453125,
+      "learning_rate": 7.079516129032259e-05,
+      "loss": 0.2272,
+      "step": 18608
+    },
+    {
+      "epoch": 0.297744,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.079354838709677e-05,
+      "loss": 0.1815,
+      "step": 18609
+    },
+    {
+      "epoch": 0.29776,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.079193548387097e-05,
+      "loss": 0.1877,
+      "step": 18610
+    },
+    {
+      "epoch": 0.297776,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.079032258064516e-05,
+      "loss": 0.1445,
+      "step": 18611
+    },
+    {
+      "epoch": 0.297792,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.078870967741936e-05,
+      "loss": 0.1927,
+      "step": 18612
+    },
+    {
+      "epoch": 0.297808,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.078709677419356e-05,
+      "loss": 0.1834,
+      "step": 18613
+    },
+    {
+      "epoch": 0.297824,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.078548387096776e-05,
+      "loss": 0.1659,
+      "step": 18614
+    },
+    {
+      "epoch": 0.29784,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.078387096774194e-05,
+      "loss": 0.1194,
+      "step": 18615
+    },
+    {
+      "epoch": 0.297856,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.078225806451613e-05,
+      "loss": 0.1516,
+      "step": 18616
+    },
+    {
+      "epoch": 0.297872,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.078064516129033e-05,
+      "loss": 0.1402,
+      "step": 18617
+    },
+    {
+      "epoch": 0.297888,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.077903225806451e-05,
+      "loss": 0.138,
+      "step": 18618
+    },
+    {
+      "epoch": 0.297904,
+      "grad_norm": 0.48828125,
+      "learning_rate": 7.077741935483871e-05,
+      "loss": 0.146,
+      "step": 18619
+    },
+    {
+      "epoch": 0.29792,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.07758064516129e-05,
+      "loss": 0.154,
+      "step": 18620
+    },
+    {
+      "epoch": 0.297936,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.07741935483871e-05,
+      "loss": 0.1813,
+      "step": 18621
+    },
+    {
+      "epoch": 0.297952,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.077258064516129e-05,
+      "loss": 0.1655,
+      "step": 18622
+    },
+    {
+      "epoch": 0.297968,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.077096774193548e-05,
+      "loss": 0.1665,
+      "step": 18623
+    },
+    {
+      "epoch": 0.297984,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.076935483870968e-05,
+      "loss": 0.1727,
+      "step": 18624
+    },
+    {
+      "epoch": 0.298,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.076774193548388e-05,
+      "loss": 0.1373,
+      "step": 18625
+    },
+    {
+      "epoch": 0.298016,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.076612903225807e-05,
+      "loss": 0.1584,
+      "step": 18626
+    },
+    {
+      "epoch": 0.298032,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.076451612903227e-05,
+      "loss": 0.1662,
+      "step": 18627
+    },
+    {
+      "epoch": 0.298048,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.076290322580646e-05,
+      "loss": 0.1454,
+      "step": 18628
+    },
+    {
+      "epoch": 0.298064,
+      "grad_norm": 1.2890625,
+      "learning_rate": 7.076129032258066e-05,
+      "loss": 0.1924,
+      "step": 18629
+    },
+    {
+      "epoch": 0.29808,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.075967741935484e-05,
+      "loss": 0.1707,
+      "step": 18630
+    },
+    {
+      "epoch": 0.298096,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.075806451612904e-05,
+      "loss": 0.1871,
+      "step": 18631
+    },
+    {
+      "epoch": 0.298112,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.075645161290323e-05,
+      "loss": 0.1501,
+      "step": 18632
+    },
+    {
+      "epoch": 0.298128,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.075483870967741e-05,
+      "loss": 0.1843,
+      "step": 18633
+    },
+    {
+      "epoch": 0.298144,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.075322580645161e-05,
+      "loss": 0.1948,
+      "step": 18634
+    },
+    {
+      "epoch": 0.29816,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.075161290322581e-05,
+      "loss": 0.1477,
+      "step": 18635
+    },
+    {
+      "epoch": 0.298176,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.075e-05,
+      "loss": 0.1774,
+      "step": 18636
+    },
+    {
+      "epoch": 0.298192,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.07483870967742e-05,
+      "loss": 0.1799,
+      "step": 18637
+    },
+    {
+      "epoch": 0.298208,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.07467741935484e-05,
+      "loss": 0.1886,
+      "step": 18638
+    },
+    {
+      "epoch": 0.298224,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.074516129032258e-05,
+      "loss": 0.1535,
+      "step": 18639
+    },
+    {
+      "epoch": 0.29824,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.074354838709678e-05,
+      "loss": 0.2012,
+      "step": 18640
+    },
+    {
+      "epoch": 0.298256,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.074193548387097e-05,
+      "loss": 0.1893,
+      "step": 18641
+    },
+    {
+      "epoch": 0.298272,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.074032258064517e-05,
+      "loss": 0.2576,
+      "step": 18642
+    },
+    {
+      "epoch": 0.298288,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.073870967741936e-05,
+      "loss": 0.1961,
+      "step": 18643
+    },
+    {
+      "epoch": 0.298304,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.073709677419355e-05,
+      "loss": 0.2018,
+      "step": 18644
+    },
+    {
+      "epoch": 0.29832,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.073548387096774e-05,
+      "loss": 0.1924,
+      "step": 18645
+    },
+    {
+      "epoch": 0.298336,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.073387096774194e-05,
+      "loss": 0.1306,
+      "step": 18646
+    },
+    {
+      "epoch": 0.298352,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.073225806451613e-05,
+      "loss": 0.1427,
+      "step": 18647
+    },
+    {
+      "epoch": 0.298368,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.073064516129033e-05,
+      "loss": 0.1474,
+      "step": 18648
+    },
+    {
+      "epoch": 0.298384,
+      "grad_norm": 0.75,
+      "learning_rate": 7.072903225806453e-05,
+      "loss": 0.1677,
+      "step": 18649
+    },
+    {
+      "epoch": 0.2984,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.072741935483871e-05,
+      "loss": 0.1731,
+      "step": 18650
+    },
+    {
+      "epoch": 0.298416,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.072580645161291e-05,
+      "loss": 0.1823,
+      "step": 18651
+    },
+    {
+      "epoch": 0.298432,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.07241935483871e-05,
+      "loss": 0.1697,
+      "step": 18652
+    },
+    {
+      "epoch": 0.298448,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.07225806451613e-05,
+      "loss": 0.1687,
+      "step": 18653
+    },
+    {
+      "epoch": 0.298464,
+      "grad_norm": 1.375,
+      "learning_rate": 7.072096774193548e-05,
+      "loss": 0.2139,
+      "step": 18654
+    },
+    {
+      "epoch": 0.29848,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.071935483870968e-05,
+      "loss": 0.1507,
+      "step": 18655
+    },
+    {
+      "epoch": 0.298496,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.071774193548387e-05,
+      "loss": 0.1806,
+      "step": 18656
+    },
+    {
+      "epoch": 0.298512,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.071612903225807e-05,
+      "loss": 0.1882,
+      "step": 18657
+    },
+    {
+      "epoch": 0.298528,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.071451612903225e-05,
+      "loss": 0.1387,
+      "step": 18658
+    },
+    {
+      "epoch": 0.298544,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.071290322580645e-05,
+      "loss": 0.164,
+      "step": 18659
+    },
+    {
+      "epoch": 0.29856,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.071129032258065e-05,
+      "loss": 0.1437,
+      "step": 18660
+    },
+    {
+      "epoch": 0.298576,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.070967741935485e-05,
+      "loss": 0.1588,
+      "step": 18661
+    },
+    {
+      "epoch": 0.298592,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.070806451612904e-05,
+      "loss": 0.1875,
+      "step": 18662
+    },
+    {
+      "epoch": 0.298608,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.070645161290323e-05,
+      "loss": 0.1846,
+      "step": 18663
+    },
+    {
+      "epoch": 0.298624,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.070483870967743e-05,
+      "loss": 0.1662,
+      "step": 18664
+    },
+    {
+      "epoch": 0.29864,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.070322580645161e-05,
+      "loss": 0.1507,
+      "step": 18665
+    },
+    {
+      "epoch": 0.298656,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.070161290322581e-05,
+      "loss": 0.1723,
+      "step": 18666
+    },
+    {
+      "epoch": 0.298672,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.07e-05,
+      "loss": 0.1408,
+      "step": 18667
+    },
+    {
+      "epoch": 0.298688,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.06983870967742e-05,
+      "loss": 0.1482,
+      "step": 18668
+    },
+    {
+      "epoch": 0.298704,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.069677419354838e-05,
+      "loss": 0.1707,
+      "step": 18669
+    },
+    {
+      "epoch": 0.29872,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.069516129032258e-05,
+      "loss": 0.1794,
+      "step": 18670
+    },
+    {
+      "epoch": 0.298736,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.069354838709677e-05,
+      "loss": 0.1556,
+      "step": 18671
+    },
+    {
+      "epoch": 0.298752,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.069193548387097e-05,
+      "loss": 0.1259,
+      "step": 18672
+    },
+    {
+      "epoch": 0.298768,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.069032258064517e-05,
+      "loss": 0.1789,
+      "step": 18673
+    },
+    {
+      "epoch": 0.298784,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.068870967741937e-05,
+      "loss": 0.1319,
+      "step": 18674
+    },
+    {
+      "epoch": 0.2988,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.068709677419355e-05,
+      "loss": 0.1694,
+      "step": 18675
+    },
+    {
+      "epoch": 0.298816,
+      "grad_norm": 1.25,
+      "learning_rate": 7.068548387096775e-05,
+      "loss": 0.1808,
+      "step": 18676
+    },
+    {
+      "epoch": 0.298832,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.068387096774194e-05,
+      "loss": 0.1758,
+      "step": 18677
+    },
+    {
+      "epoch": 0.298848,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.068225806451613e-05,
+      "loss": 0.1344,
+      "step": 18678
+    },
+    {
+      "epoch": 0.298864,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.068064516129033e-05,
+      "loss": 0.1515,
+      "step": 18679
+    },
+    {
+      "epoch": 0.29888,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.067903225806451e-05,
+      "loss": 0.1903,
+      "step": 18680
+    },
+    {
+      "epoch": 0.298896,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.067741935483871e-05,
+      "loss": 0.164,
+      "step": 18681
+    },
+    {
+      "epoch": 0.298912,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.06758064516129e-05,
+      "loss": 0.1525,
+      "step": 18682
+    },
+    {
+      "epoch": 0.298928,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.06741935483871e-05,
+      "loss": 0.1545,
+      "step": 18683
+    },
+    {
+      "epoch": 0.298944,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.06725806451613e-05,
+      "loss": 0.1549,
+      "step": 18684
+    },
+    {
+      "epoch": 0.29896,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.06709677419355e-05,
+      "loss": 0.1337,
+      "step": 18685
+    },
+    {
+      "epoch": 0.298976,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.066935483870968e-05,
+      "loss": 0.1647,
+      "step": 18686
+    },
+    {
+      "epoch": 0.298992,
+      "grad_norm": 1.578125,
+      "learning_rate": 7.066774193548388e-05,
+      "loss": 0.2221,
+      "step": 18687
+    },
+    {
+      "epoch": 0.299008,
+      "grad_norm": 1.3046875,
+      "learning_rate": 7.066612903225807e-05,
+      "loss": 0.1784,
+      "step": 18688
+    },
+    {
+      "epoch": 0.299024,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.066451612903227e-05,
+      "loss": 0.1942,
+      "step": 18689
+    },
+    {
+      "epoch": 0.29904,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.066290322580645e-05,
+      "loss": 0.152,
+      "step": 18690
+    },
+    {
+      "epoch": 0.299056,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.066129032258065e-05,
+      "loss": 0.1711,
+      "step": 18691
+    },
+    {
+      "epoch": 0.299072,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.065967741935484e-05,
+      "loss": 0.2082,
+      "step": 18692
+    },
+    {
+      "epoch": 0.299088,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.065806451612904e-05,
+      "loss": 0.1805,
+      "step": 18693
+    },
+    {
+      "epoch": 0.299104,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.065645161290322e-05,
+      "loss": 0.1668,
+      "step": 18694
+    },
+    {
+      "epoch": 0.29912,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.065483870967742e-05,
+      "loss": 0.1527,
+      "step": 18695
+    },
+    {
+      "epoch": 0.299136,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.065322580645162e-05,
+      "loss": 0.1698,
+      "step": 18696
+    },
+    {
+      "epoch": 0.299152,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.065161290322581e-05,
+      "loss": 0.1247,
+      "step": 18697
+    },
+    {
+      "epoch": 0.299168,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.065000000000001e-05,
+      "loss": 0.1807,
+      "step": 18698
+    },
+    {
+      "epoch": 0.299184,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.06483870967742e-05,
+      "loss": 0.186,
+      "step": 18699
+    },
+    {
+      "epoch": 0.2992,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.06467741935484e-05,
+      "loss": 0.1858,
+      "step": 18700
+    },
+    {
+      "epoch": 0.299216,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.064516129032258e-05,
+      "loss": 0.1575,
+      "step": 18701
+    },
+    {
+      "epoch": 0.299232,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.064354838709678e-05,
+      "loss": 0.1604,
+      "step": 18702
+    },
+    {
+      "epoch": 0.299248,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.064193548387097e-05,
+      "loss": 0.1536,
+      "step": 18703
+    },
+    {
+      "epoch": 0.299264,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.064032258064517e-05,
+      "loss": 0.148,
+      "step": 18704
+    },
+    {
+      "epoch": 0.29928,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.063870967741935e-05,
+      "loss": 0.1654,
+      "step": 18705
+    },
+    {
+      "epoch": 0.299296,
+      "grad_norm": 0.625,
+      "learning_rate": 7.063709677419355e-05,
+      "loss": 0.1423,
+      "step": 18706
+    },
+    {
+      "epoch": 0.299312,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.063548387096774e-05,
+      "loss": 0.1725,
+      "step": 18707
+    },
+    {
+      "epoch": 0.299328,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.063387096774194e-05,
+      "loss": 0.1532,
+      "step": 18708
+    },
+    {
+      "epoch": 0.299344,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.063225806451614e-05,
+      "loss": 0.1457,
+      "step": 18709
+    },
+    {
+      "epoch": 0.29936,
+      "grad_norm": 1.328125,
+      "learning_rate": 7.063064516129032e-05,
+      "loss": 0.1595,
+      "step": 18710
+    },
+    {
+      "epoch": 0.299376,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.062903225806452e-05,
+      "loss": 0.1791,
+      "step": 18711
+    },
+    {
+      "epoch": 0.299392,
+      "grad_norm": 1.3046875,
+      "learning_rate": 7.062741935483871e-05,
+      "loss": 0.1925,
+      "step": 18712
+    },
+    {
+      "epoch": 0.299408,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.062580645161291e-05,
+      "loss": 0.1659,
+      "step": 18713
+    },
+    {
+      "epoch": 0.299424,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.06241935483871e-05,
+      "loss": 0.1733,
+      "step": 18714
+    },
+    {
+      "epoch": 0.29944,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.06225806451613e-05,
+      "loss": 0.1825,
+      "step": 18715
+    },
+    {
+      "epoch": 0.299456,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.062096774193548e-05,
+      "loss": 0.1808,
+      "step": 18716
+    },
+    {
+      "epoch": 0.299472,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.061935483870968e-05,
+      "loss": 0.2039,
+      "step": 18717
+    },
+    {
+      "epoch": 0.299488,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.061774193548387e-05,
+      "loss": 0.1602,
+      "step": 18718
+    },
+    {
+      "epoch": 0.299504,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.061612903225807e-05,
+      "loss": 0.1814,
+      "step": 18719
+    },
+    {
+      "epoch": 0.29952,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.061451612903227e-05,
+      "loss": 0.1614,
+      "step": 18720
+    },
+    {
+      "epoch": 0.299536,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.061290322580647e-05,
+      "loss": 0.1869,
+      "step": 18721
+    },
+    {
+      "epoch": 0.299552,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.061129032258065e-05,
+      "loss": 0.1653,
+      "step": 18722
+    },
+    {
+      "epoch": 0.299568,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.060967741935485e-05,
+      "loss": 0.1301,
+      "step": 18723
+    },
+    {
+      "epoch": 0.299584,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.060806451612904e-05,
+      "loss": 0.1911,
+      "step": 18724
+    },
+    {
+      "epoch": 0.2996,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.060645161290322e-05,
+      "loss": 0.1467,
+      "step": 18725
+    },
+    {
+      "epoch": 0.299616,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.060483870967742e-05,
+      "loss": 0.1199,
+      "step": 18726
+    },
+    {
+      "epoch": 0.299632,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.060322580645161e-05,
+      "loss": 0.1923,
+      "step": 18727
+    },
+    {
+      "epoch": 0.299648,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.060161290322581e-05,
+      "loss": 0.2306,
+      "step": 18728
+    },
+    {
+      "epoch": 0.299664,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.06e-05,
+      "loss": 0.1657,
+      "step": 18729
+    },
+    {
+      "epoch": 0.29968,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.05983870967742e-05,
+      "loss": 0.1646,
+      "step": 18730
+    },
+    {
+      "epoch": 0.299696,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.059677419354838e-05,
+      "loss": 0.1829,
+      "step": 18731
+    },
+    {
+      "epoch": 0.299712,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.059516129032258e-05,
+      "loss": 0.2261,
+      "step": 18732
+    },
+    {
+      "epoch": 0.299728,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.059354838709678e-05,
+      "loss": 0.2163,
+      "step": 18733
+    },
+    {
+      "epoch": 0.299744,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.059193548387098e-05,
+      "loss": 0.2086,
+      "step": 18734
+    },
+    {
+      "epoch": 0.29976,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.059032258064517e-05,
+      "loss": 0.1988,
+      "step": 18735
+    },
+    {
+      "epoch": 0.299776,
+      "grad_norm": 1.5,
+      "learning_rate": 7.058870967741937e-05,
+      "loss": 0.1735,
+      "step": 18736
+    },
+    {
+      "epoch": 0.299792,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.058709677419355e-05,
+      "loss": 0.178,
+      "step": 18737
+    },
+    {
+      "epoch": 0.299808,
+      "grad_norm": 0.875,
+      "learning_rate": 7.058548387096775e-05,
+      "loss": 0.1684,
+      "step": 18738
+    },
+    {
+      "epoch": 0.299824,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.058387096774194e-05,
+      "loss": 0.1694,
+      "step": 18739
+    },
+    {
+      "epoch": 0.29984,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.058225806451612e-05,
+      "loss": 0.1841,
+      "step": 18740
+    },
+    {
+      "epoch": 0.299856,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.058064516129032e-05,
+      "loss": 0.1947,
+      "step": 18741
+    },
+    {
+      "epoch": 0.299872,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.057903225806451e-05,
+      "loss": 0.1646,
+      "step": 18742
+    },
+    {
+      "epoch": 0.299888,
+      "grad_norm": 1.28125,
+      "learning_rate": 7.057741935483871e-05,
+      "loss": 0.1844,
+      "step": 18743
+    },
+    {
+      "epoch": 0.299904,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.057580645161291e-05,
+      "loss": 0.1826,
+      "step": 18744
+    },
+    {
+      "epoch": 0.29992,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.057419354838711e-05,
+      "loss": 0.1766,
+      "step": 18745
+    },
+    {
+      "epoch": 0.299936,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.05725806451613e-05,
+      "loss": 0.1819,
+      "step": 18746
+    },
+    {
+      "epoch": 0.299952,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.05709677419355e-05,
+      "loss": 0.1612,
+      "step": 18747
+    },
+    {
+      "epoch": 0.299968,
+      "grad_norm": 0.625,
+      "learning_rate": 7.056935483870968e-05,
+      "loss": 0.1416,
+      "step": 18748
+    },
+    {
+      "epoch": 0.299984,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.056774193548388e-05,
+      "loss": 0.1441,
+      "step": 18749
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.056612903225807e-05,
+      "loss": 0.1539,
+      "step": 18750
+    },
+    {
+      "epoch": 0.300016,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.056451612903226e-05,
+      "loss": 0.149,
+      "step": 18751
+    },
+    {
+      "epoch": 0.300032,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.056290322580645e-05,
+      "loss": 0.1498,
+      "step": 18752
+    },
+    {
+      "epoch": 0.300048,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.056129032258065e-05,
+      "loss": 0.1673,
+      "step": 18753
+    },
+    {
+      "epoch": 0.300064,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.055967741935484e-05,
+      "loss": 0.1475,
+      "step": 18754
+    },
+    {
+      "epoch": 0.30008,
+      "grad_norm": 0.482421875,
+      "learning_rate": 7.055806451612904e-05,
+      "loss": 0.1375,
+      "step": 18755
+    },
+    {
+      "epoch": 0.300096,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.055645161290324e-05,
+      "loss": 0.1961,
+      "step": 18756
+    },
+    {
+      "epoch": 0.300112,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.055483870967742e-05,
+      "loss": 0.1791,
+      "step": 18757
+    },
+    {
+      "epoch": 0.300128,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.055322580645162e-05,
+      "loss": 0.1362,
+      "step": 18758
+    },
+    {
+      "epoch": 0.300144,
+      "grad_norm": 1.3671875,
+      "learning_rate": 7.055161290322581e-05,
+      "loss": 0.2098,
+      "step": 18759
+    },
+    {
+      "epoch": 0.30016,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.055000000000001e-05,
+      "loss": 0.1735,
+      "step": 18760
+    },
+    {
+      "epoch": 0.300176,
+      "grad_norm": 1.25,
+      "learning_rate": 7.05483870967742e-05,
+      "loss": 0.1664,
+      "step": 18761
+    },
+    {
+      "epoch": 0.300192,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.054677419354839e-05,
+      "loss": 0.1735,
+      "step": 18762
+    },
+    {
+      "epoch": 0.300208,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.054516129032258e-05,
+      "loss": 0.1638,
+      "step": 18763
+    },
+    {
+      "epoch": 0.300224,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.054354838709678e-05,
+      "loss": 0.1328,
+      "step": 18764
+    },
+    {
+      "epoch": 0.30024,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.054193548387096e-05,
+      "loss": 0.1586,
+      "step": 18765
+    },
+    {
+      "epoch": 0.300256,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.054032258064516e-05,
+      "loss": 0.1984,
+      "step": 18766
+    },
+    {
+      "epoch": 0.300272,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.053870967741935e-05,
+      "loss": 0.1623,
+      "step": 18767
+    },
+    {
+      "epoch": 0.300288,
+      "grad_norm": 1.3125,
+      "learning_rate": 7.053709677419355e-05,
+      "loss": 0.184,
+      "step": 18768
+    },
+    {
+      "epoch": 0.300304,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.053548387096775e-05,
+      "loss": 0.1765,
+      "step": 18769
+    },
+    {
+      "epoch": 0.30032,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.053387096774195e-05,
+      "loss": 0.2005,
+      "step": 18770
+    },
+    {
+      "epoch": 0.300336,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.053225806451614e-05,
+      "loss": 0.1753,
+      "step": 18771
+    },
+    {
+      "epoch": 0.300352,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.053064516129032e-05,
+      "loss": 0.1495,
+      "step": 18772
+    },
+    {
+      "epoch": 0.300368,
+      "grad_norm": 1.2578125,
+      "learning_rate": 7.052903225806452e-05,
+      "loss": 0.1826,
+      "step": 18773
+    },
+    {
+      "epoch": 0.300384,
+      "grad_norm": 0.478515625,
+      "learning_rate": 7.052741935483871e-05,
+      "loss": 0.1228,
+      "step": 18774
+    },
+    {
+      "epoch": 0.3004,
+      "grad_norm": 1.3359375,
+      "learning_rate": 7.052580645161291e-05,
+      "loss": 0.2173,
+      "step": 18775
+    },
+    {
+      "epoch": 0.300416,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.052419354838709e-05,
+      "loss": 0.1684,
+      "step": 18776
+    },
+    {
+      "epoch": 0.300432,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.052258064516129e-05,
+      "loss": 0.1803,
+      "step": 18777
+    },
+    {
+      "epoch": 0.300448,
+      "grad_norm": 1.0,
+      "learning_rate": 7.052096774193548e-05,
+      "loss": 0.179,
+      "step": 18778
+    },
+    {
+      "epoch": 0.300464,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.051935483870968e-05,
+      "loss": 0.1297,
+      "step": 18779
+    },
+    {
+      "epoch": 0.30048,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.051774193548388e-05,
+      "loss": 0.1508,
+      "step": 18780
+    },
+    {
+      "epoch": 0.300496,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.051612903225808e-05,
+      "loss": 0.1522,
+      "step": 18781
+    },
+    {
+      "epoch": 0.300512,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.051451612903226e-05,
+      "loss": 0.1636,
+      "step": 18782
+    },
+    {
+      "epoch": 0.300528,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.051290322580646e-05,
+      "loss": 0.154,
+      "step": 18783
+    },
+    {
+      "epoch": 0.300544,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.051129032258065e-05,
+      "loss": 0.1628,
+      "step": 18784
+    },
+    {
+      "epoch": 0.30056,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.050967741935485e-05,
+      "loss": 0.1952,
+      "step": 18785
+    },
+    {
+      "epoch": 0.300576,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.050806451612903e-05,
+      "loss": 0.1487,
+      "step": 18786
+    },
+    {
+      "epoch": 0.300592,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.050645161290322e-05,
+      "loss": 0.2029,
+      "step": 18787
+    },
+    {
+      "epoch": 0.300608,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.050483870967742e-05,
+      "loss": 0.1676,
+      "step": 18788
+    },
+    {
+      "epoch": 0.300624,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.05032258064516e-05,
+      "loss": 0.1454,
+      "step": 18789
+    },
+    {
+      "epoch": 0.30064,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.05016129032258e-05,
+      "loss": 0.177,
+      "step": 18790
+    },
+    {
+      "epoch": 0.300656,
+      "grad_norm": 1.3203125,
+      "learning_rate": 7.05e-05,
+      "loss": 0.1711,
+      "step": 18791
+    },
+    {
+      "epoch": 0.300672,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.049838709677419e-05,
+      "loss": 0.1591,
+      "step": 18792
+    },
+    {
+      "epoch": 0.300688,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.049677419354839e-05,
+      "loss": 0.1929,
+      "step": 18793
+    },
+    {
+      "epoch": 0.300704,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.049516129032259e-05,
+      "loss": 0.1418,
+      "step": 18794
+    },
+    {
+      "epoch": 0.30072,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.049354838709678e-05,
+      "loss": 0.1606,
+      "step": 18795
+    },
+    {
+      "epoch": 0.300736,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.049193548387098e-05,
+      "loss": 0.1622,
+      "step": 18796
+    },
+    {
+      "epoch": 0.300752,
+      "grad_norm": 1.28125,
+      "learning_rate": 7.049032258064516e-05,
+      "loss": 0.2261,
+      "step": 18797
+    },
+    {
+      "epoch": 0.300768,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.048870967741936e-05,
+      "loss": 0.1538,
+      "step": 18798
+    },
+    {
+      "epoch": 0.300784,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.048709677419355e-05,
+      "loss": 0.1632,
+      "step": 18799
+    },
+    {
+      "epoch": 0.3008,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.048548387096775e-05,
+      "loss": 0.1705,
+      "step": 18800
+    },
+    {
+      "epoch": 0.300816,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.048387096774193e-05,
+      "loss": 0.1546,
+      "step": 18801
+    },
+    {
+      "epoch": 0.300832,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.048225806451613e-05,
+      "loss": 0.1741,
+      "step": 18802
+    },
+    {
+      "epoch": 0.300848,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.048064516129032e-05,
+      "loss": 0.175,
+      "step": 18803
+    },
+    {
+      "epoch": 0.300864,
+      "grad_norm": 1.28125,
+      "learning_rate": 7.047903225806452e-05,
+      "loss": 0.1714,
+      "step": 18804
+    },
+    {
+      "epoch": 0.30088,
+      "grad_norm": 1.328125,
+      "learning_rate": 7.047741935483872e-05,
+      "loss": 0.1616,
+      "step": 18805
+    },
+    {
+      "epoch": 0.300896,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.04758064516129e-05,
+      "loss": 0.211,
+      "step": 18806
+    },
+    {
+      "epoch": 0.300912,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.04741935483871e-05,
+      "loss": 0.1581,
+      "step": 18807
+    },
+    {
+      "epoch": 0.300928,
+      "grad_norm": 1.4609375,
+      "learning_rate": 7.047258064516129e-05,
+      "loss": 0.2218,
+      "step": 18808
+    },
+    {
+      "epoch": 0.300944,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.047096774193549e-05,
+      "loss": 0.1579,
+      "step": 18809
+    },
+    {
+      "epoch": 0.30096,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.046935483870968e-05,
+      "loss": 0.1214,
+      "step": 18810
+    },
+    {
+      "epoch": 0.300976,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.046774193548388e-05,
+      "loss": 0.1928,
+      "step": 18811
+    },
+    {
+      "epoch": 0.300992,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.046612903225806e-05,
+      "loss": 0.2104,
+      "step": 18812
+    },
+    {
+      "epoch": 0.301008,
+      "grad_norm": 1.4375,
+      "learning_rate": 7.046451612903226e-05,
+      "loss": 0.1479,
+      "step": 18813
+    },
+    {
+      "epoch": 0.301024,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.046290322580645e-05,
+      "loss": 0.1374,
+      "step": 18814
+    },
+    {
+      "epoch": 0.30104,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.046129032258065e-05,
+      "loss": 0.1484,
+      "step": 18815
+    },
+    {
+      "epoch": 0.301056,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.045967741935485e-05,
+      "loss": 0.1748,
+      "step": 18816
+    },
+    {
+      "epoch": 0.301072,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.045806451612905e-05,
+      "loss": 0.1449,
+      "step": 18817
+    },
+    {
+      "epoch": 0.301088,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.045645161290323e-05,
+      "loss": 0.1736,
+      "step": 18818
+    },
+    {
+      "epoch": 0.301104,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.045483870967742e-05,
+      "loss": 0.2187,
+      "step": 18819
+    },
+    {
+      "epoch": 0.30112,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.045322580645162e-05,
+      "loss": 0.1924,
+      "step": 18820
+    },
+    {
+      "epoch": 0.301136,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.04516129032258e-05,
+      "loss": 0.1768,
+      "step": 18821
+    },
+    {
+      "epoch": 0.301152,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.045e-05,
+      "loss": 0.1681,
+      "step": 18822
+    },
+    {
+      "epoch": 0.301168,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.044838709677419e-05,
+      "loss": 0.1811,
+      "step": 18823
+    },
+    {
+      "epoch": 0.301184,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.044677419354839e-05,
+      "loss": 0.1787,
+      "step": 18824
+    },
+    {
+      "epoch": 0.3012,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.044516129032258e-05,
+      "loss": 0.2112,
+      "step": 18825
+    },
+    {
+      "epoch": 0.301216,
+      "grad_norm": 1.0,
+      "learning_rate": 7.044354838709678e-05,
+      "loss": 0.1715,
+      "step": 18826
+    },
+    {
+      "epoch": 0.301232,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.044193548387096e-05,
+      "loss": 0.1515,
+      "step": 18827
+    },
+    {
+      "epoch": 0.301248,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.044032258064516e-05,
+      "loss": 0.1779,
+      "step": 18828
+    },
+    {
+      "epoch": 0.301264,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.043870967741936e-05,
+      "loss": 0.1634,
+      "step": 18829
+    },
+    {
+      "epoch": 0.30128,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.043709677419356e-05,
+      "loss": 0.1741,
+      "step": 18830
+    },
+    {
+      "epoch": 0.301296,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.043548387096775e-05,
+      "loss": 0.1878,
+      "step": 18831
+    },
+    {
+      "epoch": 0.301312,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.043387096774195e-05,
+      "loss": 0.1581,
+      "step": 18832
+    },
+    {
+      "epoch": 0.301328,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.043225806451613e-05,
+      "loss": 0.1964,
+      "step": 18833
+    },
+    {
+      "epoch": 0.301344,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.043064516129032e-05,
+      "loss": 0.1613,
+      "step": 18834
+    },
+    {
+      "epoch": 0.30136,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.042903225806452e-05,
+      "loss": 0.1794,
+      "step": 18835
+    },
+    {
+      "epoch": 0.301376,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.04274193548387e-05,
+      "loss": 0.1677,
+      "step": 18836
+    },
+    {
+      "epoch": 0.301392,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.04258064516129e-05,
+      "loss": 0.1411,
+      "step": 18837
+    },
+    {
+      "epoch": 0.301408,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.042419354838709e-05,
+      "loss": 0.1668,
+      "step": 18838
+    },
+    {
+      "epoch": 0.301424,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.042258064516129e-05,
+      "loss": 0.1829,
+      "step": 18839
+    },
+    {
+      "epoch": 0.30144,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.042096774193549e-05,
+      "loss": 0.1989,
+      "step": 18840
+    },
+    {
+      "epoch": 0.301456,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.041935483870969e-05,
+      "loss": 0.1818,
+      "step": 18841
+    },
+    {
+      "epoch": 0.301472,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.041774193548388e-05,
+      "loss": 0.1754,
+      "step": 18842
+    },
+    {
+      "epoch": 0.301488,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.041612903225808e-05,
+      "loss": 0.1635,
+      "step": 18843
+    },
+    {
+      "epoch": 0.301504,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.041451612903226e-05,
+      "loss": 0.1656,
+      "step": 18844
+    },
+    {
+      "epoch": 0.30152,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.041290322580646e-05,
+      "loss": 0.1331,
+      "step": 18845
+    },
+    {
+      "epoch": 0.301536,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.041129032258065e-05,
+      "loss": 0.179,
+      "step": 18846
+    },
+    {
+      "epoch": 0.301552,
+      "grad_norm": 1.296875,
+      "learning_rate": 7.040967741935485e-05,
+      "loss": 0.1826,
+      "step": 18847
+    },
+    {
+      "epoch": 0.301568,
+      "grad_norm": 1.578125,
+      "learning_rate": 7.040806451612903e-05,
+      "loss": 0.1821,
+      "step": 18848
+    },
+    {
+      "epoch": 0.301584,
+      "grad_norm": 0.46875,
+      "learning_rate": 7.040645161290322e-05,
+      "loss": 0.1597,
+      "step": 18849
+    },
+    {
+      "epoch": 0.3016,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.040483870967742e-05,
+      "loss": 0.1677,
+      "step": 18850
+    },
+    {
+      "epoch": 0.301616,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.040322580645162e-05,
+      "loss": 0.1948,
+      "step": 18851
+    },
+    {
+      "epoch": 0.301632,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.040161290322582e-05,
+      "loss": 0.1744,
+      "step": 18852
+    },
+    {
+      "epoch": 0.301648,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.04e-05,
+      "loss": 0.1703,
+      "step": 18853
+    },
+    {
+      "epoch": 0.301664,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.03983870967742e-05,
+      "loss": 0.1525,
+      "step": 18854
+    },
+    {
+      "epoch": 0.30168,
+      "grad_norm": 1.125,
+      "learning_rate": 7.039677419354839e-05,
+      "loss": 0.1838,
+      "step": 18855
+    },
+    {
+      "epoch": 0.301696,
+      "grad_norm": 1.296875,
+      "learning_rate": 7.039516129032259e-05,
+      "loss": 0.1896,
+      "step": 18856
+    },
+    {
+      "epoch": 0.301712,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.039354838709678e-05,
+      "loss": 0.1984,
+      "step": 18857
+    },
+    {
+      "epoch": 0.301728,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.039193548387097e-05,
+      "loss": 0.1589,
+      "step": 18858
+    },
+    {
+      "epoch": 0.301744,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.039032258064516e-05,
+      "loss": 0.1628,
+      "step": 18859
+    },
+    {
+      "epoch": 0.30176,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.038870967741936e-05,
+      "loss": 0.1742,
+      "step": 18860
+    },
+    {
+      "epoch": 0.301776,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.038709677419355e-05,
+      "loss": 0.1627,
+      "step": 18861
+    },
+    {
+      "epoch": 0.301792,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.038548387096775e-05,
+      "loss": 0.1648,
+      "step": 18862
+    },
+    {
+      "epoch": 0.301808,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.038387096774193e-05,
+      "loss": 0.1824,
+      "step": 18863
+    },
+    {
+      "epoch": 0.301824,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.038225806451613e-05,
+      "loss": 0.1746,
+      "step": 18864
+    },
+    {
+      "epoch": 0.30184,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.038064516129033e-05,
+      "loss": 0.1869,
+      "step": 18865
+    },
+    {
+      "epoch": 0.301856,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.037903225806452e-05,
+      "loss": 0.1658,
+      "step": 18866
+    },
+    {
+      "epoch": 0.301872,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.037741935483872e-05,
+      "loss": 0.1715,
+      "step": 18867
+    },
+    {
+      "epoch": 0.301888,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.03758064516129e-05,
+      "loss": 0.1678,
+      "step": 18868
+    },
+    {
+      "epoch": 0.301904,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.03741935483871e-05,
+      "loss": 0.1675,
+      "step": 18869
+    },
+    {
+      "epoch": 0.30192,
+      "grad_norm": 0.5078125,
+      "learning_rate": 7.037258064516129e-05,
+      "loss": 0.1242,
+      "step": 18870
+    },
+    {
+      "epoch": 0.301936,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.037096774193549e-05,
+      "loss": 0.1548,
+      "step": 18871
+    },
+    {
+      "epoch": 0.301952,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.036935483870967e-05,
+      "loss": 0.191,
+      "step": 18872
+    },
+    {
+      "epoch": 0.301968,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.036774193548387e-05,
+      "loss": 0.1823,
+      "step": 18873
+    },
+    {
+      "epoch": 0.301984,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.036612903225806e-05,
+      "loss": 0.1429,
+      "step": 18874
+    },
+    {
+      "epoch": 0.302,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.036451612903226e-05,
+      "loss": 0.1924,
+      "step": 18875
+    },
+    {
+      "epoch": 0.302016,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.036290322580646e-05,
+      "loss": 0.1557,
+      "step": 18876
+    },
+    {
+      "epoch": 0.302032,
+      "grad_norm": 0.494140625,
+      "learning_rate": 7.036129032258066e-05,
+      "loss": 0.1687,
+      "step": 18877
+    },
+    {
+      "epoch": 0.302048,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.035967741935485e-05,
+      "loss": 0.1525,
+      "step": 18878
+    },
+    {
+      "epoch": 0.302064,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.035806451612904e-05,
+      "loss": 0.1553,
+      "step": 18879
+    },
+    {
+      "epoch": 0.30208,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.035645161290323e-05,
+      "loss": 0.2016,
+      "step": 18880
+    },
+    {
+      "epoch": 0.302096,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.035483870967742e-05,
+      "loss": 0.1388,
+      "step": 18881
+    },
+    {
+      "epoch": 0.302112,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.035322580645162e-05,
+      "loss": 0.1862,
+      "step": 18882
+    },
+    {
+      "epoch": 0.302128,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.03516129032258e-05,
+      "loss": 0.149,
+      "step": 18883
+    },
+    {
+      "epoch": 0.302144,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.035e-05,
+      "loss": 0.168,
+      "step": 18884
+    },
+    {
+      "epoch": 0.30216,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.034838709677419e-05,
+      "loss": 0.2043,
+      "step": 18885
+    },
+    {
+      "epoch": 0.302176,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.034677419354839e-05,
+      "loss": 0.144,
+      "step": 18886
+    },
+    {
+      "epoch": 0.302192,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.034516129032259e-05,
+      "loss": 0.1578,
+      "step": 18887
+    },
+    {
+      "epoch": 0.302208,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.034354838709677e-05,
+      "loss": 0.1674,
+      "step": 18888
+    },
+    {
+      "epoch": 0.302224,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.034193548387097e-05,
+      "loss": 0.1664,
+      "step": 18889
+    },
+    {
+      "epoch": 0.30224,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.034032258064517e-05,
+      "loss": 0.1389,
+      "step": 18890
+    },
+    {
+      "epoch": 0.302256,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.033870967741936e-05,
+      "loss": 0.1517,
+      "step": 18891
+    },
+    {
+      "epoch": 0.302272,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.033709677419356e-05,
+      "loss": 0.1553,
+      "step": 18892
+    },
+    {
+      "epoch": 0.302288,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.033548387096774e-05,
+      "loss": 0.1505,
+      "step": 18893
+    },
+    {
+      "epoch": 0.302304,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.033387096774194e-05,
+      "loss": 0.2317,
+      "step": 18894
+    },
+    {
+      "epoch": 0.30232,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.033225806451613e-05,
+      "loss": 0.1959,
+      "step": 18895
+    },
+    {
+      "epoch": 0.302336,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.033064516129032e-05,
+      "loss": 0.1627,
+      "step": 18896
+    },
+    {
+      "epoch": 0.302352,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.032903225806452e-05,
+      "loss": 0.1673,
+      "step": 18897
+    },
+    {
+      "epoch": 0.302368,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.03274193548387e-05,
+      "loss": 0.1698,
+      "step": 18898
+    },
+    {
+      "epoch": 0.302384,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.03258064516129e-05,
+      "loss": 0.1467,
+      "step": 18899
+    },
+    {
+      "epoch": 0.3024,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.03241935483871e-05,
+      "loss": 0.2017,
+      "step": 18900
+    },
+    {
+      "epoch": 0.302416,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.03225806451613e-05,
+      "loss": 0.1766,
+      "step": 18901
+    },
+    {
+      "epoch": 0.302432,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.032096774193549e-05,
+      "loss": 0.1684,
+      "step": 18902
+    },
+    {
+      "epoch": 0.302448,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.031935483870969e-05,
+      "loss": 0.1932,
+      "step": 18903
+    },
+    {
+      "epoch": 0.302464,
+      "grad_norm": 1.0,
+      "learning_rate": 7.031774193548387e-05,
+      "loss": 0.1651,
+      "step": 18904
+    },
+    {
+      "epoch": 0.30248,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.031612903225807e-05,
+      "loss": 0.1744,
+      "step": 18905
+    },
+    {
+      "epoch": 0.302496,
+      "grad_norm": 0.75,
+      "learning_rate": 7.031451612903226e-05,
+      "loss": 0.1638,
+      "step": 18906
+    },
+    {
+      "epoch": 0.302512,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.031290322580646e-05,
+      "loss": 0.1679,
+      "step": 18907
+    },
+    {
+      "epoch": 0.302528,
+      "grad_norm": 0.625,
+      "learning_rate": 7.031129032258064e-05,
+      "loss": 0.1765,
+      "step": 18908
+    },
+    {
+      "epoch": 0.302544,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.030967741935484e-05,
+      "loss": 0.2087,
+      "step": 18909
+    },
+    {
+      "epoch": 0.30256,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.030806451612903e-05,
+      "loss": 0.1856,
+      "step": 18910
+    },
+    {
+      "epoch": 0.302576,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.030645161290323e-05,
+      "loss": 0.1409,
+      "step": 18911
+    },
+    {
+      "epoch": 0.302592,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.030483870967743e-05,
+      "loss": 0.1793,
+      "step": 18912
+    },
+    {
+      "epoch": 0.302608,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.030322580645162e-05,
+      "loss": 0.175,
+      "step": 18913
+    },
+    {
+      "epoch": 0.302624,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.030161290322582e-05,
+      "loss": 0.1733,
+      "step": 18914
+    },
+    {
+      "epoch": 0.30264,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.03e-05,
+      "loss": 0.182,
+      "step": 18915
+    },
+    {
+      "epoch": 0.302656,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.02983870967742e-05,
+      "loss": 0.1945,
+      "step": 18916
+    },
+    {
+      "epoch": 0.302672,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.029677419354839e-05,
+      "loss": 0.1829,
+      "step": 18917
+    },
+    {
+      "epoch": 0.302688,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.029516129032259e-05,
+      "loss": 0.1677,
+      "step": 18918
+    },
+    {
+      "epoch": 0.302704,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.029354838709677e-05,
+      "loss": 0.1713,
+      "step": 18919
+    },
+    {
+      "epoch": 0.30272,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.029193548387097e-05,
+      "loss": 0.1729,
+      "step": 18920
+    },
+    {
+      "epoch": 0.302736,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.029032258064516e-05,
+      "loss": 0.2006,
+      "step": 18921
+    },
+    {
+      "epoch": 0.302752,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.028870967741936e-05,
+      "loss": 0.1474,
+      "step": 18922
+    },
+    {
+      "epoch": 0.302768,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.028709677419354e-05,
+      "loss": 0.1795,
+      "step": 18923
+    },
+    {
+      "epoch": 0.302784,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.028548387096774e-05,
+      "loss": 0.1511,
+      "step": 18924
+    },
+    {
+      "epoch": 0.3028,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.028387096774194e-05,
+      "loss": 0.2078,
+      "step": 18925
+    },
+    {
+      "epoch": 0.302816,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.028225806451614e-05,
+      "loss": 0.1619,
+      "step": 18926
+    },
+    {
+      "epoch": 0.302832,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.028064516129033e-05,
+      "loss": 0.1642,
+      "step": 18927
+    },
+    {
+      "epoch": 0.302848,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.027903225806452e-05,
+      "loss": 0.1786,
+      "step": 18928
+    },
+    {
+      "epoch": 0.302864,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.027741935483871e-05,
+      "loss": 0.1589,
+      "step": 18929
+    },
+    {
+      "epoch": 0.30288,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.02758064516129e-05,
+      "loss": 0.1634,
+      "step": 18930
+    },
+    {
+      "epoch": 0.302896,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.02741935483871e-05,
+      "loss": 0.1924,
+      "step": 18931
+    },
+    {
+      "epoch": 0.302912,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.027258064516129e-05,
+      "loss": 0.1664,
+      "step": 18932
+    },
+    {
+      "epoch": 0.302928,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.027096774193549e-05,
+      "loss": 0.1966,
+      "step": 18933
+    },
+    {
+      "epoch": 0.302944,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.026935483870967e-05,
+      "loss": 0.1341,
+      "step": 18934
+    },
+    {
+      "epoch": 0.30296,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.026774193548387e-05,
+      "loss": 0.1697,
+      "step": 18935
+    },
+    {
+      "epoch": 0.302976,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.026612903225807e-05,
+      "loss": 0.1523,
+      "step": 18936
+    },
+    {
+      "epoch": 0.302992,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.026451612903227e-05,
+      "loss": 0.1786,
+      "step": 18937
+    },
+    {
+      "epoch": 0.303008,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.026290322580646e-05,
+      "loss": 0.192,
+      "step": 18938
+    },
+    {
+      "epoch": 0.303024,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.026129032258066e-05,
+      "loss": 0.1465,
+      "step": 18939
+    },
+    {
+      "epoch": 0.30304,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.025967741935484e-05,
+      "loss": 0.1554,
+      "step": 18940
+    },
+    {
+      "epoch": 0.303056,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.025806451612904e-05,
+      "loss": 0.1869,
+      "step": 18941
+    },
+    {
+      "epoch": 0.303072,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.025645161290323e-05,
+      "loss": 0.1595,
+      "step": 18942
+    },
+    {
+      "epoch": 0.303088,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.025483870967741e-05,
+      "loss": 0.1817,
+      "step": 18943
+    },
+    {
+      "epoch": 0.303104,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.025322580645161e-05,
+      "loss": 0.2022,
+      "step": 18944
+    },
+    {
+      "epoch": 0.30312,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.02516129032258e-05,
+      "loss": 0.1746,
+      "step": 18945
+    },
+    {
+      "epoch": 0.303136,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.025e-05,
+      "loss": 0.1333,
+      "step": 18946
+    },
+    {
+      "epoch": 0.303152,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.02483870967742e-05,
+      "loss": 0.1568,
+      "step": 18947
+    },
+    {
+      "epoch": 0.303168,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.02467741935484e-05,
+      "loss": 0.144,
+      "step": 18948
+    },
+    {
+      "epoch": 0.303184,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.024516129032259e-05,
+      "loss": 0.1853,
+      "step": 18949
+    },
+    {
+      "epoch": 0.3032,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.024354838709678e-05,
+      "loss": 0.1484,
+      "step": 18950
+    },
+    {
+      "epoch": 0.303216,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.024193548387097e-05,
+      "loss": 0.1425,
+      "step": 18951
+    },
+    {
+      "epoch": 0.303232,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.024032258064517e-05,
+      "loss": 0.164,
+      "step": 18952
+    },
+    {
+      "epoch": 0.303248,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.023870967741936e-05,
+      "loss": 0.1611,
+      "step": 18953
+    },
+    {
+      "epoch": 0.303264,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.023709677419356e-05,
+      "loss": 0.1649,
+      "step": 18954
+    },
+    {
+      "epoch": 0.30328,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.023548387096774e-05,
+      "loss": 0.1797,
+      "step": 18955
+    },
+    {
+      "epoch": 0.303296,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.023387096774194e-05,
+      "loss": 0.1683,
+      "step": 18956
+    },
+    {
+      "epoch": 0.303312,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.023225806451613e-05,
+      "loss": 0.1235,
+      "step": 18957
+    },
+    {
+      "epoch": 0.303328,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.023064516129031e-05,
+      "loss": 0.1367,
+      "step": 18958
+    },
+    {
+      "epoch": 0.303344,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.022903225806451e-05,
+      "loss": 0.1965,
+      "step": 18959
+    },
+    {
+      "epoch": 0.30336,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.022741935483871e-05,
+      "loss": 0.1731,
+      "step": 18960
+    },
+    {
+      "epoch": 0.303376,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.022580645161291e-05,
+      "loss": 0.1738,
+      "step": 18961
+    },
+    {
+      "epoch": 0.303392,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.02241935483871e-05,
+      "loss": 0.2104,
+      "step": 18962
+    },
+    {
+      "epoch": 0.303408,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.02225806451613e-05,
+      "loss": 0.193,
+      "step": 18963
+    },
+    {
+      "epoch": 0.303424,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.022096774193548e-05,
+      "loss": 0.2026,
+      "step": 18964
+    },
+    {
+      "epoch": 0.30344,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.021935483870968e-05,
+      "loss": 0.1935,
+      "step": 18965
+    },
+    {
+      "epoch": 0.303456,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.021774193548387e-05,
+      "loss": 0.1941,
+      "step": 18966
+    },
+    {
+      "epoch": 0.303472,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.021612903225807e-05,
+      "loss": 0.1647,
+      "step": 18967
+    },
+    {
+      "epoch": 0.303488,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.021451612903226e-05,
+      "loss": 0.1158,
+      "step": 18968
+    },
+    {
+      "epoch": 0.303504,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.021290322580646e-05,
+      "loss": 0.1568,
+      "step": 18969
+    },
+    {
+      "epoch": 0.30352,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.021129032258064e-05,
+      "loss": 0.1386,
+      "step": 18970
+    },
+    {
+      "epoch": 0.303536,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.020967741935484e-05,
+      "loss": 0.1847,
+      "step": 18971
+    },
+    {
+      "epoch": 0.303552,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.020806451612904e-05,
+      "loss": 0.1425,
+      "step": 18972
+    },
+    {
+      "epoch": 0.303568,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.020645161290324e-05,
+      "loss": 0.1404,
+      "step": 18973
+    },
+    {
+      "epoch": 0.303584,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.020483870967743e-05,
+      "loss": 0.1852,
+      "step": 18974
+    },
+    {
+      "epoch": 0.3036,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.020322580645161e-05,
+      "loss": 0.1633,
+      "step": 18975
+    },
+    {
+      "epoch": 0.303616,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.020161290322581e-05,
+      "loss": 0.1741,
+      "step": 18976
+    },
+    {
+      "epoch": 0.303632,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.02e-05,
+      "loss": 0.1588,
+      "step": 18977
+    },
+    {
+      "epoch": 0.303648,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.01983870967742e-05,
+      "loss": 0.1805,
+      "step": 18978
+    },
+    {
+      "epoch": 0.303664,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.019677419354838e-05,
+      "loss": 0.1696,
+      "step": 18979
+    },
+    {
+      "epoch": 0.30368,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.019516129032258e-05,
+      "loss": 0.1768,
+      "step": 18980
+    },
+    {
+      "epoch": 0.303696,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.019354838709677e-05,
+      "loss": 0.1304,
+      "step": 18981
+    },
+    {
+      "epoch": 0.303712,
+      "grad_norm": 1.2578125,
+      "learning_rate": 7.019193548387097e-05,
+      "loss": 0.1447,
+      "step": 18982
+    },
+    {
+      "epoch": 0.303728,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.019032258064516e-05,
+      "loss": 0.1552,
+      "step": 18983
+    },
+    {
+      "epoch": 0.303744,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.018870967741936e-05,
+      "loss": 0.181,
+      "step": 18984
+    },
+    {
+      "epoch": 0.30376,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.018709677419356e-05,
+      "loss": 0.1534,
+      "step": 18985
+    },
+    {
+      "epoch": 0.303776,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.018548387096775e-05,
+      "loss": 0.1581,
+      "step": 18986
+    },
+    {
+      "epoch": 0.303792,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.018387096774194e-05,
+      "loss": 0.1649,
+      "step": 18987
+    },
+    {
+      "epoch": 0.303808,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.018225806451614e-05,
+      "loss": 0.1206,
+      "step": 18988
+    },
+    {
+      "epoch": 0.303824,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.018064516129033e-05,
+      "loss": 0.1927,
+      "step": 18989
+    },
+    {
+      "epoch": 0.30384,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.017903225806451e-05,
+      "loss": 0.1571,
+      "step": 18990
+    },
+    {
+      "epoch": 0.303856,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.017741935483871e-05,
+      "loss": 0.1403,
+      "step": 18991
+    },
+    {
+      "epoch": 0.303872,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.01758064516129e-05,
+      "loss": 0.1663,
+      "step": 18992
+    },
+    {
+      "epoch": 0.303888,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.01741935483871e-05,
+      "loss": 0.2016,
+      "step": 18993
+    },
+    {
+      "epoch": 0.303904,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.017258064516128e-05,
+      "loss": 0.2085,
+      "step": 18994
+    },
+    {
+      "epoch": 0.30392,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.017096774193548e-05,
+      "loss": 0.1562,
+      "step": 18995
+    },
+    {
+      "epoch": 0.303936,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.016935483870968e-05,
+      "loss": 0.1443,
+      "step": 18996
+    },
+    {
+      "epoch": 0.303952,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.016774193548388e-05,
+      "loss": 0.1784,
+      "step": 18997
+    },
+    {
+      "epoch": 0.303968,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.016612903225807e-05,
+      "loss": 0.1711,
+      "step": 18998
+    },
+    {
+      "epoch": 0.303984,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.016451612903227e-05,
+      "loss": 0.1747,
+      "step": 18999
+    },
+    {
+      "epoch": 0.304,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.016290322580645e-05,
+      "loss": 0.1757,
+      "step": 19000
+    },
+    {
+      "epoch": 0.304016,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.016129032258065e-05,
+      "loss": 0.1716,
+      "step": 19001
+    },
+    {
+      "epoch": 0.304032,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.015967741935484e-05,
+      "loss": 0.1543,
+      "step": 19002
+    },
+    {
+      "epoch": 0.304048,
+      "grad_norm": 1.125,
+      "learning_rate": 7.015806451612904e-05,
+      "loss": 0.1727,
+      "step": 19003
+    },
+    {
+      "epoch": 0.304064,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.015645161290323e-05,
+      "loss": 0.1482,
+      "step": 19004
+    },
+    {
+      "epoch": 0.30408,
+      "grad_norm": 1.0,
+      "learning_rate": 7.015483870967741e-05,
+      "loss": 0.1886,
+      "step": 19005
+    },
+    {
+      "epoch": 0.304096,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.015322580645161e-05,
+      "loss": 0.1774,
+      "step": 19006
+    },
+    {
+      "epoch": 0.304112,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.015161290322581e-05,
+      "loss": 0.1996,
+      "step": 19007
+    },
+    {
+      "epoch": 0.304128,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.015000000000001e-05,
+      "loss": 0.1897,
+      "step": 19008
+    },
+    {
+      "epoch": 0.304144,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.01483870967742e-05,
+      "loss": 0.1718,
+      "step": 19009
+    },
+    {
+      "epoch": 0.30416,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.01467741935484e-05,
+      "loss": 0.1598,
+      "step": 19010
+    },
+    {
+      "epoch": 0.304176,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.014516129032258e-05,
+      "loss": 0.1414,
+      "step": 19011
+    },
+    {
+      "epoch": 0.304192,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.014354838709678e-05,
+      "loss": 0.1527,
+      "step": 19012
+    },
+    {
+      "epoch": 0.304208,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.014193548387097e-05,
+      "loss": 0.2048,
+      "step": 19013
+    },
+    {
+      "epoch": 0.304224,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.014032258064517e-05,
+      "loss": 0.163,
+      "step": 19014
+    },
+    {
+      "epoch": 0.30424,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.013870967741935e-05,
+      "loss": 0.1462,
+      "step": 19015
+    },
+    {
+      "epoch": 0.304256,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.013709677419355e-05,
+      "loss": 0.1614,
+      "step": 19016
+    },
+    {
+      "epoch": 0.304272,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.013548387096774e-05,
+      "loss": 0.146,
+      "step": 19017
+    },
+    {
+      "epoch": 0.304288,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.013387096774194e-05,
+      "loss": 0.1778,
+      "step": 19018
+    },
+    {
+      "epoch": 0.304304,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.013225806451613e-05,
+      "loss": 0.1644,
+      "step": 19019
+    },
+    {
+      "epoch": 0.30432,
+      "grad_norm": 1.296875,
+      "learning_rate": 7.013064516129033e-05,
+      "loss": 0.1743,
+      "step": 19020
+    },
+    {
+      "epoch": 0.304336,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.012903225806452e-05,
+      "loss": 0.1638,
+      "step": 19021
+    },
+    {
+      "epoch": 0.304352,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.012741935483871e-05,
+      "loss": 0.1689,
+      "step": 19022
+    },
+    {
+      "epoch": 0.304368,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.012580645161291e-05,
+      "loss": 0.2028,
+      "step": 19023
+    },
+    {
+      "epoch": 0.304384,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.01241935483871e-05,
+      "loss": 0.1659,
+      "step": 19024
+    },
+    {
+      "epoch": 0.3044,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.01225806451613e-05,
+      "loss": 0.1885,
+      "step": 19025
+    },
+    {
+      "epoch": 0.304416,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.012096774193548e-05,
+      "loss": 0.1436,
+      "step": 19026
+    },
+    {
+      "epoch": 0.304432,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.011935483870968e-05,
+      "loss": 0.1434,
+      "step": 19027
+    },
+    {
+      "epoch": 0.304448,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.011774193548387e-05,
+      "loss": 0.1696,
+      "step": 19028
+    },
+    {
+      "epoch": 0.304464,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.011612903225807e-05,
+      "loss": 0.1885,
+      "step": 19029
+    },
+    {
+      "epoch": 0.30448,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.011451612903225e-05,
+      "loss": 0.1817,
+      "step": 19030
+    },
+    {
+      "epoch": 0.304496,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.011290322580645e-05,
+      "loss": 0.1391,
+      "step": 19031
+    },
+    {
+      "epoch": 0.304512,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.011129032258065e-05,
+      "loss": 0.2032,
+      "step": 19032
+    },
+    {
+      "epoch": 0.304528,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.010967741935485e-05,
+      "loss": 0.137,
+      "step": 19033
+    },
+    {
+      "epoch": 0.304544,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.010806451612904e-05,
+      "loss": 0.1861,
+      "step": 19034
+    },
+    {
+      "epoch": 0.30456,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.010645161290324e-05,
+      "loss": 0.2026,
+      "step": 19035
+    },
+    {
+      "epoch": 0.304576,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.010483870967742e-05,
+      "loss": 0.1559,
+      "step": 19036
+    },
+    {
+      "epoch": 0.304592,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.010322580645161e-05,
+      "loss": 0.1815,
+      "step": 19037
+    },
+    {
+      "epoch": 0.304608,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.010161290322581e-05,
+      "loss": 0.1469,
+      "step": 19038
+    },
+    {
+      "epoch": 0.304624,
+      "grad_norm": 0.75,
+      "learning_rate": 7.01e-05,
+      "loss": 0.1466,
+      "step": 19039
+    },
+    {
+      "epoch": 0.30464,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.00983870967742e-05,
+      "loss": 0.1335,
+      "step": 19040
+    },
+    {
+      "epoch": 0.304656,
+      "grad_norm": 1.3671875,
+      "learning_rate": 7.009677419354838e-05,
+      "loss": 0.1539,
+      "step": 19041
+    },
+    {
+      "epoch": 0.304672,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.009516129032258e-05,
+      "loss": 0.1637,
+      "step": 19042
+    },
+    {
+      "epoch": 0.304688,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.009354838709678e-05,
+      "loss": 0.1668,
+      "step": 19043
+    },
+    {
+      "epoch": 0.304704,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.009193548387097e-05,
+      "loss": 0.1574,
+      "step": 19044
+    },
+    {
+      "epoch": 0.30472,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.009032258064517e-05,
+      "loss": 0.1784,
+      "step": 19045
+    },
+    {
+      "epoch": 0.304736,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.008870967741937e-05,
+      "loss": 0.1644,
+      "step": 19046
+    },
+    {
+      "epoch": 0.304752,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.008709677419355e-05,
+      "loss": 0.1682,
+      "step": 19047
+    },
+    {
+      "epoch": 0.304768,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.008548387096775e-05,
+      "loss": 0.1338,
+      "step": 19048
+    },
+    {
+      "epoch": 0.304784,
+      "grad_norm": 1.3515625,
+      "learning_rate": 7.008387096774194e-05,
+      "loss": 0.1947,
+      "step": 19049
+    },
+    {
+      "epoch": 0.3048,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.008225806451614e-05,
+      "loss": 0.1732,
+      "step": 19050
+    },
+    {
+      "epoch": 0.304816,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.008064516129032e-05,
+      "loss": 0.2107,
+      "step": 19051
+    },
+    {
+      "epoch": 0.304832,
+      "grad_norm": 1.2734375,
+      "learning_rate": 7.007903225806451e-05,
+      "loss": 0.2051,
+      "step": 19052
+    },
+    {
+      "epoch": 0.304848,
+      "grad_norm": 1.0,
+      "learning_rate": 7.007741935483871e-05,
+      "loss": 0.1927,
+      "step": 19053
+    },
+    {
+      "epoch": 0.304864,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.00758064516129e-05,
+      "loss": 0.175,
+      "step": 19054
+    },
+    {
+      "epoch": 0.30488,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.00741935483871e-05,
+      "loss": 0.1885,
+      "step": 19055
+    },
+    {
+      "epoch": 0.304896,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.00725806451613e-05,
+      "loss": 0.1668,
+      "step": 19056
+    },
+    {
+      "epoch": 0.304912,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.00709677419355e-05,
+      "loss": 0.182,
+      "step": 19057
+    },
+    {
+      "epoch": 0.304928,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.006935483870968e-05,
+      "loss": 0.1754,
+      "step": 19058
+    },
+    {
+      "epoch": 0.304944,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.006774193548388e-05,
+      "loss": 0.135,
+      "step": 19059
+    },
+    {
+      "epoch": 0.30496,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.006612903225807e-05,
+      "loss": 0.1669,
+      "step": 19060
+    },
+    {
+      "epoch": 0.304976,
+      "grad_norm": 1.453125,
+      "learning_rate": 7.006451612903227e-05,
+      "loss": 0.1778,
+      "step": 19061
+    },
+    {
+      "epoch": 0.304992,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.006290322580645e-05,
+      "loss": 0.1503,
+      "step": 19062
+    },
+    {
+      "epoch": 0.305008,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.006129032258065e-05,
+      "loss": 0.2021,
+      "step": 19063
+    },
+    {
+      "epoch": 0.305024,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.005967741935484e-05,
+      "loss": 0.1795,
+      "step": 19064
+    },
+    {
+      "epoch": 0.30504,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.005806451612904e-05,
+      "loss": 0.154,
+      "step": 19065
+    },
+    {
+      "epoch": 0.305056,
+      "grad_norm": 1.28125,
+      "learning_rate": 7.005645161290322e-05,
+      "loss": 0.1527,
+      "step": 19066
+    },
+    {
+      "epoch": 0.305072,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.005483870967742e-05,
+      "loss": 0.1882,
+      "step": 19067
+    },
+    {
+      "epoch": 0.305088,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.005322580645162e-05,
+      "loss": 0.1333,
+      "step": 19068
+    },
+    {
+      "epoch": 0.305104,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.005161290322581e-05,
+      "loss": 0.1161,
+      "step": 19069
+    },
+    {
+      "epoch": 0.30512,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.005000000000001e-05,
+      "loss": 0.2015,
+      "step": 19070
+    },
+    {
+      "epoch": 0.305136,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.00483870967742e-05,
+      "loss": 0.1523,
+      "step": 19071
+    },
+    {
+      "epoch": 0.305152,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.00467741935484e-05,
+      "loss": 0.1461,
+      "step": 19072
+    },
+    {
+      "epoch": 0.305168,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.004516129032258e-05,
+      "loss": 0.1608,
+      "step": 19073
+    },
+    {
+      "epoch": 0.305184,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.004354838709678e-05,
+      "loss": 0.1896,
+      "step": 19074
+    },
+    {
+      "epoch": 0.3052,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.004193548387097e-05,
+      "loss": 0.1608,
+      "step": 19075
+    },
+    {
+      "epoch": 0.305216,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.004032258064517e-05,
+      "loss": 0.1727,
+      "step": 19076
+    },
+    {
+      "epoch": 0.305232,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.003870967741935e-05,
+      "loss": 0.1567,
+      "step": 19077
+    },
+    {
+      "epoch": 0.305248,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.003709677419355e-05,
+      "loss": 0.1613,
+      "step": 19078
+    },
+    {
+      "epoch": 0.305264,
+      "grad_norm": 1.125,
+      "learning_rate": 7.003548387096774e-05,
+      "loss": 0.1534,
+      "step": 19079
+    },
+    {
+      "epoch": 0.30528,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.003387096774194e-05,
+      "loss": 0.177,
+      "step": 19080
+    },
+    {
+      "epoch": 0.305296,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.003225806451614e-05,
+      "loss": 0.1489,
+      "step": 19081
+    },
+    {
+      "epoch": 0.305312,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.003064516129034e-05,
+      "loss": 0.1608,
+      "step": 19082
+    },
+    {
+      "epoch": 0.305328,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.002903225806452e-05,
+      "loss": 0.1767,
+      "step": 19083
+    },
+    {
+      "epoch": 0.305344,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.002741935483871e-05,
+      "loss": 0.1938,
+      "step": 19084
+    },
+    {
+      "epoch": 0.30536,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.002580645161291e-05,
+      "loss": 0.2002,
+      "step": 19085
+    },
+    {
+      "epoch": 0.305376,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.00241935483871e-05,
+      "loss": 0.1795,
+      "step": 19086
+    },
+    {
+      "epoch": 0.305392,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.00225806451613e-05,
+      "loss": 0.1343,
+      "step": 19087
+    },
+    {
+      "epoch": 0.305408,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.002096774193548e-05,
+      "loss": 0.2192,
+      "step": 19088
+    },
+    {
+      "epoch": 0.305424,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.001935483870968e-05,
+      "loss": 0.2201,
+      "step": 19089
+    },
+    {
+      "epoch": 0.30544,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.001774193548387e-05,
+      "loss": 0.2003,
+      "step": 19090
+    },
+    {
+      "epoch": 0.305456,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.001612903225807e-05,
+      "loss": 0.1911,
+      "step": 19091
+    },
+    {
+      "epoch": 0.305472,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.001451612903226e-05,
+      "loss": 0.2131,
+      "step": 19092
+    },
+    {
+      "epoch": 0.305488,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.001290322580646e-05,
+      "loss": 0.1697,
+      "step": 19093
+    },
+    {
+      "epoch": 0.305504,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.001129032258065e-05,
+      "loss": 0.1644,
+      "step": 19094
+    },
+    {
+      "epoch": 0.30552,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.000967741935485e-05,
+      "loss": 0.1728,
+      "step": 19095
+    },
+    {
+      "epoch": 0.305536,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.000806451612904e-05,
+      "loss": 0.1464,
+      "step": 19096
+    },
+    {
+      "epoch": 0.305552,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.000645161290324e-05,
+      "loss": 0.2457,
+      "step": 19097
+    },
+    {
+      "epoch": 0.305568,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.000483870967742e-05,
+      "loss": 0.1648,
+      "step": 19098
+    },
+    {
+      "epoch": 0.305584,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.000322580645161e-05,
+      "loss": 0.1661,
+      "step": 19099
+    },
+    {
+      "epoch": 0.3056,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.000161290322581e-05,
+      "loss": 0.1694,
+      "step": 19100
+    },
+    {
+      "epoch": 0.305616,
+      "grad_norm": 1.78125,
+      "learning_rate": 7e-05,
+      "loss": 0.164,
+      "step": 19101
+    },
+    {
+      "epoch": 0.305632,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.99983870967742e-05,
+      "loss": 0.1651,
+      "step": 19102
+    },
+    {
+      "epoch": 0.305648,
+      "grad_norm": 1.2578125,
+      "learning_rate": 6.999677419354839e-05,
+      "loss": 0.2091,
+      "step": 19103
+    },
+    {
+      "epoch": 0.305664,
+      "grad_norm": 1.3671875,
+      "learning_rate": 6.999516129032259e-05,
+      "loss": 0.1324,
+      "step": 19104
+    },
+    {
+      "epoch": 0.30568,
+      "grad_norm": 0.85546875,
+      "learning_rate": 6.999354838709678e-05,
+      "loss": 0.1367,
+      "step": 19105
+    },
+    {
+      "epoch": 0.305696,
+      "grad_norm": 1.3984375,
+      "learning_rate": 6.999193548387098e-05,
+      "loss": 0.1829,
+      "step": 19106
+    },
+    {
+      "epoch": 0.305712,
+      "grad_norm": 1.125,
+      "learning_rate": 6.999032258064516e-05,
+      "loss": 0.2002,
+      "step": 19107
+    },
+    {
+      "epoch": 0.305728,
+      "grad_norm": 0.73046875,
+      "learning_rate": 6.998870967741936e-05,
+      "loss": 0.1779,
+      "step": 19108
+    },
+    {
+      "epoch": 0.305744,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.998709677419355e-05,
+      "loss": 0.1433,
+      "step": 19109
+    },
+    {
+      "epoch": 0.30576,
+      "grad_norm": 0.7890625,
+      "learning_rate": 6.998548387096775e-05,
+      "loss": 0.1597,
+      "step": 19110
+    },
+    {
+      "epoch": 0.305776,
+      "grad_norm": 0.921875,
+      "learning_rate": 6.998387096774194e-05,
+      "loss": 0.1573,
+      "step": 19111
+    },
+    {
+      "epoch": 0.305792,
+      "grad_norm": 0.6953125,
+      "learning_rate": 6.998225806451614e-05,
+      "loss": 0.1974,
+      "step": 19112
+    },
+    {
+      "epoch": 0.305808,
+      "grad_norm": 0.8046875,
+      "learning_rate": 6.998064516129032e-05,
+      "loss": 0.1643,
+      "step": 19113
+    },
+    {
+      "epoch": 0.305824,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.997903225806451e-05,
+      "loss": 0.1946,
+      "step": 19114
+    },
+    {
+      "epoch": 0.30584,
+      "grad_norm": 1.203125,
+      "learning_rate": 6.997741935483871e-05,
+      "loss": 0.1623,
+      "step": 19115
+    },
+    {
+      "epoch": 0.305856,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.997580645161291e-05,
+      "loss": 0.2144,
+      "step": 19116
+    },
+    {
+      "epoch": 0.305872,
+      "grad_norm": 0.80078125,
+      "learning_rate": 6.99741935483871e-05,
+      "loss": 0.1831,
+      "step": 19117
+    },
+    {
+      "epoch": 0.305888,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.997258064516129e-05,
+      "loss": 0.1563,
+      "step": 19118
+    },
+    {
+      "epoch": 0.305904,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.997096774193549e-05,
+      "loss": 0.1771,
+      "step": 19119
+    },
+    {
+      "epoch": 0.30592,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.996935483870968e-05,
+      "loss": 0.1586,
+      "step": 19120
+    },
+    {
+      "epoch": 0.305936,
+      "grad_norm": 1.1328125,
+      "learning_rate": 6.996774193548388e-05,
+      "loss": 0.1837,
+      "step": 19121
+    },
+    {
+      "epoch": 0.305952,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.996612903225806e-05,
+      "loss": 0.1676,
+      "step": 19122
+    },
+    {
+      "epoch": 0.305968,
+      "grad_norm": 0.86328125,
+      "learning_rate": 6.996451612903226e-05,
+      "loss": 0.1808,
+      "step": 19123
+    },
+    {
+      "epoch": 0.305984,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.996290322580645e-05,
+      "loss": 0.1696,
+      "step": 19124
+    },
+    {
+      "epoch": 0.306,
+      "grad_norm": 1.5078125,
+      "learning_rate": 6.996129032258065e-05,
+      "loss": 0.1618,
+      "step": 19125
+    },
+    {
+      "epoch": 0.306016,
+      "grad_norm": 0.7734375,
+      "learning_rate": 6.995967741935484e-05,
+      "loss": 0.1547,
+      "step": 19126
+    },
+    {
+      "epoch": 0.306032,
+      "grad_norm": 0.828125,
+      "learning_rate": 6.995806451612904e-05,
+      "loss": 0.1634,
+      "step": 19127
+    },
+    {
+      "epoch": 0.306048,
+      "grad_norm": 0.8359375,
+      "learning_rate": 6.995645161290323e-05,
+      "loss": 0.2075,
+      "step": 19128
+    },
+    {
+      "epoch": 0.306064,
+      "grad_norm": 1.3203125,
+      "learning_rate": 6.995483870967743e-05,
+      "loss": 0.2018,
+      "step": 19129
+    },
+    {
+      "epoch": 0.30608,
+      "grad_norm": 1.1953125,
+      "learning_rate": 6.995322580645162e-05,
+      "loss": 0.1764,
+      "step": 19130
+    },
+    {
+      "epoch": 0.306096,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.99516129032258e-05,
+      "loss": 0.1864,
+      "step": 19131
+    },
+    {
+      "epoch": 0.306112,
+      "grad_norm": 0.8046875,
+      "learning_rate": 6.995e-05,
+      "loss": 0.1664,
+      "step": 19132
+    },
+    {
+      "epoch": 0.306128,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.994838709677419e-05,
+      "loss": 0.18,
+      "step": 19133
+    },
+    {
+      "epoch": 0.306144,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.994677419354839e-05,
+      "loss": 0.111,
+      "step": 19134
+    },
+    {
+      "epoch": 0.30616,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.994516129032258e-05,
+      "loss": 0.1684,
+      "step": 19135
+    },
+    {
+      "epoch": 0.306176,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.994354838709678e-05,
+      "loss": 0.1561,
+      "step": 19136
+    },
+    {
+      "epoch": 0.306192,
+      "grad_norm": 1.34375,
+      "learning_rate": 6.994193548387096e-05,
+      "loss": 0.184,
+      "step": 19137
+    },
+    {
+      "epoch": 0.306208,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.994032258064516e-05,
+      "loss": 0.1539,
+      "step": 19138
+    },
+    {
+      "epoch": 0.306224,
+      "grad_norm": 0.67578125,
+      "learning_rate": 6.993870967741936e-05,
+      "loss": 0.1744,
+      "step": 19139
+    },
+    {
+      "epoch": 0.30624,
+      "grad_norm": 0.88671875,
+      "learning_rate": 6.993709677419355e-05,
+      "loss": 0.1744,
+      "step": 19140
+    },
+    {
+      "epoch": 0.306256,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.993548387096775e-05,
+      "loss": 0.1346,
+      "step": 19141
+    },
+    {
+      "epoch": 0.306272,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.993387096774195e-05,
+      "loss": 0.1676,
+      "step": 19142
+    },
+    {
+      "epoch": 0.306288,
+      "grad_norm": 0.99609375,
+      "learning_rate": 6.993225806451613e-05,
+      "loss": 0.1366,
+      "step": 19143
+    },
+    {
+      "epoch": 0.306304,
+      "grad_norm": 1.2265625,
+      "learning_rate": 6.993064516129033e-05,
+      "loss": 0.1655,
+      "step": 19144
+    },
+    {
+      "epoch": 0.30632,
+      "grad_norm": 0.74609375,
+      "learning_rate": 6.992903225806452e-05,
+      "loss": 0.1616,
+      "step": 19145
+    },
+    {
+      "epoch": 0.306336,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.99274193548387e-05,
+      "loss": 0.1607,
+      "step": 19146
+    },
+    {
+      "epoch": 0.306352,
+      "grad_norm": 1.265625,
+      "learning_rate": 6.99258064516129e-05,
+      "loss": 0.2251,
+      "step": 19147
+    },
+    {
+      "epoch": 0.306368,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.992419354838709e-05,
+      "loss": 0.1628,
+      "step": 19148
+    },
+    {
+      "epoch": 0.306384,
+      "grad_norm": 0.96484375,
+      "learning_rate": 6.992258064516129e-05,
+      "loss": 0.1153,
+      "step": 19149
+    },
+    {
+      "epoch": 0.3064,
+      "grad_norm": 1.2578125,
+      "learning_rate": 6.992096774193548e-05,
+      "loss": 0.1677,
+      "step": 19150
+    },
+    {
+      "epoch": 0.306416,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.991935483870968e-05,
+      "loss": 0.1888,
+      "step": 19151
+    },
+    {
+      "epoch": 0.306432,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.991774193548388e-05,
+      "loss": 0.1794,
+      "step": 19152
+    },
+    {
+      "epoch": 0.306448,
+      "grad_norm": 1.1640625,
+      "learning_rate": 6.991612903225808e-05,
+      "loss": 0.1961,
+      "step": 19153
+    },
+    {
+      "epoch": 0.306464,
+      "grad_norm": 1.5703125,
+      "learning_rate": 6.991451612903226e-05,
+      "loss": 0.1977,
+      "step": 19154
+    },
+    {
+      "epoch": 0.30648,
+      "grad_norm": 0.84765625,
+      "learning_rate": 6.991290322580646e-05,
+      "loss": 0.201,
+      "step": 19155
+    },
+    {
+      "epoch": 0.306496,
+      "grad_norm": 0.9375,
+      "learning_rate": 6.991129032258065e-05,
+      "loss": 0.1569,
+      "step": 19156
+    },
+    {
+      "epoch": 0.306512,
+      "grad_norm": 1.140625,
+      "learning_rate": 6.990967741935485e-05,
+      "loss": 0.1888,
+      "step": 19157
+    },
+    {
+      "epoch": 0.306528,
+      "grad_norm": 0.90234375,
+      "learning_rate": 6.990806451612903e-05,
+      "loss": 0.1623,
+      "step": 19158
+    },
+    {
+      "epoch": 0.306544,
+      "grad_norm": 0.81640625,
+      "learning_rate": 6.990645161290323e-05,
+      "loss": 0.1569,
+      "step": 19159
+    },
+    {
+      "epoch": 0.30656,
+      "grad_norm": 0.78515625,
+      "learning_rate": 6.990483870967742e-05,
+      "loss": 0.2101,
+      "step": 19160
+    },
+    {
+      "epoch": 0.306576,
+      "grad_norm": 0.7890625,
+      "learning_rate": 6.99032258064516e-05,
+      "loss": 0.1899,
+      "step": 19161
+    },
+    {
+      "epoch": 0.306592,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.99016129032258e-05,
+      "loss": 0.1817,
+      "step": 19162
+    },
+    {
+      "epoch": 0.306608,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.99e-05,
+      "loss": 0.1359,
+      "step": 19163
+    },
+    {
+      "epoch": 0.306624,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.98983870967742e-05,
+      "loss": 0.1542,
+      "step": 19164
+    },
+    {
+      "epoch": 0.30664,
+      "grad_norm": 0.84375,
+      "learning_rate": 6.989677419354839e-05,
+      "loss": 0.223,
+      "step": 19165
+    },
+    {
+      "epoch": 0.306656,
+      "grad_norm": 0.828125,
+      "learning_rate": 6.989516129032259e-05,
+      "loss": 0.1632,
+      "step": 19166
+    },
+    {
+      "epoch": 0.306672,
+      "grad_norm": 1.3515625,
+      "learning_rate": 6.989354838709678e-05,
+      "loss": 0.2003,
+      "step": 19167
+    },
+    {
+      "epoch": 0.306688,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.989193548387098e-05,
+      "loss": 0.1462,
+      "step": 19168
+    },
+    {
+      "epoch": 0.306704,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.989032258064516e-05,
+      "loss": 0.2084,
+      "step": 19169
+    },
+    {
+      "epoch": 0.30672,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.988870967741936e-05,
+      "loss": 0.1696,
+      "step": 19170
+    },
+    {
+      "epoch": 0.306736,
+      "grad_norm": 0.75390625,
+      "learning_rate": 6.988709677419355e-05,
+      "loss": 0.191,
+      "step": 19171
+    },
+    {
+      "epoch": 0.306752,
+      "grad_norm": 0.490234375,
+      "learning_rate": 6.988548387096775e-05,
+      "loss": 0.148,
+      "step": 19172
+    },
+    {
+      "epoch": 0.306768,
+      "grad_norm": 0.79296875,
+      "learning_rate": 6.988387096774193e-05,
+      "loss": 0.189,
+      "step": 19173
+    },
+    {
+      "epoch": 0.306784,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.988225806451613e-05,
+      "loss": 0.187,
+      "step": 19174
+    },
+    {
+      "epoch": 0.3068,
+      "grad_norm": 0.90625,
+      "learning_rate": 6.988064516129032e-05,
+      "loss": 0.1915,
+      "step": 19175
+    },
+    {
+      "epoch": 0.306816,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.987903225806452e-05,
+      "loss": 0.1563,
+      "step": 19176
+    },
+    {
+      "epoch": 0.306832,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.987741935483872e-05,
+      "loss": 0.174,
+      "step": 19177
+    },
+    {
+      "epoch": 0.306848,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.98758064516129e-05,
+      "loss": 0.1603,
+      "step": 19178
+    },
+    {
+      "epoch": 0.306864,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.98741935483871e-05,
+      "loss": 0.1683,
+      "step": 19179
+    },
+    {
+      "epoch": 0.30688,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.987258064516129e-05,
+      "loss": 0.1526,
+      "step": 19180
+    },
+    {
+      "epoch": 0.306896,
+      "grad_norm": 0.67578125,
+      "learning_rate": 6.987096774193549e-05,
+      "loss": 0.1737,
+      "step": 19181
+    },
+    {
+      "epoch": 0.306912,
+      "grad_norm": 1.2265625,
+      "learning_rate": 6.986935483870968e-05,
+      "loss": 0.1704,
+      "step": 19182
+    },
+    {
+      "epoch": 0.306928,
+      "grad_norm": 0.87890625,
+      "learning_rate": 6.986774193548388e-05,
+      "loss": 0.1522,
+      "step": 19183
+    },
+    {
+      "epoch": 0.306944,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.986612903225806e-05,
+      "loss": 0.1973,
+      "step": 19184
+    },
+    {
+      "epoch": 0.30696,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.986451612903226e-05,
+      "loss": 0.2092,
+      "step": 19185
+    },
+    {
+      "epoch": 0.306976,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.986290322580645e-05,
+      "loss": 0.1597,
+      "step": 19186
+    },
+    {
+      "epoch": 0.306992,
+      "grad_norm": 0.73046875,
+      "learning_rate": 6.986129032258065e-05,
+      "loss": 0.1873,
+      "step": 19187
+    },
+    {
+      "epoch": 0.307008,
+      "grad_norm": 0.859375,
+      "learning_rate": 6.985967741935485e-05,
+      "loss": 0.164,
+      "step": 19188
+    },
+    {
+      "epoch": 0.307024,
+      "grad_norm": 0.828125,
+      "learning_rate": 6.985806451612905e-05,
+      "loss": 0.1738,
+      "step": 19189
+    },
+    {
+      "epoch": 0.30704,
+      "grad_norm": 1.40625,
+      "learning_rate": 6.985645161290323e-05,
+      "loss": 0.1266,
+      "step": 19190
+    },
+    {
+      "epoch": 0.307056,
+      "grad_norm": 0.765625,
+      "learning_rate": 6.985483870967743e-05,
+      "loss": 0.1896,
+      "step": 19191
+    },
+    {
+      "epoch": 0.307072,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.985322580645162e-05,
+      "loss": 0.1802,
+      "step": 19192
+    },
+    {
+      "epoch": 0.307088,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.98516129032258e-05,
+      "loss": 0.1883,
+      "step": 19193
+    },
+    {
+      "epoch": 0.307104,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.985e-05,
+      "loss": 0.1855,
+      "step": 19194
+    },
+    {
+      "epoch": 0.30712,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.984838709677419e-05,
+      "loss": 0.1876,
+      "step": 19195
+    },
+    {
+      "epoch": 0.307136,
+      "grad_norm": 0.78125,
+      "learning_rate": 6.984677419354839e-05,
+      "loss": 0.1967,
+      "step": 19196
+    },
+    {
+      "epoch": 0.307152,
+      "grad_norm": 0.75390625,
+      "learning_rate": 6.984516129032258e-05,
+      "loss": 0.1776,
+      "step": 19197
+    },
+    {
+      "epoch": 0.307168,
+      "grad_norm": 0.8671875,
+      "learning_rate": 6.984354838709678e-05,
+      "loss": 0.2098,
+      "step": 19198
+    },
+    {
+      "epoch": 0.307184,
+      "grad_norm": 0.7734375,
+      "learning_rate": 6.984193548387097e-05,
+      "loss": 0.1613,
+      "step": 19199
+    },
+    {
+      "epoch": 0.3072,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.984032258064517e-05,
+      "loss": 0.1701,
+      "step": 19200
+    },
+    {
+      "epoch": 0.307216,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.983870967741936e-05,
+      "loss": 0.1698,
+      "step": 19201
+    },
+    {
+      "epoch": 0.307232,
+      "grad_norm": 0.9375,
+      "learning_rate": 6.983709677419356e-05,
+      "loss": 0.1858,
+      "step": 19202
+    },
+    {
+      "epoch": 0.307248,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.983548387096775e-05,
+      "loss": 0.1982,
+      "step": 19203
+    },
+    {
+      "epoch": 0.307264,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.983387096774195e-05,
+      "loss": 0.1774,
+      "step": 19204
+    },
+    {
+      "epoch": 0.30728,
+      "grad_norm": 0.796875,
+      "learning_rate": 6.983225806451613e-05,
+      "loss": 0.1626,
+      "step": 19205
+    },
+    {
+      "epoch": 0.307296,
+      "grad_norm": 0.69921875,
+      "learning_rate": 6.983064516129033e-05,
+      "loss": 0.1679,
+      "step": 19206
+    },
+    {
+      "epoch": 0.307312,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.982903225806452e-05,
+      "loss": 0.2185,
+      "step": 19207
+    },
+    {
+      "epoch": 0.307328,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.98274193548387e-05,
+      "loss": 0.1856,
+      "step": 19208
+    },
+    {
+      "epoch": 0.307344,
+      "grad_norm": 0.96484375,
+      "learning_rate": 6.98258064516129e-05,
+      "loss": 0.159,
+      "step": 19209
+    },
+    {
+      "epoch": 0.30736,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.982419354838709e-05,
+      "loss": 0.1654,
+      "step": 19210
+    },
+    {
+      "epoch": 0.307376,
+      "grad_norm": 1.2421875,
+      "learning_rate": 6.982258064516129e-05,
+      "loss": 0.1473,
+      "step": 19211
+    },
+    {
+      "epoch": 0.307392,
+      "grad_norm": 0.7421875,
+      "learning_rate": 6.982096774193549e-05,
+      "loss": 0.1653,
+      "step": 19212
+    },
+    {
+      "epoch": 0.307408,
+      "grad_norm": 0.796875,
+      "learning_rate": 6.981935483870969e-05,
+      "loss": 0.1405,
+      "step": 19213
+    },
+    {
+      "epoch": 0.307424,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.981774193548387e-05,
+      "loss": 0.1711,
+      "step": 19214
+    },
+    {
+      "epoch": 0.30744,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.981612903225807e-05,
+      "loss": 0.1666,
+      "step": 19215
+    },
+    {
+      "epoch": 0.307456,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.981451612903226e-05,
+      "loss": 0.149,
+      "step": 19216
+    },
+    {
+      "epoch": 0.307472,
+      "grad_norm": 0.84375,
+      "learning_rate": 6.981290322580646e-05,
+      "loss": 0.2056,
+      "step": 19217
+    },
+    {
+      "epoch": 0.307488,
+      "grad_norm": 0.90234375,
+      "learning_rate": 6.981129032258065e-05,
+      "loss": 0.2139,
+      "step": 19218
+    },
+    {
+      "epoch": 0.307504,
+      "grad_norm": 0.74609375,
+      "learning_rate": 6.980967741935485e-05,
+      "loss": 0.1611,
+      "step": 19219
+    },
+    {
+      "epoch": 0.30752,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.980806451612903e-05,
+      "loss": 0.1652,
+      "step": 19220
+    },
+    {
+      "epoch": 0.307536,
+      "grad_norm": 1.5,
+      "learning_rate": 6.980645161290323e-05,
+      "loss": 0.1654,
+      "step": 19221
+    },
+    {
+      "epoch": 0.307552,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.980483870967742e-05,
+      "loss": 0.177,
+      "step": 19222
+    },
+    {
+      "epoch": 0.307568,
+      "grad_norm": 0.99609375,
+      "learning_rate": 6.980322580645162e-05,
+      "loss": 0.17,
+      "step": 19223
+    },
+    {
+      "epoch": 0.307584,
+      "grad_norm": 0.46875,
+      "learning_rate": 6.980161290322582e-05,
+      "loss": 0.1605,
+      "step": 19224
+    },
+    {
+      "epoch": 0.3076,
+      "grad_norm": 0.98828125,
+      "learning_rate": 6.98e-05,
+      "loss": 0.1632,
+      "step": 19225
+    },
+    {
+      "epoch": 0.307616,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.97983870967742e-05,
+      "loss": 0.1878,
+      "step": 19226
+    },
+    {
+      "epoch": 0.307632,
+      "grad_norm": 1.9765625,
+      "learning_rate": 6.979677419354839e-05,
+      "loss": 0.1771,
+      "step": 19227
+    },
+    {
+      "epoch": 0.307648,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.979516129032259e-05,
+      "loss": 0.1891,
+      "step": 19228
+    },
+    {
+      "epoch": 0.307664,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.979354838709677e-05,
+      "loss": 0.1668,
+      "step": 19229
+    },
+    {
+      "epoch": 0.30768,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.979193548387097e-05,
+      "loss": 0.1604,
+      "step": 19230
+    },
+    {
+      "epoch": 0.307696,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.979032258064516e-05,
+      "loss": 0.1587,
+      "step": 19231
+    },
+    {
+      "epoch": 0.307712,
+      "grad_norm": 1.515625,
+      "learning_rate": 6.978870967741936e-05,
+      "loss": 0.2134,
+      "step": 19232
+    },
+    {
+      "epoch": 0.307728,
+      "grad_norm": 0.76953125,
+      "learning_rate": 6.978709677419355e-05,
+      "loss": 0.1663,
+      "step": 19233
+    },
+    {
+      "epoch": 0.307744,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.978548387096775e-05,
+      "loss": 0.1463,
+      "step": 19234
+    },
+    {
+      "epoch": 0.30776,
+      "grad_norm": 1.2265625,
+      "learning_rate": 6.978387096774193e-05,
+      "loss": 0.2058,
+      "step": 19235
+    },
+    {
+      "epoch": 0.307776,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.978225806451613e-05,
+      "loss": 0.1835,
+      "step": 19236
+    },
+    {
+      "epoch": 0.307792,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.978064516129033e-05,
+      "loss": 0.1715,
+      "step": 19237
+    },
+    {
+      "epoch": 0.307808,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.977903225806452e-05,
+      "loss": 0.1814,
+      "step": 19238
+    },
+    {
+      "epoch": 0.307824,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.977741935483872e-05,
+      "loss": 0.175,
+      "step": 19239
+    },
+    {
+      "epoch": 0.30784,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.97758064516129e-05,
+      "loss": 0.1844,
+      "step": 19240
+    },
+    {
+      "epoch": 0.307856,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.97741935483871e-05,
+      "loss": 0.1639,
+      "step": 19241
+    },
+    {
+      "epoch": 0.307872,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.977258064516129e-05,
+      "loss": 0.2015,
+      "step": 19242
+    },
+    {
+      "epoch": 0.307888,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.977096774193549e-05,
+      "loss": 0.2138,
+      "step": 19243
+    },
+    {
+      "epoch": 0.307904,
+      "grad_norm": 0.78515625,
+      "learning_rate": 6.976935483870967e-05,
+      "loss": 0.1697,
+      "step": 19244
+    },
+    {
+      "epoch": 0.30792,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.976774193548387e-05,
+      "loss": 0.195,
+      "step": 19245
+    },
+    {
+      "epoch": 0.307936,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.976612903225806e-05,
+      "loss": 0.1624,
+      "step": 19246
+    },
+    {
+      "epoch": 0.307952,
+      "grad_norm": 0.5078125,
+      "learning_rate": 6.976451612903226e-05,
+      "loss": 0.144,
+      "step": 19247
+    },
+    {
+      "epoch": 0.307968,
+      "grad_norm": 1.4375,
+      "learning_rate": 6.976290322580646e-05,
+      "loss": 0.1712,
+      "step": 19248
+    },
+    {
+      "epoch": 0.307984,
+      "grad_norm": 0.87109375,
+      "learning_rate": 6.976129032258066e-05,
+      "loss": 0.1407,
+      "step": 19249
+    },
+    {
+      "epoch": 0.308,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.975967741935484e-05,
+      "loss": 0.1954,
+      "step": 19250
+    },
+    {
+      "epoch": 0.308016,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.975806451612904e-05,
+      "loss": 0.1629,
+      "step": 19251
+    },
+    {
+      "epoch": 0.308032,
+      "grad_norm": 0.76171875,
+      "learning_rate": 6.975645161290323e-05,
+      "loss": 0.1641,
+      "step": 19252
+    },
+    {
+      "epoch": 0.308048,
+      "grad_norm": 0.93359375,
+      "learning_rate": 6.975483870967743e-05,
+      "loss": 0.1292,
+      "step": 19253
+    },
+    {
+      "epoch": 0.308064,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.975322580645162e-05,
+      "loss": 0.1579,
+      "step": 19254
+    },
+    {
+      "epoch": 0.30808,
+      "grad_norm": 0.8046875,
+      "learning_rate": 6.97516129032258e-05,
+      "loss": 0.2319,
+      "step": 19255
+    },
+    {
+      "epoch": 0.308096,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.975e-05,
+      "loss": 0.1673,
+      "step": 19256
+    },
+    {
+      "epoch": 0.308112,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.974838709677419e-05,
+      "loss": 0.1132,
+      "step": 19257
+    },
+    {
+      "epoch": 0.308128,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.974677419354839e-05,
+      "loss": 0.1694,
+      "step": 19258
+    },
+    {
+      "epoch": 0.308144,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.974516129032259e-05,
+      "loss": 0.1705,
+      "step": 19259
+    },
+    {
+      "epoch": 0.30816,
+      "grad_norm": 0.83203125,
+      "learning_rate": 6.974354838709679e-05,
+      "loss": 0.1866,
+      "step": 19260
+    },
+    {
+      "epoch": 0.308176,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.974193548387097e-05,
+      "loss": 0.1643,
+      "step": 19261
+    },
+    {
+      "epoch": 0.308192,
+      "grad_norm": 1.140625,
+      "learning_rate": 6.974032258064517e-05,
+      "loss": 0.1729,
+      "step": 19262
+    },
+    {
+      "epoch": 0.308208,
+      "grad_norm": 0.82421875,
+      "learning_rate": 6.973870967741936e-05,
+      "loss": 0.1633,
+      "step": 19263
+    },
+    {
+      "epoch": 0.308224,
+      "grad_norm": 1.0859375,
+      "learning_rate": 6.973709677419356e-05,
+      "loss": 0.1497,
+      "step": 19264
+    },
+    {
+      "epoch": 0.30824,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.973548387096774e-05,
+      "loss": 0.2037,
+      "step": 19265
+    },
+    {
+      "epoch": 0.308256,
+      "grad_norm": 0.8046875,
+      "learning_rate": 6.973387096774194e-05,
+      "loss": 0.1871,
+      "step": 19266
+    },
+    {
+      "epoch": 0.308272,
+      "grad_norm": 0.78515625,
+      "learning_rate": 6.973225806451613e-05,
+      "loss": 0.1386,
+      "step": 19267
+    },
+    {
+      "epoch": 0.308288,
+      "grad_norm": 0.7421875,
+      "learning_rate": 6.973064516129033e-05,
+      "loss": 0.1793,
+      "step": 19268
+    },
+    {
+      "epoch": 0.308304,
+      "grad_norm": 1.0859375,
+      "learning_rate": 6.972903225806452e-05,
+      "loss": 0.1761,
+      "step": 19269
+    },
+    {
+      "epoch": 0.30832,
+      "grad_norm": 1.0703125,
+      "learning_rate": 6.97274193548387e-05,
+      "loss": 0.2403,
+      "step": 19270
+    },
+    {
+      "epoch": 0.308336,
+      "grad_norm": 0.7890625,
+      "learning_rate": 6.97258064516129e-05,
+      "loss": 0.1482,
+      "step": 19271
+    },
+    {
+      "epoch": 0.308352,
+      "grad_norm": 0.75390625,
+      "learning_rate": 6.97241935483871e-05,
+      "loss": 0.1735,
+      "step": 19272
+    },
+    {
+      "epoch": 0.308368,
+      "grad_norm": 0.67578125,
+      "learning_rate": 6.97225806451613e-05,
+      "loss": 0.1422,
+      "step": 19273
+    },
+    {
+      "epoch": 0.308384,
+      "grad_norm": 0.859375,
+      "learning_rate": 6.972096774193549e-05,
+      "loss": 0.204,
+      "step": 19274
+    },
+    {
+      "epoch": 0.3084,
+      "grad_norm": 1.125,
+      "learning_rate": 6.971935483870969e-05,
+      "loss": 0.2036,
+      "step": 19275
+    },
+    {
+      "epoch": 0.308416,
+      "grad_norm": 0.70703125,
+      "learning_rate": 6.971774193548387e-05,
+      "loss": 0.187,
+      "step": 19276
+    },
+    {
+      "epoch": 0.308432,
+      "grad_norm": 0.87890625,
+      "learning_rate": 6.971612903225807e-05,
+      "loss": 0.1737,
+      "step": 19277
+    },
+    {
+      "epoch": 0.308448,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.971451612903226e-05,
+      "loss": 0.1892,
+      "step": 19278
+    },
+    {
+      "epoch": 0.308464,
+      "grad_norm": 0.9375,
+      "learning_rate": 6.971290322580646e-05,
+      "loss": 0.1659,
+      "step": 19279
+    },
+    {
+      "epoch": 0.30848,
+      "grad_norm": 0.9765625,
+      "learning_rate": 6.971129032258064e-05,
+      "loss": 0.1748,
+      "step": 19280
+    },
+    {
+      "epoch": 0.308496,
+      "grad_norm": 0.75,
+      "learning_rate": 6.970967741935484e-05,
+      "loss": 0.1503,
+      "step": 19281
+    },
+    {
+      "epoch": 0.308512,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.970806451612903e-05,
+      "loss": 0.1699,
+      "step": 19282
+    },
+    {
+      "epoch": 0.308528,
+      "grad_norm": 0.703125,
+      "learning_rate": 6.970645161290323e-05,
+      "loss": 0.1507,
+      "step": 19283
+    },
+    {
+      "epoch": 0.308544,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.970483870967743e-05,
+      "loss": 0.1479,
+      "step": 19284
+    },
+    {
+      "epoch": 0.30856,
+      "grad_norm": 0.73046875,
+      "learning_rate": 6.970322580645161e-05,
+      "loss": 0.1725,
+      "step": 19285
+    },
+    {
+      "epoch": 0.308576,
+      "grad_norm": 0.78515625,
+      "learning_rate": 6.970161290322581e-05,
+      "loss": 0.1714,
+      "step": 19286
+    },
+    {
+      "epoch": 0.308592,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.97e-05,
+      "loss": 0.2191,
+      "step": 19287
+    },
+    {
+      "epoch": 0.308608,
+      "grad_norm": 1.0,
+      "learning_rate": 6.96983870967742e-05,
+      "loss": 0.135,
+      "step": 19288
+    },
+    {
+      "epoch": 0.308624,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.969677419354839e-05,
+      "loss": 0.1373,
+      "step": 19289
+    },
+    {
+      "epoch": 0.30864,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.969516129032259e-05,
+      "loss": 0.1701,
+      "step": 19290
+    },
+    {
+      "epoch": 0.308656,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.969354838709677e-05,
+      "loss": 0.1602,
+      "step": 19291
+    },
+    {
+      "epoch": 0.308672,
+      "grad_norm": 0.83984375,
+      "learning_rate": 6.969193548387097e-05,
+      "loss": 0.1578,
+      "step": 19292
+    },
+    {
+      "epoch": 0.308688,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.969032258064516e-05,
+      "loss": 0.1646,
+      "step": 19293
+    },
+    {
+      "epoch": 0.308704,
+      "grad_norm": 0.99609375,
+      "learning_rate": 6.968870967741936e-05,
+      "loss": 0.2074,
+      "step": 19294
+    },
+    {
+      "epoch": 0.30872,
+      "grad_norm": 1.2734375,
+      "learning_rate": 6.968709677419356e-05,
+      "loss": 0.1867,
+      "step": 19295
+    },
+    {
+      "epoch": 0.308736,
+      "grad_norm": 0.8515625,
+      "learning_rate": 6.968548387096774e-05,
+      "loss": 0.1652,
+      "step": 19296
+    },
+    {
+      "epoch": 0.308752,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.968387096774194e-05,
+      "loss": 0.1606,
+      "step": 19297
+    },
+    {
+      "epoch": 0.308768,
+      "grad_norm": 0.82421875,
+      "learning_rate": 6.968225806451614e-05,
+      "loss": 0.174,
+      "step": 19298
+    },
+    {
+      "epoch": 0.308784,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.968064516129033e-05,
+      "loss": 0.2028,
+      "step": 19299
+    },
+    {
+      "epoch": 0.3088,
+      "grad_norm": 0.7890625,
+      "learning_rate": 6.967903225806453e-05,
+      "loss": 0.1439,
+      "step": 19300
+    },
+    {
+      "epoch": 0.308816,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.967741935483871e-05,
+      "loss": 0.1185,
+      "step": 19301
+    },
+    {
+      "epoch": 0.308832,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.96758064516129e-05,
+      "loss": 0.2008,
+      "step": 19302
+    },
+    {
+      "epoch": 0.308848,
+      "grad_norm": 0.8125,
+      "learning_rate": 6.96741935483871e-05,
+      "loss": 0.2041,
+      "step": 19303
+    },
+    {
+      "epoch": 0.308864,
+      "grad_norm": 0.8046875,
+      "learning_rate": 6.967258064516129e-05,
+      "loss": 0.169,
+      "step": 19304
+    },
+    {
+      "epoch": 0.30888,
+      "grad_norm": 0.7890625,
+      "learning_rate": 6.967096774193549e-05,
+      "loss": 0.1704,
+      "step": 19305
+    },
+    {
+      "epoch": 0.308896,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.966935483870967e-05,
+      "loss": 0.1751,
+      "step": 19306
+    },
+    {
+      "epoch": 0.308912,
+      "grad_norm": 0.4609375,
+      "learning_rate": 6.966774193548387e-05,
+      "loss": 0.1591,
+      "step": 19307
+    },
+    {
+      "epoch": 0.308928,
+      "grad_norm": 0.94921875,
+      "learning_rate": 6.966612903225807e-05,
+      "loss": 0.1725,
+      "step": 19308
+    },
+    {
+      "epoch": 0.308944,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.966451612903227e-05,
+      "loss": 0.16,
+      "step": 19309
+    },
+    {
+      "epoch": 0.30896,
+      "grad_norm": 0.875,
+      "learning_rate": 6.966290322580646e-05,
+      "loss": 0.2102,
+      "step": 19310
+    },
+    {
+      "epoch": 0.308976,
+      "grad_norm": 0.90234375,
+      "learning_rate": 6.966129032258066e-05,
+      "loss": 0.1961,
+      "step": 19311
+    },
+    {
+      "epoch": 0.308992,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.965967741935484e-05,
+      "loss": 0.14,
+      "step": 19312
+    },
+    {
+      "epoch": 0.309008,
+      "grad_norm": 0.8125,
+      "learning_rate": 6.965806451612904e-05,
+      "loss": 0.1588,
+      "step": 19313
+    },
+    {
+      "epoch": 0.309024,
+      "grad_norm": 1.4296875,
+      "learning_rate": 6.965645161290323e-05,
+      "loss": 0.1652,
+      "step": 19314
+    },
+    {
+      "epoch": 0.30904,
+      "grad_norm": 0.625,
+      "learning_rate": 6.965483870967743e-05,
+      "loss": 0.1635,
+      "step": 19315
+    },
+    {
+      "epoch": 0.309056,
+      "grad_norm": 0.890625,
+      "learning_rate": 6.965322580645161e-05,
+      "loss": 0.1927,
+      "step": 19316
+    },
+    {
+      "epoch": 0.309072,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.96516129032258e-05,
+      "loss": 0.1975,
+      "step": 19317
+    },
+    {
+      "epoch": 0.309088,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.965e-05,
+      "loss": 0.1799,
+      "step": 19318
+    },
+    {
+      "epoch": 0.309104,
+      "grad_norm": 0.7734375,
+      "learning_rate": 6.96483870967742e-05,
+      "loss": 0.1676,
+      "step": 19319
+    },
+    {
+      "epoch": 0.30912,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.96467741935484e-05,
+      "loss": 0.1565,
+      "step": 19320
+    },
+    {
+      "epoch": 0.309136,
+      "grad_norm": 0.87890625,
+      "learning_rate": 6.964516129032258e-05,
+      "loss": 0.1609,
+      "step": 19321
+    },
+    {
+      "epoch": 0.309152,
+      "grad_norm": 0.90625,
+      "learning_rate": 6.964354838709678e-05,
+      "loss": 0.2018,
+      "step": 19322
+    },
+    {
+      "epoch": 0.309168,
+      "grad_norm": 0.6953125,
+      "learning_rate": 6.964193548387097e-05,
+      "loss": 0.1218,
+      "step": 19323
+    },
+    {
+      "epoch": 0.309184,
+      "grad_norm": 0.84765625,
+      "learning_rate": 6.964032258064517e-05,
+      "loss": 0.16,
+      "step": 19324
+    },
+    {
+      "epoch": 0.3092,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.963870967741936e-05,
+      "loss": 0.1815,
+      "step": 19325
+    },
+    {
+      "epoch": 0.309216,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.963709677419356e-05,
+      "loss": 0.1809,
+      "step": 19326
+    },
+    {
+      "epoch": 0.309232,
+      "grad_norm": 1.140625,
+      "learning_rate": 6.963548387096774e-05,
+      "loss": 0.1772,
+      "step": 19327
+    },
+    {
+      "epoch": 0.309248,
+      "grad_norm": 0.8671875,
+      "learning_rate": 6.963387096774194e-05,
+      "loss": 0.2091,
+      "step": 19328
+    },
+    {
+      "epoch": 0.309264,
+      "grad_norm": 0.74609375,
+      "learning_rate": 6.963225806451613e-05,
+      "loss": 0.213,
+      "step": 19329
+    },
+    {
+      "epoch": 0.30928,
+      "grad_norm": 0.94921875,
+      "learning_rate": 6.963064516129033e-05,
+      "loss": 0.1784,
+      "step": 19330
+    },
+    {
+      "epoch": 0.309296,
+      "grad_norm": 0.96484375,
+      "learning_rate": 6.962903225806451e-05,
+      "loss": 0.2127,
+      "step": 19331
+    },
+    {
+      "epoch": 0.309312,
+      "grad_norm": 0.75,
+      "learning_rate": 6.962741935483871e-05,
+      "loss": 0.1439,
+      "step": 19332
+    },
+    {
+      "epoch": 0.309328,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.962580645161291e-05,
+      "loss": 0.1891,
+      "step": 19333
+    },
+    {
+      "epoch": 0.309344,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.96241935483871e-05,
+      "loss": 0.1781,
+      "step": 19334
+    },
+    {
+      "epoch": 0.30936,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.96225806451613e-05,
+      "loss": 0.1623,
+      "step": 19335
+    },
+    {
+      "epoch": 0.309376,
+      "grad_norm": 0.625,
+      "learning_rate": 6.962096774193548e-05,
+      "loss": 0.1677,
+      "step": 19336
+    },
+    {
+      "epoch": 0.309392,
+      "grad_norm": 0.78515625,
+      "learning_rate": 6.961935483870968e-05,
+      "loss": 0.1831,
+      "step": 19337
+    },
+    {
+      "epoch": 0.309408,
+      "grad_norm": 0.7578125,
+      "learning_rate": 6.961774193548387e-05,
+      "loss": 0.1488,
+      "step": 19338
+    },
+    {
+      "epoch": 0.309424,
+      "grad_norm": 0.7421875,
+      "learning_rate": 6.961612903225807e-05,
+      "loss": 0.1683,
+      "step": 19339
+    },
+    {
+      "epoch": 0.30944,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.961451612903226e-05,
+      "loss": 0.1583,
+      "step": 19340
+    },
+    {
+      "epoch": 0.309456,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.961290322580645e-05,
+      "loss": 0.1345,
+      "step": 19341
+    },
+    {
+      "epoch": 0.309472,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.961129032258064e-05,
+      "loss": 0.1623,
+      "step": 19342
+    },
+    {
+      "epoch": 0.309488,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.960967741935484e-05,
+      "loss": 0.1629,
+      "step": 19343
+    },
+    {
+      "epoch": 0.309504,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.960806451612904e-05,
+      "loss": 0.1738,
+      "step": 19344
+    },
+    {
+      "epoch": 0.30952,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.960645161290324e-05,
+      "loss": 0.1697,
+      "step": 19345
+    },
+    {
+      "epoch": 0.309536,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.960483870967743e-05,
+      "loss": 0.1696,
+      "step": 19346
+    },
+    {
+      "epoch": 0.309552,
+      "grad_norm": 0.83203125,
+      "learning_rate": 6.960322580645161e-05,
+      "loss": 0.154,
+      "step": 19347
+    },
+    {
+      "epoch": 0.309568,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.960161290322581e-05,
+      "loss": 0.1712,
+      "step": 19348
+    },
+    {
+      "epoch": 0.309584,
+      "grad_norm": 0.6953125,
+      "learning_rate": 6.96e-05,
+      "loss": 0.1755,
+      "step": 19349
+    },
+    {
+      "epoch": 0.3096,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.95983870967742e-05,
+      "loss": 0.1729,
+      "step": 19350
+    },
+    {
+      "epoch": 0.309616,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.959677419354838e-05,
+      "loss": 0.157,
+      "step": 19351
+    },
+    {
+      "epoch": 0.309632,
+      "grad_norm": 0.8515625,
+      "learning_rate": 6.959516129032258e-05,
+      "loss": 0.1859,
+      "step": 19352
+    },
+    {
+      "epoch": 0.309648,
+      "grad_norm": 0.80078125,
+      "learning_rate": 6.959354838709677e-05,
+      "loss": 0.1624,
+      "step": 19353
+    },
+    {
+      "epoch": 0.309664,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.959193548387097e-05,
+      "loss": 0.1738,
+      "step": 19354
+    },
+    {
+      "epoch": 0.30968,
+      "grad_norm": 0.85546875,
+      "learning_rate": 6.959032258064517e-05,
+      "loss": 0.1601,
+      "step": 19355
+    },
+    {
+      "epoch": 0.309696,
+      "grad_norm": 1.1953125,
+      "learning_rate": 6.958870967741937e-05,
+      "loss": 0.1465,
+      "step": 19356
+    },
+    {
+      "epoch": 0.309712,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.958709677419355e-05,
+      "loss": 0.1465,
+      "step": 19357
+    },
+    {
+      "epoch": 0.309728,
+      "grad_norm": 0.79296875,
+      "learning_rate": 6.958548387096775e-05,
+      "loss": 0.1991,
+      "step": 19358
+    },
+    {
+      "epoch": 0.309744,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.958387096774194e-05,
+      "loss": 0.1692,
+      "step": 19359
+    },
+    {
+      "epoch": 0.30976,
+      "grad_norm": 0.81640625,
+      "learning_rate": 6.958225806451614e-05,
+      "loss": 0.1744,
+      "step": 19360
+    },
+    {
+      "epoch": 0.309776,
+      "grad_norm": 0.765625,
+      "learning_rate": 6.958064516129033e-05,
+      "loss": 0.1592,
+      "step": 19361
+    },
+    {
+      "epoch": 0.309792,
+      "grad_norm": 0.82421875,
+      "learning_rate": 6.957903225806453e-05,
+      "loss": 0.1562,
+      "step": 19362
+    },
+    {
+      "epoch": 0.309808,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.957741935483871e-05,
+      "loss": 0.164,
+      "step": 19363
+    },
+    {
+      "epoch": 0.309824,
+      "grad_norm": 0.75390625,
+      "learning_rate": 6.95758064516129e-05,
+      "loss": 0.1748,
+      "step": 19364
+    },
+    {
+      "epoch": 0.30984,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.95741935483871e-05,
+      "loss": 0.2158,
+      "step": 19365
+    },
+    {
+      "epoch": 0.309856,
+      "grad_norm": 0.83203125,
+      "learning_rate": 6.957258064516128e-05,
+      "loss": 0.1635,
+      "step": 19366
+    },
+    {
+      "epoch": 0.309872,
+      "grad_norm": 0.8203125,
+      "learning_rate": 6.957096774193548e-05,
+      "loss": 0.1664,
+      "step": 19367
+    },
+    {
+      "epoch": 0.309888,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.956935483870968e-05,
+      "loss": 0.2025,
+      "step": 19368
+    },
+    {
+      "epoch": 0.309904,
+      "grad_norm": 0.8046875,
+      "learning_rate": 6.956774193548388e-05,
+      "loss": 0.1461,
+      "step": 19369
+    },
+    {
+      "epoch": 0.30992,
+      "grad_norm": 1.2421875,
+      "learning_rate": 6.956612903225807e-05,
+      "loss": 0.1809,
+      "step": 19370
+    },
+    {
+      "epoch": 0.309936,
+      "grad_norm": 0.76953125,
+      "learning_rate": 6.956451612903227e-05,
+      "loss": 0.1796,
+      "step": 19371
+    },
+    {
+      "epoch": 0.309952,
+      "grad_norm": 1.34375,
+      "learning_rate": 6.956290322580645e-05,
+      "loss": 0.2089,
+      "step": 19372
+    },
+    {
+      "epoch": 0.309968,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.956129032258065e-05,
+      "loss": 0.2313,
+      "step": 19373
+    },
+    {
+      "epoch": 0.309984,
+      "grad_norm": 0.80859375,
+      "learning_rate": 6.955967741935484e-05,
+      "loss": 0.1981,
+      "step": 19374
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.78515625,
+      "learning_rate": 6.955806451612904e-05,
+      "loss": 0.1431,
+      "step": 19375
+    },
+    {
+      "epoch": 0.310016,
+      "grad_norm": 0.75390625,
+      "learning_rate": 6.955645161290323e-05,
+      "loss": 0.2087,
+      "step": 19376
+    },
+    {
+      "epoch": 0.310032,
+      "grad_norm": 0.70703125,
+      "learning_rate": 6.955483870967742e-05,
+      "loss": 0.1625,
+      "step": 19377
+    },
+    {
+      "epoch": 0.310048,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.955322580645161e-05,
+      "loss": 0.1723,
+      "step": 19378
+    },
+    {
+      "epoch": 0.310064,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.955161290322581e-05,
+      "loss": 0.1435,
+      "step": 19379
+    },
+    {
+      "epoch": 0.31008,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.955000000000001e-05,
+      "loss": 0.1346,
+      "step": 19380
+    },
+    {
+      "epoch": 0.310096,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.95483870967742e-05,
+      "loss": 0.1591,
+      "step": 19381
+    },
+    {
+      "epoch": 0.310112,
+      "grad_norm": 1.125,
+      "learning_rate": 6.95467741935484e-05,
+      "loss": 0.2151,
+      "step": 19382
+    },
+    {
+      "epoch": 0.310128,
+      "grad_norm": 0.703125,
+      "learning_rate": 6.954516129032258e-05,
+      "loss": 0.1906,
+      "step": 19383
+    },
+    {
+      "epoch": 0.310144,
+      "grad_norm": 0.5078125,
+      "learning_rate": 6.954354838709678e-05,
+      "loss": 0.1903,
+      "step": 19384
+    },
+    {
+      "epoch": 0.31016,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.954193548387097e-05,
+      "loss": 0.1765,
+      "step": 19385
+    },
+    {
+      "epoch": 0.310176,
+      "grad_norm": 0.8203125,
+      "learning_rate": 6.954032258064517e-05,
+      "loss": 0.1683,
+      "step": 19386
+    },
+    {
+      "epoch": 0.310192,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.953870967741935e-05,
+      "loss": 0.1654,
+      "step": 19387
+    },
+    {
+      "epoch": 0.310208,
+      "grad_norm": 0.87109375,
+      "learning_rate": 6.953709677419355e-05,
+      "loss": 0.1609,
+      "step": 19388
+    },
+    {
+      "epoch": 0.310224,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.953548387096774e-05,
+      "loss": 0.1608,
+      "step": 19389
+    },
+    {
+      "epoch": 0.31024,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.953387096774194e-05,
+      "loss": 0.1448,
+      "step": 19390
+    },
+    {
+      "epoch": 0.310256,
+      "grad_norm": 0.80078125,
+      "learning_rate": 6.953225806451614e-05,
+      "loss": 0.1933,
+      "step": 19391
+    },
+    {
+      "epoch": 0.310272,
+      "grad_norm": 0.82421875,
+      "learning_rate": 6.953064516129032e-05,
+      "loss": 0.1651,
+      "step": 19392
+    },
+    {
+      "epoch": 0.310288,
+      "grad_norm": 0.765625,
+      "learning_rate": 6.952903225806452e-05,
+      "loss": 0.1808,
+      "step": 19393
+    },
+    {
+      "epoch": 0.310304,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.952741935483871e-05,
+      "loss": 0.1337,
+      "step": 19394
+    },
+    {
+      "epoch": 0.31032,
+      "grad_norm": 0.890625,
+      "learning_rate": 6.952580645161291e-05,
+      "loss": 0.1912,
+      "step": 19395
+    },
+    {
+      "epoch": 0.310336,
+      "grad_norm": 0.69921875,
+      "learning_rate": 6.95241935483871e-05,
+      "loss": 0.1528,
+      "step": 19396
+    },
+    {
+      "epoch": 0.310352,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.95225806451613e-05,
+      "loss": 0.1743,
+      "step": 19397
+    },
+    {
+      "epoch": 0.310368,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.952096774193548e-05,
+      "loss": 0.2002,
+      "step": 19398
+    },
+    {
+      "epoch": 0.310384,
+      "grad_norm": 0.80078125,
+      "learning_rate": 6.951935483870968e-05,
+      "loss": 0.1691,
+      "step": 19399
+    },
+    {
+      "epoch": 0.3104,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.951774193548387e-05,
+      "loss": 0.1311,
+      "step": 19400
+    },
+    {
+      "epoch": 0.310416,
+      "grad_norm": 0.76953125,
+      "learning_rate": 6.951612903225807e-05,
+      "loss": 0.171,
+      "step": 19401
+    },
+    {
+      "epoch": 0.310432,
+      "grad_norm": 0.94921875,
+      "learning_rate": 6.951451612903225e-05,
+      "loss": 0.1893,
+      "step": 19402
+    },
+    {
+      "epoch": 0.310448,
+      "grad_norm": 0.703125,
+      "learning_rate": 6.951290322580645e-05,
+      "loss": 0.1883,
+      "step": 19403
+    },
+    {
+      "epoch": 0.310464,
+      "grad_norm": 0.6953125,
+      "learning_rate": 6.951129032258065e-05,
+      "loss": 0.1724,
+      "step": 19404
+    },
+    {
+      "epoch": 0.31048,
+      "grad_norm": 0.96484375,
+      "learning_rate": 6.950967741935485e-05,
+      "loss": 0.1748,
+      "step": 19405
+    },
+    {
+      "epoch": 0.310496,
+      "grad_norm": 1.1328125,
+      "learning_rate": 6.950806451612904e-05,
+      "loss": 0.2063,
+      "step": 19406
+    },
+    {
+      "epoch": 0.310512,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.950645161290324e-05,
+      "loss": 0.1648,
+      "step": 19407
+    },
+    {
+      "epoch": 0.310528,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.950483870967742e-05,
+      "loss": 0.2224,
+      "step": 19408
+    },
+    {
+      "epoch": 0.310544,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.950322580645162e-05,
+      "loss": 0.1698,
+      "step": 19409
+    },
+    {
+      "epoch": 0.31056,
+      "grad_norm": 0.984375,
+      "learning_rate": 6.950161290322581e-05,
+      "loss": 0.2197,
+      "step": 19410
+    },
+    {
+      "epoch": 0.310576,
+      "grad_norm": 0.7890625,
+      "learning_rate": 6.95e-05,
+      "loss": 0.1373,
+      "step": 19411
+    },
+    {
+      "epoch": 0.310592,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.94983870967742e-05,
+      "loss": 0.1471,
+      "step": 19412
+    },
+    {
+      "epoch": 0.310608,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.949677419354838e-05,
+      "loss": 0.1291,
+      "step": 19413
+    },
+    {
+      "epoch": 0.310624,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.949516129032258e-05,
+      "loss": 0.1518,
+      "step": 19414
+    },
+    {
+      "epoch": 0.31064,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.949354838709678e-05,
+      "loss": 0.1713,
+      "step": 19415
+    },
+    {
+      "epoch": 0.310656,
+      "grad_norm": 0.75,
+      "learning_rate": 6.949193548387098e-05,
+      "loss": 0.1635,
+      "step": 19416
+    },
+    {
+      "epoch": 0.310672,
+      "grad_norm": 0.75390625,
+      "learning_rate": 6.949032258064517e-05,
+      "loss": 0.1709,
+      "step": 19417
+    },
+    {
+      "epoch": 0.310688,
+      "grad_norm": 1.1953125,
+      "learning_rate": 6.948870967741937e-05,
+      "loss": 0.1534,
+      "step": 19418
+    },
+    {
+      "epoch": 0.310704,
+      "grad_norm": 0.8515625,
+      "learning_rate": 6.948709677419355e-05,
+      "loss": 0.1904,
+      "step": 19419
+    },
+    {
+      "epoch": 0.31072,
+      "grad_norm": 0.96484375,
+      "learning_rate": 6.948548387096775e-05,
+      "loss": 0.1678,
+      "step": 19420
+    },
+    {
+      "epoch": 0.310736,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.948387096774194e-05,
+      "loss": 0.1456,
+      "step": 19421
+    },
+    {
+      "epoch": 0.310752,
+      "grad_norm": 1.8125,
+      "learning_rate": 6.948225806451614e-05,
+      "loss": 0.1281,
+      "step": 19422
+    },
+    {
+      "epoch": 0.310768,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.948064516129032e-05,
+      "loss": 0.1707,
+      "step": 19423
+    },
+    {
+      "epoch": 0.310784,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.947903225806452e-05,
+      "loss": 0.136,
+      "step": 19424
+    },
+    {
+      "epoch": 0.3108,
+      "grad_norm": 1.3046875,
+      "learning_rate": 6.947741935483871e-05,
+      "loss": 0.1819,
+      "step": 19425
+    },
+    {
+      "epoch": 0.310816,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.94758064516129e-05,
+      "loss": 0.1714,
+      "step": 19426
+    },
+    {
+      "epoch": 0.310832,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.94741935483871e-05,
+      "loss": 0.1732,
+      "step": 19427
+    },
+    {
+      "epoch": 0.310848,
+      "grad_norm": 1.2734375,
+      "learning_rate": 6.94725806451613e-05,
+      "loss": 0.2058,
+      "step": 19428
+    },
+    {
+      "epoch": 0.310864,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.94709677419355e-05,
+      "loss": 0.1667,
+      "step": 19429
+    },
+    {
+      "epoch": 0.31088,
+      "grad_norm": 0.83203125,
+      "learning_rate": 6.946935483870968e-05,
+      "loss": 0.1384,
+      "step": 19430
+    },
+    {
+      "epoch": 0.310896,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.946774193548388e-05,
+      "loss": 0.1724,
+      "step": 19431
+    },
+    {
+      "epoch": 0.310912,
+      "grad_norm": 1.140625,
+      "learning_rate": 6.946612903225807e-05,
+      "loss": 0.1963,
+      "step": 19432
+    },
+    {
+      "epoch": 0.310928,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.946451612903227e-05,
+      "loss": 0.1568,
+      "step": 19433
+    },
+    {
+      "epoch": 0.310944,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.946290322580645e-05,
+      "loss": 0.1798,
+      "step": 19434
+    },
+    {
+      "epoch": 0.31096,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.946129032258065e-05,
+      "loss": 0.1937,
+      "step": 19435
+    },
+    {
+      "epoch": 0.310976,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.945967741935484e-05,
+      "loss": 0.1808,
+      "step": 19436
+    },
+    {
+      "epoch": 0.310992,
+      "grad_norm": 0.89453125,
+      "learning_rate": 6.945806451612904e-05,
+      "loss": 0.1152,
+      "step": 19437
+    },
+    {
+      "epoch": 0.311008,
+      "grad_norm": 0.765625,
+      "learning_rate": 6.945645161290322e-05,
+      "loss": 0.1957,
+      "step": 19438
+    },
+    {
+      "epoch": 0.311024,
+      "grad_norm": 0.70703125,
+      "learning_rate": 6.945483870967742e-05,
+      "loss": 0.1767,
+      "step": 19439
+    },
+    {
+      "epoch": 0.31104,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.945322580645162e-05,
+      "loss": 0.1522,
+      "step": 19440
+    },
+    {
+      "epoch": 0.311056,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.945161290322581e-05,
+      "loss": 0.2026,
+      "step": 19441
+    },
+    {
+      "epoch": 0.311072,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.945000000000001e-05,
+      "loss": 0.142,
+      "step": 19442
+    },
+    {
+      "epoch": 0.311088,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.94483870967742e-05,
+      "loss": 0.1812,
+      "step": 19443
+    },
+    {
+      "epoch": 0.311104,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.94467741935484e-05,
+      "loss": 0.1821,
+      "step": 19444
+    },
+    {
+      "epoch": 0.31112,
+      "grad_norm": 0.890625,
+      "learning_rate": 6.944516129032258e-05,
+      "loss": 0.2182,
+      "step": 19445
+    },
+    {
+      "epoch": 0.311136,
+      "grad_norm": 0.77734375,
+      "learning_rate": 6.944354838709678e-05,
+      "loss": 0.1445,
+      "step": 19446
+    },
+    {
+      "epoch": 0.311152,
+      "grad_norm": 1.390625,
+      "learning_rate": 6.944193548387097e-05,
+      "loss": 0.1714,
+      "step": 19447
+    },
+    {
+      "epoch": 0.311168,
+      "grad_norm": 0.75390625,
+      "learning_rate": 6.944032258064516e-05,
+      "loss": 0.1182,
+      "step": 19448
+    },
+    {
+      "epoch": 0.311184,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.943870967741935e-05,
+      "loss": 0.1724,
+      "step": 19449
+    },
+    {
+      "epoch": 0.3112,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.943709677419355e-05,
+      "loss": 0.1325,
+      "step": 19450
+    },
+    {
+      "epoch": 0.311216,
+      "grad_norm": 0.8984375,
+      "learning_rate": 6.943548387096775e-05,
+      "loss": 0.1823,
+      "step": 19451
+    },
+    {
+      "epoch": 0.311232,
+      "grad_norm": 0.703125,
+      "learning_rate": 6.943387096774195e-05,
+      "loss": 0.1601,
+      "step": 19452
+    },
+    {
+      "epoch": 0.311248,
+      "grad_norm": 0.828125,
+      "learning_rate": 6.943225806451614e-05,
+      "loss": 0.19,
+      "step": 19453
+    },
+    {
+      "epoch": 0.311264,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.943064516129034e-05,
+      "loss": 0.1511,
+      "step": 19454
+    },
+    {
+      "epoch": 0.31128,
+      "grad_norm": 0.7890625,
+      "learning_rate": 6.942903225806452e-05,
+      "loss": 0.1762,
+      "step": 19455
+    },
+    {
+      "epoch": 0.311296,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.942741935483871e-05,
+      "loss": 0.1286,
+      "step": 19456
+    },
+    {
+      "epoch": 0.311312,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.942580645161291e-05,
+      "loss": 0.1742,
+      "step": 19457
+    },
+    {
+      "epoch": 0.311328,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.94241935483871e-05,
+      "loss": 0.1861,
+      "step": 19458
+    },
+    {
+      "epoch": 0.311344,
+      "grad_norm": 0.8359375,
+      "learning_rate": 6.942258064516129e-05,
+      "loss": 0.1663,
+      "step": 19459
+    },
+    {
+      "epoch": 0.31136,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.942096774193548e-05,
+      "loss": 0.1337,
+      "step": 19460
+    },
+    {
+      "epoch": 0.311376,
+      "grad_norm": 1.1953125,
+      "learning_rate": 6.941935483870968e-05,
+      "loss": 0.1446,
+      "step": 19461
+    },
+    {
+      "epoch": 0.311392,
+      "grad_norm": 0.79296875,
+      "learning_rate": 6.941774193548386e-05,
+      "loss": 0.1603,
+      "step": 19462
+    },
+    {
+      "epoch": 0.311408,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.941612903225806e-05,
+      "loss": 0.2012,
+      "step": 19463
+    },
+    {
+      "epoch": 0.311424,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.941451612903226e-05,
+      "loss": 0.1984,
+      "step": 19464
+    },
+    {
+      "epoch": 0.31144,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.941290322580646e-05,
+      "loss": 0.1976,
+      "step": 19465
+    },
+    {
+      "epoch": 0.311456,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.941129032258065e-05,
+      "loss": 0.1364,
+      "step": 19466
+    },
+    {
+      "epoch": 0.311472,
+      "grad_norm": 0.78515625,
+      "learning_rate": 6.940967741935485e-05,
+      "loss": 0.1647,
+      "step": 19467
+    },
+    {
+      "epoch": 0.311488,
+      "grad_norm": 1.3828125,
+      "learning_rate": 6.940806451612904e-05,
+      "loss": 0.1786,
+      "step": 19468
+    },
+    {
+      "epoch": 0.311504,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.940645161290323e-05,
+      "loss": 0.1285,
+      "step": 19469
+    },
+    {
+      "epoch": 0.31152,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.940483870967742e-05,
+      "loss": 0.1603,
+      "step": 19470
+    },
+    {
+      "epoch": 0.311536,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.940322580645162e-05,
+      "loss": 0.195,
+      "step": 19471
+    },
+    {
+      "epoch": 0.311552,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.94016129032258e-05,
+      "loss": 0.171,
+      "step": 19472
+    },
+    {
+      "epoch": 0.311568,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.939999999999999e-05,
+      "loss": 0.1692,
+      "step": 19473
+    },
+    {
+      "epoch": 0.311584,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.939838709677419e-05,
+      "loss": 0.1673,
+      "step": 19474
+    },
+    {
+      "epoch": 0.3116,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.939677419354839e-05,
+      "loss": 0.1555,
+      "step": 19475
+    },
+    {
+      "epoch": 0.311616,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.939516129032259e-05,
+      "loss": 0.1486,
+      "step": 19476
+    },
+    {
+      "epoch": 0.311632,
+      "grad_norm": 0.8515625,
+      "learning_rate": 6.939354838709678e-05,
+      "loss": 0.129,
+      "step": 19477
+    },
+    {
+      "epoch": 0.311648,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.939193548387098e-05,
+      "loss": 0.1673,
+      "step": 19478
+    },
+    {
+      "epoch": 0.311664,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.939032258064516e-05,
+      "loss": 0.1773,
+      "step": 19479
+    },
+    {
+      "epoch": 0.31168,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.938870967741936e-05,
+      "loss": 0.1519,
+      "step": 19480
+    },
+    {
+      "epoch": 0.311696,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.938709677419355e-05,
+      "loss": 0.1828,
+      "step": 19481
+    },
+    {
+      "epoch": 0.311712,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.938548387096775e-05,
+      "loss": 0.144,
+      "step": 19482
+    },
+    {
+      "epoch": 0.311728,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.938387096774193e-05,
+      "loss": 0.2133,
+      "step": 19483
+    },
+    {
+      "epoch": 0.311744,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.938225806451613e-05,
+      "loss": 0.1552,
+      "step": 19484
+    },
+    {
+      "epoch": 0.31176,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.938064516129032e-05,
+      "loss": 0.1587,
+      "step": 19485
+    },
+    {
+      "epoch": 0.311776,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.937903225806452e-05,
+      "loss": 0.1795,
+      "step": 19486
+    },
+    {
+      "epoch": 0.311792,
+      "grad_norm": 0.78515625,
+      "learning_rate": 6.93774193548387e-05,
+      "loss": 0.1853,
+      "step": 19487
+    },
+    {
+      "epoch": 0.311808,
+      "grad_norm": 0.85546875,
+      "learning_rate": 6.93758064516129e-05,
+      "loss": 0.147,
+      "step": 19488
+    },
+    {
+      "epoch": 0.311824,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.93741935483871e-05,
+      "loss": 0.1865,
+      "step": 19489
+    },
+    {
+      "epoch": 0.31184,
+      "grad_norm": 0.9375,
+      "learning_rate": 6.937258064516129e-05,
+      "loss": 0.1823,
+      "step": 19490
+    },
+    {
+      "epoch": 0.311856,
+      "grad_norm": 0.84765625,
+      "learning_rate": 6.937096774193549e-05,
+      "loss": 0.1913,
+      "step": 19491
+    },
+    {
+      "epoch": 0.311872,
+      "grad_norm": 0.8671875,
+      "learning_rate": 6.936935483870968e-05,
+      "loss": 0.2302,
+      "step": 19492
+    },
+    {
+      "epoch": 0.311888,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.936774193548388e-05,
+      "loss": 0.1657,
+      "step": 19493
+    },
+    {
+      "epoch": 0.311904,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.936612903225806e-05,
+      "loss": 0.147,
+      "step": 19494
+    },
+    {
+      "epoch": 0.31192,
+      "grad_norm": 0.76953125,
+      "learning_rate": 6.936451612903226e-05,
+      "loss": 0.1792,
+      "step": 19495
+    },
+    {
+      "epoch": 0.311936,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.936290322580645e-05,
+      "loss": 0.1423,
+      "step": 19496
+    },
+    {
+      "epoch": 0.311952,
+      "grad_norm": 0.83203125,
+      "learning_rate": 6.936129032258065e-05,
+      "loss": 0.1769,
+      "step": 19497
+    },
+    {
+      "epoch": 0.311968,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.935967741935483e-05,
+      "loss": 0.1502,
+      "step": 19498
+    },
+    {
+      "epoch": 0.311984,
+      "grad_norm": 1.1953125,
+      "learning_rate": 6.935806451612903e-05,
+      "loss": 0.1645,
+      "step": 19499
+    },
+    {
+      "epoch": 0.312,
+      "grad_norm": 0.69921875,
+      "learning_rate": 6.935645161290323e-05,
+      "loss": 0.1224,
+      "step": 19500
+    },
+    {
+      "epoch": 0.312016,
+      "grad_norm": 0.69921875,
+      "learning_rate": 6.935483870967743e-05,
+      "loss": 0.1472,
+      "step": 19501
+    },
+    {
+      "epoch": 0.312032,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.935322580645162e-05,
+      "loss": 0.188,
+      "step": 19502
+    },
+    {
+      "epoch": 0.312048,
+      "grad_norm": 0.7421875,
+      "learning_rate": 6.93516129032258e-05,
+      "loss": 0.1841,
+      "step": 19503
+    },
+    {
+      "epoch": 0.312064,
+      "grad_norm": 0.8203125,
+      "learning_rate": 6.935e-05,
+      "loss": 0.1428,
+      "step": 19504
+    },
+    {
+      "epoch": 0.31208,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.934838709677419e-05,
+      "loss": 0.1664,
+      "step": 19505
+    },
+    {
+      "epoch": 0.312096,
+      "grad_norm": 0.8046875,
+      "learning_rate": 6.934677419354839e-05,
+      "loss": 0.1547,
+      "step": 19506
+    },
+    {
+      "epoch": 0.312112,
+      "grad_norm": 0.83984375,
+      "learning_rate": 6.934516129032258e-05,
+      "loss": 0.1918,
+      "step": 19507
+    },
+    {
+      "epoch": 0.312128,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.934354838709678e-05,
+      "loss": 0.1458,
+      "step": 19508
+    },
+    {
+      "epoch": 0.312144,
+      "grad_norm": 0.78515625,
+      "learning_rate": 6.934193548387096e-05,
+      "loss": 0.1498,
+      "step": 19509
+    },
+    {
+      "epoch": 0.31216,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.934032258064516e-05,
+      "loss": 0.1476,
+      "step": 19510
+    },
+    {
+      "epoch": 0.312176,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.933870967741936e-05,
+      "loss": 0.1942,
+      "step": 19511
+    },
+    {
+      "epoch": 0.312192,
+      "grad_norm": 0.87109375,
+      "learning_rate": 6.933709677419356e-05,
+      "loss": 0.1588,
+      "step": 19512
+    },
+    {
+      "epoch": 0.312208,
+      "grad_norm": 0.70703125,
+      "learning_rate": 6.933548387096775e-05,
+      "loss": 0.172,
+      "step": 19513
+    },
+    {
+      "epoch": 0.312224,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.933387096774195e-05,
+      "loss": 0.1467,
+      "step": 19514
+    },
+    {
+      "epoch": 0.31224,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.933225806451613e-05,
+      "loss": 0.1846,
+      "step": 19515
+    },
+    {
+      "epoch": 0.312256,
+      "grad_norm": 0.890625,
+      "learning_rate": 6.933064516129033e-05,
+      "loss": 0.1585,
+      "step": 19516
+    },
+    {
+      "epoch": 0.312272,
+      "grad_norm": 0.984375,
+      "learning_rate": 6.932903225806452e-05,
+      "loss": 0.2016,
+      "step": 19517
+    },
+    {
+      "epoch": 0.312288,
+      "grad_norm": 1.28125,
+      "learning_rate": 6.932741935483872e-05,
+      "loss": 0.2018,
+      "step": 19518
+    },
+    {
+      "epoch": 0.312304,
+      "grad_norm": 0.921875,
+      "learning_rate": 6.93258064516129e-05,
+      "loss": 0.1842,
+      "step": 19519
+    },
+    {
+      "epoch": 0.31232,
+      "grad_norm": 0.88671875,
+      "learning_rate": 6.932419354838709e-05,
+      "loss": 0.1595,
+      "step": 19520
+    },
+    {
+      "epoch": 0.312336,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.932258064516129e-05,
+      "loss": 0.2185,
+      "step": 19521
+    },
+    {
+      "epoch": 0.312352,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.932096774193548e-05,
+      "loss": 0.1965,
+      "step": 19522
+    },
+    {
+      "epoch": 0.312368,
+      "grad_norm": 0.625,
+      "learning_rate": 6.931935483870968e-05,
+      "loss": 0.1573,
+      "step": 19523
+    },
+    {
+      "epoch": 0.312384,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.931774193548388e-05,
+      "loss": 0.1691,
+      "step": 19524
+    },
+    {
+      "epoch": 0.3124,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.931612903225808e-05,
+      "loss": 0.1482,
+      "step": 19525
+    },
+    {
+      "epoch": 0.312416,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.931451612903226e-05,
+      "loss": 0.123,
+      "step": 19526
+    },
+    {
+      "epoch": 0.312432,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.931290322580646e-05,
+      "loss": 0.1435,
+      "step": 19527
+    },
+    {
+      "epoch": 0.312448,
+      "grad_norm": 1.0859375,
+      "learning_rate": 6.931129032258065e-05,
+      "loss": 0.2008,
+      "step": 19528
+    },
+    {
+      "epoch": 0.312464,
+      "grad_norm": 1.1640625,
+      "learning_rate": 6.930967741935485e-05,
+      "loss": 0.1809,
+      "step": 19529
+    },
+    {
+      "epoch": 0.31248,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.930806451612903e-05,
+      "loss": 0.1528,
+      "step": 19530
+    },
+    {
+      "epoch": 0.312496,
+      "grad_norm": 0.69921875,
+      "learning_rate": 6.930645161290323e-05,
+      "loss": 0.1578,
+      "step": 19531
+    },
+    {
+      "epoch": 0.312512,
+      "grad_norm": 0.85546875,
+      "learning_rate": 6.930483870967742e-05,
+      "loss": 0.1886,
+      "step": 19532
+    },
+    {
+      "epoch": 0.312528,
+      "grad_norm": 0.80859375,
+      "learning_rate": 6.930322580645162e-05,
+      "loss": 0.1688,
+      "step": 19533
+    },
+    {
+      "epoch": 0.312544,
+      "grad_norm": 1.1796875,
+      "learning_rate": 6.93016129032258e-05,
+      "loss": 0.1819,
+      "step": 19534
+    },
+    {
+      "epoch": 0.31256,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.93e-05,
+      "loss": 0.1387,
+      "step": 19535
+    },
+    {
+      "epoch": 0.312576,
+      "grad_norm": 0.82421875,
+      "learning_rate": 6.92983870967742e-05,
+      "loss": 0.1929,
+      "step": 19536
+    },
+    {
+      "epoch": 0.312592,
+      "grad_norm": 0.99609375,
+      "learning_rate": 6.929677419354839e-05,
+      "loss": 0.1903,
+      "step": 19537
+    },
+    {
+      "epoch": 0.312608,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.929516129032259e-05,
+      "loss": 0.1744,
+      "step": 19538
+    },
+    {
+      "epoch": 0.312624,
+      "grad_norm": 0.859375,
+      "learning_rate": 6.929354838709678e-05,
+      "loss": 0.2181,
+      "step": 19539
+    },
+    {
+      "epoch": 0.31264,
+      "grad_norm": 0.76171875,
+      "learning_rate": 6.929193548387098e-05,
+      "loss": 0.1759,
+      "step": 19540
+    },
+    {
+      "epoch": 0.312656,
+      "grad_norm": 0.625,
+      "learning_rate": 6.929032258064516e-05,
+      "loss": 0.1778,
+      "step": 19541
+    },
+    {
+      "epoch": 0.312672,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.928870967741936e-05,
+      "loss": 0.1635,
+      "step": 19542
+    },
+    {
+      "epoch": 0.312688,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.928709677419355e-05,
+      "loss": 0.1591,
+      "step": 19543
+    },
+    {
+      "epoch": 0.312704,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.928548387096775e-05,
+      "loss": 0.144,
+      "step": 19544
+    },
+    {
+      "epoch": 0.31272,
+      "grad_norm": 0.859375,
+      "learning_rate": 6.928387096774193e-05,
+      "loss": 0.1937,
+      "step": 19545
+    },
+    {
+      "epoch": 0.312736,
+      "grad_norm": 1.3828125,
+      "learning_rate": 6.928225806451613e-05,
+      "loss": 0.1844,
+      "step": 19546
+    },
+    {
+      "epoch": 0.312752,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.928064516129033e-05,
+      "loss": 0.1673,
+      "step": 19547
+    },
+    {
+      "epoch": 0.312768,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.927903225806452e-05,
+      "loss": 0.1591,
+      "step": 19548
+    },
+    {
+      "epoch": 0.312784,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.927741935483872e-05,
+      "loss": 0.2005,
+      "step": 19549
+    },
+    {
+      "epoch": 0.3128,
+      "grad_norm": 1.21875,
+      "learning_rate": 6.92758064516129e-05,
+      "loss": 0.1618,
+      "step": 19550
+    },
+    {
+      "epoch": 0.312816,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.92741935483871e-05,
+      "loss": 0.1504,
+      "step": 19551
+    },
+    {
+      "epoch": 0.312832,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.927258064516129e-05,
+      "loss": 0.1732,
+      "step": 19552
+    },
+    {
+      "epoch": 0.312848,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.927096774193549e-05,
+      "loss": 0.2161,
+      "step": 19553
+    },
+    {
+      "epoch": 0.312864,
+      "grad_norm": 0.88671875,
+      "learning_rate": 6.926935483870967e-05,
+      "loss": 0.1947,
+      "step": 19554
+    },
+    {
+      "epoch": 0.31288,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.926774193548387e-05,
+      "loss": 0.1605,
+      "step": 19555
+    },
+    {
+      "epoch": 0.312896,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.926612903225806e-05,
+      "loss": 0.1337,
+      "step": 19556
+    },
+    {
+      "epoch": 0.312912,
+      "grad_norm": 1.25,
+      "learning_rate": 6.926451612903226e-05,
+      "loss": 0.1784,
+      "step": 19557
+    },
+    {
+      "epoch": 0.312928,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.926290322580645e-05,
+      "loss": 0.1767,
+      "step": 19558
+    },
+    {
+      "epoch": 0.312944,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.926129032258065e-05,
+      "loss": 0.1535,
+      "step": 19559
+    },
+    {
+      "epoch": 0.31296,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.925967741935485e-05,
+      "loss": 0.1418,
+      "step": 19560
+    },
+    {
+      "epoch": 0.312976,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.925806451612905e-05,
+      "loss": 0.1655,
+      "step": 19561
+    },
+    {
+      "epoch": 0.312992,
+      "grad_norm": 0.90625,
+      "learning_rate": 6.925645161290323e-05,
+      "loss": 0.1673,
+      "step": 19562
+    },
+    {
+      "epoch": 0.313008,
+      "grad_norm": 0.81640625,
+      "learning_rate": 6.925483870967743e-05,
+      "loss": 0.1374,
+      "step": 19563
+    },
+    {
+      "epoch": 0.313024,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.925322580645162e-05,
+      "loss": 0.1597,
+      "step": 19564
+    },
+    {
+      "epoch": 0.31304,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.92516129032258e-05,
+      "loss": 0.201,
+      "step": 19565
+    },
+    {
+      "epoch": 0.313056,
+      "grad_norm": 0.875,
+      "learning_rate": 6.925e-05,
+      "loss": 0.1865,
+      "step": 19566
+    },
+    {
+      "epoch": 0.313072,
+      "grad_norm": 0.875,
+      "learning_rate": 6.924838709677419e-05,
+      "loss": 0.1788,
+      "step": 19567
+    },
+    {
+      "epoch": 0.313088,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.924677419354839e-05,
+      "loss": 0.1736,
+      "step": 19568
+    },
+    {
+      "epoch": 0.313104,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.924516129032257e-05,
+      "loss": 0.1645,
+      "step": 19569
+    },
+    {
+      "epoch": 0.31312,
+      "grad_norm": 0.404296875,
+      "learning_rate": 6.924354838709677e-05,
+      "loss": 0.1232,
+      "step": 19570
+    },
+    {
+      "epoch": 0.313136,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.924193548387097e-05,
+      "loss": 0.1877,
+      "step": 19571
+    },
+    {
+      "epoch": 0.313152,
+      "grad_norm": 0.83984375,
+      "learning_rate": 6.924032258064517e-05,
+      "loss": 0.1696,
+      "step": 19572
+    },
+    {
+      "epoch": 0.313168,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.923870967741936e-05,
+      "loss": 0.1598,
+      "step": 19573
+    },
+    {
+      "epoch": 0.313184,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.923709677419356e-05,
+      "loss": 0.1764,
+      "step": 19574
+    },
+    {
+      "epoch": 0.3132,
+      "grad_norm": 0.765625,
+      "learning_rate": 6.923548387096775e-05,
+      "loss": 0.1472,
+      "step": 19575
+    },
+    {
+      "epoch": 0.313216,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.923387096774194e-05,
+      "loss": 0.155,
+      "step": 19576
+    },
+    {
+      "epoch": 0.313232,
+      "grad_norm": 0.7578125,
+      "learning_rate": 6.923225806451613e-05,
+      "loss": 0.1902,
+      "step": 19577
+    },
+    {
+      "epoch": 0.313248,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.923064516129033e-05,
+      "loss": 0.1755,
+      "step": 19578
+    },
+    {
+      "epoch": 0.313264,
+      "grad_norm": 0.91796875,
+      "learning_rate": 6.922903225806452e-05,
+      "loss": 0.14,
+      "step": 19579
+    },
+    {
+      "epoch": 0.31328,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.922741935483872e-05,
+      "loss": 0.1684,
+      "step": 19580
+    },
+    {
+      "epoch": 0.313296,
+      "grad_norm": 0.6953125,
+      "learning_rate": 6.92258064516129e-05,
+      "loss": 0.1471,
+      "step": 19581
+    },
+    {
+      "epoch": 0.313312,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.922419354838709e-05,
+      "loss": 0.1572,
+      "step": 19582
+    },
+    {
+      "epoch": 0.313328,
+      "grad_norm": 0.703125,
+      "learning_rate": 6.922258064516129e-05,
+      "loss": 0.1436,
+      "step": 19583
+    },
+    {
+      "epoch": 0.313344,
+      "grad_norm": 0.75390625,
+      "learning_rate": 6.922096774193549e-05,
+      "loss": 0.1765,
+      "step": 19584
+    },
+    {
+      "epoch": 0.31336,
+      "grad_norm": 0.9453125,
+      "learning_rate": 6.921935483870969e-05,
+      "loss": 0.1632,
+      "step": 19585
+    },
+    {
+      "epoch": 0.313376,
+      "grad_norm": 0.875,
+      "learning_rate": 6.921774193548387e-05,
+      "loss": 0.1596,
+      "step": 19586
+    },
+    {
+      "epoch": 0.313392,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.921612903225807e-05,
+      "loss": 0.1755,
+      "step": 19587
+    },
+    {
+      "epoch": 0.313408,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.921451612903226e-05,
+      "loss": 0.181,
+      "step": 19588
+    },
+    {
+      "epoch": 0.313424,
+      "grad_norm": 0.98828125,
+      "learning_rate": 6.921290322580646e-05,
+      "loss": 0.1577,
+      "step": 19589
+    },
+    {
+      "epoch": 0.31344,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.921129032258064e-05,
+      "loss": 0.1855,
+      "step": 19590
+    },
+    {
+      "epoch": 0.313456,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.920967741935484e-05,
+      "loss": 0.193,
+      "step": 19591
+    },
+    {
+      "epoch": 0.313472,
+      "grad_norm": 0.9453125,
+      "learning_rate": 6.920806451612903e-05,
+      "loss": 0.167,
+      "step": 19592
+    },
+    {
+      "epoch": 0.313488,
+      "grad_norm": 1.1328125,
+      "learning_rate": 6.920645161290323e-05,
+      "loss": 0.2335,
+      "step": 19593
+    },
+    {
+      "epoch": 0.313504,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.920483870967742e-05,
+      "loss": 0.1689,
+      "step": 19594
+    },
+    {
+      "epoch": 0.31352,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.920322580645162e-05,
+      "loss": 0.1962,
+      "step": 19595
+    },
+    {
+      "epoch": 0.313536,
+      "grad_norm": 0.78125,
+      "learning_rate": 6.920161290322582e-05,
+      "loss": 0.1275,
+      "step": 19596
+    },
+    {
+      "epoch": 0.313552,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.92e-05,
+      "loss": 0.1394,
+      "step": 19597
+    },
+    {
+      "epoch": 0.313568,
+      "grad_norm": 0.828125,
+      "learning_rate": 6.91983870967742e-05,
+      "loss": 0.1864,
+      "step": 19598
+    },
+    {
+      "epoch": 0.313584,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.919677419354839e-05,
+      "loss": 0.1744,
+      "step": 19599
+    },
+    {
+      "epoch": 0.3136,
+      "grad_norm": 1.125,
+      "learning_rate": 6.919516129032259e-05,
+      "loss": 0.2461,
+      "step": 19600
+    },
+    {
+      "epoch": 0.313616,
+      "grad_norm": 0.67578125,
+      "learning_rate": 6.919354838709677e-05,
+      "loss": 0.1689,
+      "step": 19601
+    },
+    {
+      "epoch": 0.313632,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.919193548387097e-05,
+      "loss": 0.1739,
+      "step": 19602
+    },
+    {
+      "epoch": 0.313648,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.919032258064516e-05,
+      "loss": 0.1738,
+      "step": 19603
+    },
+    {
+      "epoch": 0.313664,
+      "grad_norm": 0.7890625,
+      "learning_rate": 6.918870967741936e-05,
+      "loss": 0.1666,
+      "step": 19604
+    },
+    {
+      "epoch": 0.31368,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.918709677419354e-05,
+      "loss": 0.1333,
+      "step": 19605
+    },
+    {
+      "epoch": 0.313696,
+      "grad_norm": 0.81640625,
+      "learning_rate": 6.918548387096774e-05,
+      "loss": 0.1964,
+      "step": 19606
+    },
+    {
+      "epoch": 0.313712,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.918387096774194e-05,
+      "loss": 0.1451,
+      "step": 19607
+    },
+    {
+      "epoch": 0.313728,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.918225806451614e-05,
+      "loss": 0.1695,
+      "step": 19608
+    },
+    {
+      "epoch": 0.313744,
+      "grad_norm": 0.7421875,
+      "learning_rate": 6.918064516129033e-05,
+      "loss": 0.1679,
+      "step": 19609
+    },
+    {
+      "epoch": 0.31376,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.917903225806453e-05,
+      "loss": 0.1568,
+      "step": 19610
+    },
+    {
+      "epoch": 0.313776,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.917741935483872e-05,
+      "loss": 0.1796,
+      "step": 19611
+    },
+    {
+      "epoch": 0.313792,
+      "grad_norm": 0.9921875,
+      "learning_rate": 6.91758064516129e-05,
+      "loss": 0.1961,
+      "step": 19612
+    },
+    {
+      "epoch": 0.313808,
+      "grad_norm": 0.9921875,
+      "learning_rate": 6.91741935483871e-05,
+      "loss": 0.1702,
+      "step": 19613
+    },
+    {
+      "epoch": 0.313824,
+      "grad_norm": 0.625,
+      "learning_rate": 6.917258064516129e-05,
+      "loss": 0.1619,
+      "step": 19614
+    },
+    {
+      "epoch": 0.31384,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.917096774193549e-05,
+      "loss": 0.2073,
+      "step": 19615
+    },
+    {
+      "epoch": 0.313856,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.916935483870967e-05,
+      "loss": 0.1776,
+      "step": 19616
+    },
+    {
+      "epoch": 0.313872,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.916774193548387e-05,
+      "loss": 0.1385,
+      "step": 19617
+    },
+    {
+      "epoch": 0.313888,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.916612903225806e-05,
+      "loss": 0.1684,
+      "step": 19618
+    },
+    {
+      "epoch": 0.313904,
+      "grad_norm": 0.72265625,
+      "learning_rate": 6.916451612903226e-05,
+      "loss": 0.1746,
+      "step": 19619
+    },
+    {
+      "epoch": 0.31392,
+      "grad_norm": 1.3046875,
+      "learning_rate": 6.916290322580646e-05,
+      "loss": 0.1665,
+      "step": 19620
+    },
+    {
+      "epoch": 0.313936,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.916129032258066e-05,
+      "loss": 0.138,
+      "step": 19621
+    },
+    {
+      "epoch": 0.313952,
+      "grad_norm": 0.87109375,
+      "learning_rate": 6.915967741935484e-05,
+      "loss": 0.2012,
+      "step": 19622
+    },
+    {
+      "epoch": 0.313968,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.915806451612904e-05,
+      "loss": 0.1705,
+      "step": 19623
+    },
+    {
+      "epoch": 0.313984,
+      "grad_norm": 0.75,
+      "learning_rate": 6.915645161290323e-05,
+      "loss": 0.1595,
+      "step": 19624
+    },
+    {
+      "epoch": 0.314,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.915483870967743e-05,
+      "loss": 0.1437,
+      "step": 19625
+    },
+    {
+      "epoch": 0.314016,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.915322580645161e-05,
+      "loss": 0.1616,
+      "step": 19626
+    },
+    {
+      "epoch": 0.314032,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.915161290322581e-05,
+      "loss": 0.1888,
+      "step": 19627
+    },
+    {
+      "epoch": 0.314048,
+      "grad_norm": 0.8359375,
+      "learning_rate": 6.915e-05,
+      "loss": 0.1998,
+      "step": 19628
+    },
+    {
+      "epoch": 0.314064,
+      "grad_norm": 0.80078125,
+      "learning_rate": 6.914838709677419e-05,
+      "loss": 0.2305,
+      "step": 19629
+    },
+    {
+      "epoch": 0.31408,
+      "grad_norm": 0.81640625,
+      "learning_rate": 6.914677419354839e-05,
+      "loss": 0.1742,
+      "step": 19630
+    },
+    {
+      "epoch": 0.314096,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.914516129032259e-05,
+      "loss": 0.1995,
+      "step": 19631
+    },
+    {
+      "epoch": 0.314112,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.914354838709679e-05,
+      "loss": 0.1427,
+      "step": 19632
+    },
+    {
+      "epoch": 0.314128,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.914193548387097e-05,
+      "loss": 0.2177,
+      "step": 19633
+    },
+    {
+      "epoch": 0.314144,
+      "grad_norm": 0.890625,
+      "learning_rate": 6.914032258064517e-05,
+      "loss": 0.1535,
+      "step": 19634
+    },
+    {
+      "epoch": 0.31416,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.913870967741936e-05,
+      "loss": 0.1632,
+      "step": 19635
+    },
+    {
+      "epoch": 0.314176,
+      "grad_norm": 0.7734375,
+      "learning_rate": 6.913709677419356e-05,
+      "loss": 0.1648,
+      "step": 19636
+    },
+    {
+      "epoch": 0.314192,
+      "grad_norm": 0.890625,
+      "learning_rate": 6.913548387096774e-05,
+      "loss": 0.2187,
+      "step": 19637
+    },
+    {
+      "epoch": 0.314208,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.913387096774194e-05,
+      "loss": 0.1922,
+      "step": 19638
+    },
+    {
+      "epoch": 0.314224,
+      "grad_norm": 0.87109375,
+      "learning_rate": 6.913225806451613e-05,
+      "loss": 0.1792,
+      "step": 19639
+    },
+    {
+      "epoch": 0.31424,
+      "grad_norm": 0.72265625,
+      "learning_rate": 6.913064516129033e-05,
+      "loss": 0.152,
+      "step": 19640
+    },
+    {
+      "epoch": 0.314256,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.912903225806451e-05,
+      "loss": 0.1608,
+      "step": 19641
+    },
+    {
+      "epoch": 0.314272,
+      "grad_norm": 0.76953125,
+      "learning_rate": 6.912741935483871e-05,
+      "loss": 0.1799,
+      "step": 19642
+    },
+    {
+      "epoch": 0.314288,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.912580645161291e-05,
+      "loss": 0.1785,
+      "step": 19643
+    },
+    {
+      "epoch": 0.314304,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.91241935483871e-05,
+      "loss": 0.1339,
+      "step": 19644
+    },
+    {
+      "epoch": 0.31432,
+      "grad_norm": 0.69921875,
+      "learning_rate": 6.91225806451613e-05,
+      "loss": 0.1701,
+      "step": 19645
+    },
+    {
+      "epoch": 0.314336,
+      "grad_norm": 0.94921875,
+      "learning_rate": 6.912096774193549e-05,
+      "loss": 0.2,
+      "step": 19646
+    },
+    {
+      "epoch": 0.314352,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.911935483870968e-05,
+      "loss": 0.1993,
+      "step": 19647
+    },
+    {
+      "epoch": 0.314368,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.911774193548387e-05,
+      "loss": 0.1405,
+      "step": 19648
+    },
+    {
+      "epoch": 0.314384,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.911612903225807e-05,
+      "loss": 0.201,
+      "step": 19649
+    },
+    {
+      "epoch": 0.3144,
+      "grad_norm": 0.67578125,
+      "learning_rate": 6.911451612903226e-05,
+      "loss": 0.1679,
+      "step": 19650
+    },
+    {
+      "epoch": 0.314416,
+      "grad_norm": 0.8984375,
+      "learning_rate": 6.911290322580646e-05,
+      "loss": 0.1816,
+      "step": 19651
+    },
+    {
+      "epoch": 0.314432,
+      "grad_norm": 1.140625,
+      "learning_rate": 6.911129032258064e-05,
+      "loss": 0.1354,
+      "step": 19652
+    },
+    {
+      "epoch": 0.314448,
+      "grad_norm": 0.70703125,
+      "learning_rate": 6.910967741935484e-05,
+      "loss": 0.1631,
+      "step": 19653
+    },
+    {
+      "epoch": 0.314464,
+      "grad_norm": 0.84375,
+      "learning_rate": 6.910806451612903e-05,
+      "loss": 0.2112,
+      "step": 19654
+    },
+    {
+      "epoch": 0.31448,
+      "grad_norm": 0.50390625,
+      "learning_rate": 6.910645161290323e-05,
+      "loss": 0.1604,
+      "step": 19655
+    },
+    {
+      "epoch": 0.314496,
+      "grad_norm": 0.72265625,
+      "learning_rate": 6.910483870967743e-05,
+      "loss": 0.1444,
+      "step": 19656
+    },
+    {
+      "epoch": 0.314512,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.910322580645163e-05,
+      "loss": 0.1504,
+      "step": 19657
+    },
+    {
+      "epoch": 0.314528,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.910161290322581e-05,
+      "loss": 0.1328,
+      "step": 19658
+    },
+    {
+      "epoch": 0.314544,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.91e-05,
+      "loss": 0.1334,
+      "step": 19659
+    },
+    {
+      "epoch": 0.31456,
+      "grad_norm": 0.87109375,
+      "learning_rate": 6.90983870967742e-05,
+      "loss": 0.1734,
+      "step": 19660
+    },
+    {
+      "epoch": 0.314576,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.909677419354838e-05,
+      "loss": 0.1661,
+      "step": 19661
+    },
+    {
+      "epoch": 0.314592,
+      "grad_norm": 0.98828125,
+      "learning_rate": 6.909516129032258e-05,
+      "loss": 0.1759,
+      "step": 19662
+    },
+    {
+      "epoch": 0.314608,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.909354838709677e-05,
+      "loss": 0.197,
+      "step": 19663
+    },
+    {
+      "epoch": 0.314624,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.909193548387097e-05,
+      "loss": 0.1722,
+      "step": 19664
+    },
+    {
+      "epoch": 0.31464,
+      "grad_norm": 0.8046875,
+      "learning_rate": 6.909032258064516e-05,
+      "loss": 0.1554,
+      "step": 19665
+    },
+    {
+      "epoch": 0.314656,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.908870967741936e-05,
+      "loss": 0.1609,
+      "step": 19666
+    },
+    {
+      "epoch": 0.314672,
+      "grad_norm": 1.5390625,
+      "learning_rate": 6.908709677419356e-05,
+      "loss": 0.2141,
+      "step": 19667
+    },
+    {
+      "epoch": 0.314688,
+      "grad_norm": 0.7421875,
+      "learning_rate": 6.908548387096776e-05,
+      "loss": 0.1707,
+      "step": 19668
+    },
+    {
+      "epoch": 0.314704,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.908387096774194e-05,
+      "loss": 0.1693,
+      "step": 19669
+    },
+    {
+      "epoch": 0.31472,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.908225806451614e-05,
+      "loss": 0.1523,
+      "step": 19670
+    },
+    {
+      "epoch": 0.314736,
+      "grad_norm": 0.83984375,
+      "learning_rate": 6.908064516129033e-05,
+      "loss": 0.2278,
+      "step": 19671
+    },
+    {
+      "epoch": 0.314752,
+      "grad_norm": 0.8046875,
+      "learning_rate": 6.907903225806453e-05,
+      "loss": 0.1641,
+      "step": 19672
+    },
+    {
+      "epoch": 0.314768,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.907741935483871e-05,
+      "loss": 0.183,
+      "step": 19673
+    },
+    {
+      "epoch": 0.314784,
+      "grad_norm": 1.203125,
+      "learning_rate": 6.90758064516129e-05,
+      "loss": 0.1953,
+      "step": 19674
+    },
+    {
+      "epoch": 0.3148,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.90741935483871e-05,
+      "loss": 0.1476,
+      "step": 19675
+    },
+    {
+      "epoch": 0.314816,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.907258064516128e-05,
+      "loss": 0.1437,
+      "step": 19676
+    },
+    {
+      "epoch": 0.314832,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.907096774193548e-05,
+      "loss": 0.1575,
+      "step": 19677
+    },
+    {
+      "epoch": 0.314848,
+      "grad_norm": 1.21875,
+      "learning_rate": 6.906935483870967e-05,
+      "loss": 0.197,
+      "step": 19678
+    },
+    {
+      "epoch": 0.314864,
+      "grad_norm": 0.94921875,
+      "learning_rate": 6.906774193548387e-05,
+      "loss": 0.1772,
+      "step": 19679
+    },
+    {
+      "epoch": 0.31488,
+      "grad_norm": 0.8125,
+      "learning_rate": 6.906612903225807e-05,
+      "loss": 0.1733,
+      "step": 19680
+    },
+    {
+      "epoch": 0.314896,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.906451612903227e-05,
+      "loss": 0.1806,
+      "step": 19681
+    },
+    {
+      "epoch": 0.314912,
+      "grad_norm": 0.8125,
+      "learning_rate": 6.906290322580646e-05,
+      "loss": 0.2068,
+      "step": 19682
+    },
+    {
+      "epoch": 0.314928,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.906129032258065e-05,
+      "loss": 0.1515,
+      "step": 19683
+    },
+    {
+      "epoch": 0.314944,
+      "grad_norm": 0.9921875,
+      "learning_rate": 6.905967741935484e-05,
+      "loss": 0.2088,
+      "step": 19684
+    },
+    {
+      "epoch": 0.31496,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.905806451612904e-05,
+      "loss": 0.122,
+      "step": 19685
+    },
+    {
+      "epoch": 0.314976,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.905645161290323e-05,
+      "loss": 0.1862,
+      "step": 19686
+    },
+    {
+      "epoch": 0.314992,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.905483870967743e-05,
+      "loss": 0.1752,
+      "step": 19687
+    },
+    {
+      "epoch": 0.315008,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.905322580645161e-05,
+      "loss": 0.1541,
+      "step": 19688
+    },
+    {
+      "epoch": 0.315024,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.905161290322581e-05,
+      "loss": 0.1624,
+      "step": 19689
+    },
+    {
+      "epoch": 0.31504,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.905e-05,
+      "loss": 0.1912,
+      "step": 19690
+    },
+    {
+      "epoch": 0.315056,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.90483870967742e-05,
+      "loss": 0.2061,
+      "step": 19691
+    },
+    {
+      "epoch": 0.315072,
+      "grad_norm": 0.83203125,
+      "learning_rate": 6.90467741935484e-05,
+      "loss": 0.18,
+      "step": 19692
+    },
+    {
+      "epoch": 0.315088,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.904516129032258e-05,
+      "loss": 0.1711,
+      "step": 19693
+    },
+    {
+      "epoch": 0.315104,
+      "grad_norm": 0.87109375,
+      "learning_rate": 6.904354838709678e-05,
+      "loss": 0.1651,
+      "step": 19694
+    },
+    {
+      "epoch": 0.31512,
+      "grad_norm": 0.75390625,
+      "learning_rate": 6.904193548387097e-05,
+      "loss": 0.1917,
+      "step": 19695
+    },
+    {
+      "epoch": 0.315136,
+      "grad_norm": 0.625,
+      "learning_rate": 6.904032258064517e-05,
+      "loss": 0.1562,
+      "step": 19696
+    },
+    {
+      "epoch": 0.315152,
+      "grad_norm": 0.765625,
+      "learning_rate": 6.903870967741935e-05,
+      "loss": 0.1437,
+      "step": 19697
+    },
+    {
+      "epoch": 0.315168,
+      "grad_norm": 0.76953125,
+      "learning_rate": 6.903709677419355e-05,
+      "loss": 0.1241,
+      "step": 19698
+    },
+    {
+      "epoch": 0.315184,
+      "grad_norm": 1.2421875,
+      "learning_rate": 6.903548387096774e-05,
+      "loss": 0.1749,
+      "step": 19699
+    },
+    {
+      "epoch": 0.3152,
+      "grad_norm": 0.9453125,
+      "learning_rate": 6.903387096774194e-05,
+      "loss": 0.1396,
+      "step": 19700
+    },
+    {
+      "epoch": 0.315216,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.903225806451613e-05,
+      "loss": 0.1314,
+      "step": 19701
+    },
+    {
+      "epoch": 0.315232,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.903064516129033e-05,
+      "loss": 0.1478,
+      "step": 19702
+    },
+    {
+      "epoch": 0.315248,
+      "grad_norm": 1.3046875,
+      "learning_rate": 6.902903225806453e-05,
+      "loss": 0.1804,
+      "step": 19703
+    },
+    {
+      "epoch": 0.315264,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.902741935483872e-05,
+      "loss": 0.2158,
+      "step": 19704
+    },
+    {
+      "epoch": 0.31528,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.902580645161291e-05,
+      "loss": 0.2004,
+      "step": 19705
+    },
+    {
+      "epoch": 0.315296,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.90241935483871e-05,
+      "loss": 0.1477,
+      "step": 19706
+    },
+    {
+      "epoch": 0.315312,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.90225806451613e-05,
+      "loss": 0.151,
+      "step": 19707
+    },
+    {
+      "epoch": 0.315328,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.902096774193548e-05,
+      "loss": 0.1898,
+      "step": 19708
+    },
+    {
+      "epoch": 0.315344,
+      "grad_norm": 0.89453125,
+      "learning_rate": 6.901935483870968e-05,
+      "loss": 0.1429,
+      "step": 19709
+    },
+    {
+      "epoch": 0.31536,
+      "grad_norm": 0.96484375,
+      "learning_rate": 6.901774193548387e-05,
+      "loss": 0.1668,
+      "step": 19710
+    },
+    {
+      "epoch": 0.315376,
+      "grad_norm": 0.87109375,
+      "learning_rate": 6.901612903225807e-05,
+      "loss": 0.1792,
+      "step": 19711
+    },
+    {
+      "epoch": 0.315392,
+      "grad_norm": 0.69921875,
+      "learning_rate": 6.901451612903225e-05,
+      "loss": 0.1592,
+      "step": 19712
+    },
+    {
+      "epoch": 0.315408,
+      "grad_norm": 0.796875,
+      "learning_rate": 6.901290322580645e-05,
+      "loss": 0.1799,
+      "step": 19713
+    },
+    {
+      "epoch": 0.315424,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.901129032258064e-05,
+      "loss": 0.1667,
+      "step": 19714
+    },
+    {
+      "epoch": 0.31544,
+      "grad_norm": 0.84765625,
+      "learning_rate": 6.900967741935484e-05,
+      "loss": 0.1491,
+      "step": 19715
+    },
+    {
+      "epoch": 0.315456,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.900806451612904e-05,
+      "loss": 0.1612,
+      "step": 19716
+    },
+    {
+      "epoch": 0.315472,
+      "grad_norm": 0.84375,
+      "learning_rate": 6.900645161290324e-05,
+      "loss": 0.1774,
+      "step": 19717
+    },
+    {
+      "epoch": 0.315488,
+      "grad_norm": 0.921875,
+      "learning_rate": 6.900483870967742e-05,
+      "loss": 0.2067,
+      "step": 19718
+    },
+    {
+      "epoch": 0.315504,
+      "grad_norm": 0.625,
+      "learning_rate": 6.900322580645162e-05,
+      "loss": 0.1659,
+      "step": 19719
+    },
+    {
+      "epoch": 0.31552,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.900161290322581e-05,
+      "loss": 0.1554,
+      "step": 19720
+    },
+    {
+      "epoch": 0.315536,
+      "grad_norm": 0.93359375,
+      "learning_rate": 6.9e-05,
+      "loss": 0.2159,
+      "step": 19721
+    },
+    {
+      "epoch": 0.315552,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.89983870967742e-05,
+      "loss": 0.2013,
+      "step": 19722
+    },
+    {
+      "epoch": 0.315568,
+      "grad_norm": 1.0703125,
+      "learning_rate": 6.899677419354838e-05,
+      "loss": 0.1321,
+      "step": 19723
+    },
+    {
+      "epoch": 0.315584,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.899516129032258e-05,
+      "loss": 0.1638,
+      "step": 19724
+    },
+    {
+      "epoch": 0.3156,
+      "grad_norm": 0.765625,
+      "learning_rate": 6.899354838709677e-05,
+      "loss": 0.1402,
+      "step": 19725
+    },
+    {
+      "epoch": 0.315616,
+      "grad_norm": 0.86328125,
+      "learning_rate": 6.899193548387097e-05,
+      "loss": 0.156,
+      "step": 19726
+    },
+    {
+      "epoch": 0.315632,
+      "grad_norm": 0.7890625,
+      "learning_rate": 6.899032258064517e-05,
+      "loss": 0.1965,
+      "step": 19727
+    },
+    {
+      "epoch": 0.315648,
+      "grad_norm": 0.90234375,
+      "learning_rate": 6.898870967741937e-05,
+      "loss": 0.2053,
+      "step": 19728
+    },
+    {
+      "epoch": 0.315664,
+      "grad_norm": 1.1328125,
+      "learning_rate": 6.898709677419355e-05,
+      "loss": 0.175,
+      "step": 19729
+    },
+    {
+      "epoch": 0.31568,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.898548387096775e-05,
+      "loss": 0.1753,
+      "step": 19730
+    },
+    {
+      "epoch": 0.315696,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.898387096774194e-05,
+      "loss": 0.1495,
+      "step": 19731
+    },
+    {
+      "epoch": 0.315712,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.898225806451614e-05,
+      "loss": 0.209,
+      "step": 19732
+    },
+    {
+      "epoch": 0.315728,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.898064516129032e-05,
+      "loss": 0.1612,
+      "step": 19733
+    },
+    {
+      "epoch": 0.315744,
+      "grad_norm": 0.69921875,
+      "learning_rate": 6.897903225806452e-05,
+      "loss": 0.1283,
+      "step": 19734
+    },
+    {
+      "epoch": 0.31576,
+      "grad_norm": 0.86328125,
+      "learning_rate": 6.897741935483871e-05,
+      "loss": 0.1644,
+      "step": 19735
+    },
+    {
+      "epoch": 0.315776,
+      "grad_norm": 0.828125,
+      "learning_rate": 6.89758064516129e-05,
+      "loss": 0.1394,
+      "step": 19736
+    },
+    {
+      "epoch": 0.315792,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.89741935483871e-05,
+      "loss": 0.1251,
+      "step": 19737
+    },
+    {
+      "epoch": 0.315808,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.89725806451613e-05,
+      "loss": 0.1789,
+      "step": 19738
+    },
+    {
+      "epoch": 0.315824,
+      "grad_norm": 0.72265625,
+      "learning_rate": 6.897096774193548e-05,
+      "loss": 0.1641,
+      "step": 19739
+    },
+    {
+      "epoch": 0.31584,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.896935483870968e-05,
+      "loss": 0.1712,
+      "step": 19740
+    },
+    {
+      "epoch": 0.315856,
+      "grad_norm": 0.79296875,
+      "learning_rate": 6.896774193548388e-05,
+      "loss": 0.1467,
+      "step": 19741
+    },
+    {
+      "epoch": 0.315872,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.896612903225807e-05,
+      "loss": 0.1232,
+      "step": 19742
+    },
+    {
+      "epoch": 0.315888,
+      "grad_norm": 0.76953125,
+      "learning_rate": 6.896451612903227e-05,
+      "loss": 0.1536,
+      "step": 19743
+    },
+    {
+      "epoch": 0.315904,
+      "grad_norm": 0.8515625,
+      "learning_rate": 6.896290322580645e-05,
+      "loss": 0.157,
+      "step": 19744
+    },
+    {
+      "epoch": 0.31592,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.896129032258065e-05,
+      "loss": 0.1677,
+      "step": 19745
+    },
+    {
+      "epoch": 0.315936,
+      "grad_norm": 0.921875,
+      "learning_rate": 6.895967741935484e-05,
+      "loss": 0.1687,
+      "step": 19746
+    },
+    {
+      "epoch": 0.315952,
+      "grad_norm": 0.7578125,
+      "learning_rate": 6.895806451612904e-05,
+      "loss": 0.149,
+      "step": 19747
+    },
+    {
+      "epoch": 0.315968,
+      "grad_norm": 0.69921875,
+      "learning_rate": 6.895645161290322e-05,
+      "loss": 0.1175,
+      "step": 19748
+    },
+    {
+      "epoch": 0.315984,
+      "grad_norm": 0.78515625,
+      "learning_rate": 6.895483870967742e-05,
+      "loss": 0.1638,
+      "step": 19749
+    },
+    {
+      "epoch": 0.316,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.895322580645161e-05,
+      "loss": 0.1325,
+      "step": 19750
+    },
+    {
+      "epoch": 0.316016,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.895161290322581e-05,
+      "loss": 0.1506,
+      "step": 19751
+    },
+    {
+      "epoch": 0.316032,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.895000000000001e-05,
+      "loss": 0.2144,
+      "step": 19752
+    },
+    {
+      "epoch": 0.316048,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.89483870967742e-05,
+      "loss": 0.1651,
+      "step": 19753
+    },
+    {
+      "epoch": 0.316064,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.89467741935484e-05,
+      "loss": 0.1901,
+      "step": 19754
+    },
+    {
+      "epoch": 0.31608,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.894516129032258e-05,
+      "loss": 0.2015,
+      "step": 19755
+    },
+    {
+      "epoch": 0.316096,
+      "grad_norm": 0.76171875,
+      "learning_rate": 6.894354838709678e-05,
+      "loss": 0.1662,
+      "step": 19756
+    },
+    {
+      "epoch": 0.316112,
+      "grad_norm": 1.2421875,
+      "learning_rate": 6.894193548387097e-05,
+      "loss": 0.1871,
+      "step": 19757
+    },
+    {
+      "epoch": 0.316128,
+      "grad_norm": 0.72265625,
+      "learning_rate": 6.894032258064517e-05,
+      "loss": 0.1469,
+      "step": 19758
+    },
+    {
+      "epoch": 0.316144,
+      "grad_norm": 0.875,
+      "learning_rate": 6.893870967741935e-05,
+      "loss": 0.175,
+      "step": 19759
+    },
+    {
+      "epoch": 0.31616,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.893709677419355e-05,
+      "loss": 0.1695,
+      "step": 19760
+    },
+    {
+      "epoch": 0.316176,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.893548387096774e-05,
+      "loss": 0.1937,
+      "step": 19761
+    },
+    {
+      "epoch": 0.316192,
+      "grad_norm": 0.78515625,
+      "learning_rate": 6.893387096774194e-05,
+      "loss": 0.1739,
+      "step": 19762
+    },
+    {
+      "epoch": 0.316208,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.893225806451614e-05,
+      "loss": 0.154,
+      "step": 19763
+    },
+    {
+      "epoch": 0.316224,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.893064516129034e-05,
+      "loss": 0.1554,
+      "step": 19764
+    },
+    {
+      "epoch": 0.31624,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.892903225806452e-05,
+      "loss": 0.1713,
+      "step": 19765
+    },
+    {
+      "epoch": 0.316256,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.892741935483872e-05,
+      "loss": 0.1649,
+      "step": 19766
+    },
+    {
+      "epoch": 0.316272,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.892580645161291e-05,
+      "loss": 0.1702,
+      "step": 19767
+    },
+    {
+      "epoch": 0.316288,
+      "grad_norm": 0.859375,
+      "learning_rate": 6.89241935483871e-05,
+      "loss": 0.1492,
+      "step": 19768
+    },
+    {
+      "epoch": 0.316304,
+      "grad_norm": 0.8359375,
+      "learning_rate": 6.89225806451613e-05,
+      "loss": 0.1891,
+      "step": 19769
+    },
+    {
+      "epoch": 0.31632,
+      "grad_norm": 0.8125,
+      "learning_rate": 6.892096774193548e-05,
+      "loss": 0.1762,
+      "step": 19770
+    },
+    {
+      "epoch": 0.316336,
+      "grad_norm": 0.75390625,
+      "learning_rate": 6.891935483870968e-05,
+      "loss": 0.1754,
+      "step": 19771
+    },
+    {
+      "epoch": 0.316352,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.891774193548387e-05,
+      "loss": 0.1773,
+      "step": 19772
+    },
+    {
+      "epoch": 0.316368,
+      "grad_norm": 1.234375,
+      "learning_rate": 6.891612903225807e-05,
+      "loss": 0.166,
+      "step": 19773
+    },
+    {
+      "epoch": 0.316384,
+      "grad_norm": 0.78515625,
+      "learning_rate": 6.891451612903225e-05,
+      "loss": 0.179,
+      "step": 19774
+    },
+    {
+      "epoch": 0.3164,
+      "grad_norm": 0.78515625,
+      "learning_rate": 6.891290322580645e-05,
+      "loss": 0.1639,
+      "step": 19775
+    },
+    {
+      "epoch": 0.316416,
+      "grad_norm": 0.75,
+      "learning_rate": 6.891129032258065e-05,
+      "loss": 0.1568,
+      "step": 19776
+    },
+    {
+      "epoch": 0.316432,
+      "grad_norm": 0.82421875,
+      "learning_rate": 6.890967741935485e-05,
+      "loss": 0.1729,
+      "step": 19777
+    },
+    {
+      "epoch": 0.316448,
+      "grad_norm": 0.984375,
+      "learning_rate": 6.890806451612904e-05,
+      "loss": 0.2051,
+      "step": 19778
+    },
+    {
+      "epoch": 0.316464,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.890645161290324e-05,
+      "loss": 0.1932,
+      "step": 19779
+    },
+    {
+      "epoch": 0.31648,
+      "grad_norm": 1.0703125,
+      "learning_rate": 6.890483870967742e-05,
+      "loss": 0.1795,
+      "step": 19780
+    },
+    {
+      "epoch": 0.316496,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.890322580645162e-05,
+      "loss": 0.1778,
+      "step": 19781
+    },
+    {
+      "epoch": 0.316512,
+      "grad_norm": 0.98828125,
+      "learning_rate": 6.890161290322581e-05,
+      "loss": 0.2322,
+      "step": 19782
+    },
+    {
+      "epoch": 0.316528,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.89e-05,
+      "loss": 0.1477,
+      "step": 19783
+    },
+    {
+      "epoch": 0.316544,
+      "grad_norm": 1.5,
+      "learning_rate": 6.88983870967742e-05,
+      "loss": 0.1934,
+      "step": 19784
+    },
+    {
+      "epoch": 0.31656,
+      "grad_norm": 0.8359375,
+      "learning_rate": 6.889677419354838e-05,
+      "loss": 0.2203,
+      "step": 19785
+    },
+    {
+      "epoch": 0.316576,
+      "grad_norm": 0.67578125,
+      "learning_rate": 6.889516129032258e-05,
+      "loss": 0.1921,
+      "step": 19786
+    },
+    {
+      "epoch": 0.316592,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.889354838709678e-05,
+      "loss": 0.1915,
+      "step": 19787
+    },
+    {
+      "epoch": 0.316608,
+      "grad_norm": 0.88671875,
+      "learning_rate": 6.889193548387098e-05,
+      "loss": 0.19,
+      "step": 19788
+    },
+    {
+      "epoch": 0.316624,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.889032258064516e-05,
+      "loss": 0.1919,
+      "step": 19789
+    },
+    {
+      "epoch": 0.31664,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.888870967741936e-05,
+      "loss": 0.1425,
+      "step": 19790
+    },
+    {
+      "epoch": 0.316656,
+      "grad_norm": 0.8125,
+      "learning_rate": 6.888709677419355e-05,
+      "loss": 0.167,
+      "step": 19791
+    },
+    {
+      "epoch": 0.316672,
+      "grad_norm": 0.7890625,
+      "learning_rate": 6.888548387096775e-05,
+      "loss": 0.1433,
+      "step": 19792
+    },
+    {
+      "epoch": 0.316688,
+      "grad_norm": 0.90625,
+      "learning_rate": 6.888387096774194e-05,
+      "loss": 0.1914,
+      "step": 19793
+    },
+    {
+      "epoch": 0.316704,
+      "grad_norm": 0.83984375,
+      "learning_rate": 6.888225806451614e-05,
+      "loss": 0.1719,
+      "step": 19794
+    },
+    {
+      "epoch": 0.31672,
+      "grad_norm": 0.75,
+      "learning_rate": 6.888064516129032e-05,
+      "loss": 0.151,
+      "step": 19795
+    },
+    {
+      "epoch": 0.316736,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.887903225806452e-05,
+      "loss": 0.1448,
+      "step": 19796
+    },
+    {
+      "epoch": 0.316752,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.887741935483871e-05,
+      "loss": 0.2106,
+      "step": 19797
+    },
+    {
+      "epoch": 0.316768,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.887580645161291e-05,
+      "loss": 0.1975,
+      "step": 19798
+    },
+    {
+      "epoch": 0.316784,
+      "grad_norm": 0.88671875,
+      "learning_rate": 6.887419354838711e-05,
+      "loss": 0.1428,
+      "step": 19799
+    },
+    {
+      "epoch": 0.3168,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.887258064516129e-05,
+      "loss": 0.1653,
+      "step": 19800
+    },
+    {
+      "epoch": 0.316816,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.887096774193549e-05,
+      "loss": 0.1453,
+      "step": 19801
+    },
+    {
+      "epoch": 0.316832,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.886935483870968e-05,
+      "loss": 0.1469,
+      "step": 19802
+    },
+    {
+      "epoch": 0.316848,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.886774193548388e-05,
+      "loss": 0.1668,
+      "step": 19803
+    },
+    {
+      "epoch": 0.316864,
+      "grad_norm": 0.70703125,
+      "learning_rate": 6.886612903225806e-05,
+      "loss": 0.1787,
+      "step": 19804
+    },
+    {
+      "epoch": 0.31688,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.886451612903226e-05,
+      "loss": 0.1838,
+      "step": 19805
+    },
+    {
+      "epoch": 0.316896,
+      "grad_norm": 0.51953125,
+      "learning_rate": 6.886290322580645e-05,
+      "loss": 0.1613,
+      "step": 19806
+    },
+    {
+      "epoch": 0.316912,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.886129032258065e-05,
+      "loss": 0.1578,
+      "step": 19807
+    },
+    {
+      "epoch": 0.316928,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.885967741935484e-05,
+      "loss": 0.2281,
+      "step": 19808
+    },
+    {
+      "epoch": 0.316944,
+      "grad_norm": 0.8359375,
+      "learning_rate": 6.885806451612904e-05,
+      "loss": 0.1446,
+      "step": 19809
+    },
+    {
+      "epoch": 0.31696,
+      "grad_norm": 1.5703125,
+      "learning_rate": 6.885645161290322e-05,
+      "loss": 0.1682,
+      "step": 19810
+    },
+    {
+      "epoch": 0.316976,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.885483870967742e-05,
+      "loss": 0.1601,
+      "step": 19811
+    },
+    {
+      "epoch": 0.316992,
+      "grad_norm": 0.703125,
+      "learning_rate": 6.885322580645162e-05,
+      "loss": 0.1696,
+      "step": 19812
+    },
+    {
+      "epoch": 0.317008,
+      "grad_norm": 0.828125,
+      "learning_rate": 6.885161290322582e-05,
+      "loss": 0.1687,
+      "step": 19813
+    },
+    {
+      "epoch": 0.317024,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.885e-05,
+      "loss": 0.1598,
+      "step": 19814
+    },
+    {
+      "epoch": 0.31704,
+      "grad_norm": 1.2578125,
+      "learning_rate": 6.884838709677419e-05,
+      "loss": 0.158,
+      "step": 19815
+    },
+    {
+      "epoch": 0.317056,
+      "grad_norm": 1.2421875,
+      "learning_rate": 6.884677419354839e-05,
+      "loss": 0.1538,
+      "step": 19816
+    },
+    {
+      "epoch": 0.317072,
+      "grad_norm": 0.80078125,
+      "learning_rate": 6.884516129032258e-05,
+      "loss": 0.1724,
+      "step": 19817
+    },
+    {
+      "epoch": 0.317088,
+      "grad_norm": 1.2265625,
+      "learning_rate": 6.884354838709678e-05,
+      "loss": 0.1957,
+      "step": 19818
+    },
+    {
+      "epoch": 0.317104,
+      "grad_norm": 0.875,
+      "learning_rate": 6.884193548387096e-05,
+      "loss": 0.1451,
+      "step": 19819
+    },
+    {
+      "epoch": 0.31712,
+      "grad_norm": 1.265625,
+      "learning_rate": 6.884032258064516e-05,
+      "loss": 0.1462,
+      "step": 19820
+    },
+    {
+      "epoch": 0.317136,
+      "grad_norm": 1.0859375,
+      "learning_rate": 6.883870967741935e-05,
+      "loss": 0.1673,
+      "step": 19821
+    },
+    {
+      "epoch": 0.317152,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.883709677419355e-05,
+      "loss": 0.1418,
+      "step": 19822
+    },
+    {
+      "epoch": 0.317168,
+      "grad_norm": 0.8359375,
+      "learning_rate": 6.883548387096775e-05,
+      "loss": 0.1384,
+      "step": 19823
+    },
+    {
+      "epoch": 0.317184,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.883387096774195e-05,
+      "loss": 0.1407,
+      "step": 19824
+    },
+    {
+      "epoch": 0.3172,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.883225806451613e-05,
+      "loss": 0.1668,
+      "step": 19825
+    },
+    {
+      "epoch": 0.317216,
+      "grad_norm": 0.77734375,
+      "learning_rate": 6.883064516129033e-05,
+      "loss": 0.198,
+      "step": 19826
+    },
+    {
+      "epoch": 0.317232,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.882903225806452e-05,
+      "loss": 0.1597,
+      "step": 19827
+    },
+    {
+      "epoch": 0.317248,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.882741935483872e-05,
+      "loss": 0.1556,
+      "step": 19828
+    },
+    {
+      "epoch": 0.317264,
+      "grad_norm": 0.76171875,
+      "learning_rate": 6.88258064516129e-05,
+      "loss": 0.1642,
+      "step": 19829
+    },
+    {
+      "epoch": 0.31728,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.882419354838709e-05,
+      "loss": 0.1825,
+      "step": 19830
+    },
+    {
+      "epoch": 0.317296,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.882258064516129e-05,
+      "loss": 0.1832,
+      "step": 19831
+    },
+    {
+      "epoch": 0.317312,
+      "grad_norm": 0.67578125,
+      "learning_rate": 6.882096774193548e-05,
+      "loss": 0.1521,
+      "step": 19832
+    },
+    {
+      "epoch": 0.317328,
+      "grad_norm": 0.8125,
+      "learning_rate": 6.881935483870968e-05,
+      "loss": 0.1236,
+      "step": 19833
+    },
+    {
+      "epoch": 0.317344,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.881774193548386e-05,
+      "loss": 0.1832,
+      "step": 19834
+    },
+    {
+      "epoch": 0.31736,
+      "grad_norm": 0.7890625,
+      "learning_rate": 6.881612903225806e-05,
+      "loss": 0.1813,
+      "step": 19835
+    },
+    {
+      "epoch": 0.317376,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.881451612903226e-05,
+      "loss": 0.1528,
+      "step": 19836
+    },
+    {
+      "epoch": 0.317392,
+      "grad_norm": 0.82421875,
+      "learning_rate": 6.881290322580646e-05,
+      "loss": 0.1826,
+      "step": 19837
+    },
+    {
+      "epoch": 0.317408,
+      "grad_norm": 0.85546875,
+      "learning_rate": 6.881129032258065e-05,
+      "loss": 0.2377,
+      "step": 19838
+    },
+    {
+      "epoch": 0.317424,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.880967741935485e-05,
+      "loss": 0.1499,
+      "step": 19839
+    },
+    {
+      "epoch": 0.31744,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.880806451612903e-05,
+      "loss": 0.1587,
+      "step": 19840
+    },
+    {
+      "epoch": 0.317456,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.880645161290323e-05,
+      "loss": 0.1711,
+      "step": 19841
+    },
+    {
+      "epoch": 0.317472,
+      "grad_norm": 0.625,
+      "learning_rate": 6.880483870967742e-05,
+      "loss": 0.1272,
+      "step": 19842
+    },
+    {
+      "epoch": 0.317488,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.880322580645162e-05,
+      "loss": 0.1491,
+      "step": 19843
+    },
+    {
+      "epoch": 0.317504,
+      "grad_norm": 1.171875,
+      "learning_rate": 6.88016129032258e-05,
+      "loss": 0.2352,
+      "step": 19844
+    },
+    {
+      "epoch": 0.31752,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.879999999999999e-05,
+      "loss": 0.1545,
+      "step": 19845
+    },
+    {
+      "epoch": 0.317536,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.879838709677419e-05,
+      "loss": 0.1531,
+      "step": 19846
+    },
+    {
+      "epoch": 0.317552,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.879677419354839e-05,
+      "loss": 0.1554,
+      "step": 19847
+    },
+    {
+      "epoch": 0.317568,
+      "grad_norm": 1.4453125,
+      "learning_rate": 6.879516129032259e-05,
+      "loss": 0.1872,
+      "step": 19848
+    },
+    {
+      "epoch": 0.317584,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.879354838709678e-05,
+      "loss": 0.1519,
+      "step": 19849
+    },
+    {
+      "epoch": 0.3176,
+      "grad_norm": 0.859375,
+      "learning_rate": 6.879193548387098e-05,
+      "loss": 0.1649,
+      "step": 19850
+    },
+    {
+      "epoch": 0.317616,
+      "grad_norm": 1.3515625,
+      "learning_rate": 6.879032258064516e-05,
+      "loss": 0.1754,
+      "step": 19851
+    },
+    {
+      "epoch": 0.317632,
+      "grad_norm": 0.9296875,
+      "learning_rate": 6.878870967741936e-05,
+      "loss": 0.1736,
+      "step": 19852
+    },
+    {
+      "epoch": 0.317648,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.878709677419355e-05,
+      "loss": 0.1718,
+      "step": 19853
+    },
+    {
+      "epoch": 0.317664,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.878548387096775e-05,
+      "loss": 0.136,
+      "step": 19854
+    },
+    {
+      "epoch": 0.31768,
+      "grad_norm": 0.78125,
+      "learning_rate": 6.878387096774193e-05,
+      "loss": 0.1434,
+      "step": 19855
+    },
+    {
+      "epoch": 0.317696,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.878225806451613e-05,
+      "loss": 0.2009,
+      "step": 19856
+    },
+    {
+      "epoch": 0.317712,
+      "grad_norm": 0.74609375,
+      "learning_rate": 6.878064516129032e-05,
+      "loss": 0.1727,
+      "step": 19857
+    },
+    {
+      "epoch": 0.317728,
+      "grad_norm": 1.125,
+      "learning_rate": 6.877903225806452e-05,
+      "loss": 0.1744,
+      "step": 19858
+    },
+    {
+      "epoch": 0.317744,
+      "grad_norm": 1.1796875,
+      "learning_rate": 6.877741935483872e-05,
+      "loss": 0.158,
+      "step": 19859
+    },
+    {
+      "epoch": 0.31776,
+      "grad_norm": 0.84765625,
+      "learning_rate": 6.877580645161292e-05,
+      "loss": 0.1576,
+      "step": 19860
+    },
+    {
+      "epoch": 0.317776,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.87741935483871e-05,
+      "loss": 0.2222,
+      "step": 19861
+    },
+    {
+      "epoch": 0.317792,
+      "grad_norm": 0.82421875,
+      "learning_rate": 6.877258064516129e-05,
+      "loss": 0.2555,
+      "step": 19862
+    },
+    {
+      "epoch": 0.317808,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.877096774193549e-05,
+      "loss": 0.2015,
+      "step": 19863
+    },
+    {
+      "epoch": 0.317824,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.876935483870968e-05,
+      "loss": 0.1269,
+      "step": 19864
+    },
+    {
+      "epoch": 0.31784,
+      "grad_norm": 0.921875,
+      "learning_rate": 6.876774193548388e-05,
+      "loss": 0.1582,
+      "step": 19865
+    },
+    {
+      "epoch": 0.317856,
+      "grad_norm": 0.9296875,
+      "learning_rate": 6.876612903225806e-05,
+      "loss": 0.1606,
+      "step": 19866
+    },
+    {
+      "epoch": 0.317872,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.876451612903226e-05,
+      "loss": 0.1368,
+      "step": 19867
+    },
+    {
+      "epoch": 0.317888,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.876290322580645e-05,
+      "loss": 0.2088,
+      "step": 19868
+    },
+    {
+      "epoch": 0.317904,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.876129032258065e-05,
+      "loss": 0.1368,
+      "step": 19869
+    },
+    {
+      "epoch": 0.31792,
+      "grad_norm": 0.84765625,
+      "learning_rate": 6.875967741935483e-05,
+      "loss": 0.1764,
+      "step": 19870
+    },
+    {
+      "epoch": 0.317936,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.875806451612903e-05,
+      "loss": 0.1838,
+      "step": 19871
+    },
+    {
+      "epoch": 0.317952,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.875645161290323e-05,
+      "loss": 0.1643,
+      "step": 19872
+    },
+    {
+      "epoch": 0.317968,
+      "grad_norm": 0.79296875,
+      "learning_rate": 6.875483870967743e-05,
+      "loss": 0.1423,
+      "step": 19873
+    },
+    {
+      "epoch": 0.317984,
+      "grad_norm": 0.99609375,
+      "learning_rate": 6.875322580645162e-05,
+      "loss": 0.1556,
+      "step": 19874
+    },
+    {
+      "epoch": 0.318,
+      "grad_norm": 0.625,
+      "learning_rate": 6.875161290322582e-05,
+      "loss": 0.1704,
+      "step": 19875
+    },
+    {
+      "epoch": 0.318016,
+      "grad_norm": 0.7421875,
+      "learning_rate": 6.875e-05,
+      "loss": 0.1675,
+      "step": 19876
+    },
+    {
+      "epoch": 0.318032,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.874838709677419e-05,
+      "loss": 0.1449,
+      "step": 19877
+    },
+    {
+      "epoch": 0.318048,
+      "grad_norm": 1.3203125,
+      "learning_rate": 6.874677419354839e-05,
+      "loss": 0.1814,
+      "step": 19878
+    },
+    {
+      "epoch": 0.318064,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.874516129032258e-05,
+      "loss": 0.2046,
+      "step": 19879
+    },
+    {
+      "epoch": 0.31808,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.874354838709678e-05,
+      "loss": 0.1702,
+      "step": 19880
+    },
+    {
+      "epoch": 0.318096,
+      "grad_norm": 0.80859375,
+      "learning_rate": 6.874193548387096e-05,
+      "loss": 0.1613,
+      "step": 19881
+    },
+    {
+      "epoch": 0.318112,
+      "grad_norm": 0.9296875,
+      "learning_rate": 6.874032258064516e-05,
+      "loss": 0.2379,
+      "step": 19882
+    },
+    {
+      "epoch": 0.318128,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.873870967741936e-05,
+      "loss": 0.1544,
+      "step": 19883
+    },
+    {
+      "epoch": 0.318144,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.873709677419356e-05,
+      "loss": 0.1675,
+      "step": 19884
+    },
+    {
+      "epoch": 0.31816,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.873548387096775e-05,
+      "loss": 0.1366,
+      "step": 19885
+    },
+    {
+      "epoch": 0.318176,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.873387096774195e-05,
+      "loss": 0.1302,
+      "step": 19886
+    },
+    {
+      "epoch": 0.318192,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.873225806451613e-05,
+      "loss": 0.1652,
+      "step": 19887
+    },
+    {
+      "epoch": 0.318208,
+      "grad_norm": 0.7734375,
+      "learning_rate": 6.873064516129033e-05,
+      "loss": 0.1896,
+      "step": 19888
+    },
+    {
+      "epoch": 0.318224,
+      "grad_norm": 0.75390625,
+      "learning_rate": 6.872903225806452e-05,
+      "loss": 0.1709,
+      "step": 19889
+    },
+    {
+      "epoch": 0.31824,
+      "grad_norm": 0.8046875,
+      "learning_rate": 6.872741935483872e-05,
+      "loss": 0.2298,
+      "step": 19890
+    },
+    {
+      "epoch": 0.318256,
+      "grad_norm": 0.67578125,
+      "learning_rate": 6.87258064516129e-05,
+      "loss": 0.1795,
+      "step": 19891
+    },
+    {
+      "epoch": 0.318272,
+      "grad_norm": 0.828125,
+      "learning_rate": 6.872419354838709e-05,
+      "loss": 0.1776,
+      "step": 19892
+    },
+    {
+      "epoch": 0.318288,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.872258064516129e-05,
+      "loss": 0.191,
+      "step": 19893
+    },
+    {
+      "epoch": 0.318304,
+      "grad_norm": 0.625,
+      "learning_rate": 6.872096774193549e-05,
+      "loss": 0.1535,
+      "step": 19894
+    },
+    {
+      "epoch": 0.31832,
+      "grad_norm": 1.140625,
+      "learning_rate": 6.871935483870969e-05,
+      "loss": 0.187,
+      "step": 19895
+    },
+    {
+      "epoch": 0.318336,
+      "grad_norm": 1.1640625,
+      "learning_rate": 6.871774193548387e-05,
+      "loss": 0.2129,
+      "step": 19896
+    },
+    {
+      "epoch": 0.318352,
+      "grad_norm": 0.7734375,
+      "learning_rate": 6.871612903225807e-05,
+      "loss": 0.1642,
+      "step": 19897
+    },
+    {
+      "epoch": 0.318368,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.871451612903226e-05,
+      "loss": 0.1823,
+      "step": 19898
+    },
+    {
+      "epoch": 0.318384,
+      "grad_norm": 0.87890625,
+      "learning_rate": 6.871290322580646e-05,
+      "loss": 0.2071,
+      "step": 19899
+    },
+    {
+      "epoch": 0.3184,
+      "grad_norm": 0.765625,
+      "learning_rate": 6.871129032258065e-05,
+      "loss": 0.1935,
+      "step": 19900
+    },
+    {
+      "epoch": 0.318416,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.870967741935485e-05,
+      "loss": 0.1812,
+      "step": 19901
+    },
+    {
+      "epoch": 0.318432,
+      "grad_norm": 0.7421875,
+      "learning_rate": 6.870806451612903e-05,
+      "loss": 0.1987,
+      "step": 19902
+    },
+    {
+      "epoch": 0.318448,
+      "grad_norm": 0.890625,
+      "learning_rate": 6.870645161290323e-05,
+      "loss": 0.1637,
+      "step": 19903
+    },
+    {
+      "epoch": 0.318464,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.870483870967742e-05,
+      "loss": 0.1708,
+      "step": 19904
+    },
+    {
+      "epoch": 0.31848,
+      "grad_norm": 0.89453125,
+      "learning_rate": 6.870322580645162e-05,
+      "loss": 0.1584,
+      "step": 19905
+    },
+    {
+      "epoch": 0.318496,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.87016129032258e-05,
+      "loss": 0.1629,
+      "step": 19906
+    },
+    {
+      "epoch": 0.318512,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.87e-05,
+      "loss": 0.17,
+      "step": 19907
+    },
+    {
+      "epoch": 0.318528,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.86983870967742e-05,
+      "loss": 0.1951,
+      "step": 19908
+    },
+    {
+      "epoch": 0.318544,
+      "grad_norm": 1.109375,
+      "learning_rate": 6.869677419354839e-05,
+      "loss": 0.1815,
+      "step": 19909
+    },
+    {
+      "epoch": 0.31856,
+      "grad_norm": 1.3125,
+      "learning_rate": 6.869516129032259e-05,
+      "loss": 0.1421,
+      "step": 19910
+    },
+    {
+      "epoch": 0.318576,
+      "grad_norm": 1.109375,
+      "learning_rate": 6.869354838709677e-05,
+      "loss": 0.2025,
+      "step": 19911
+    },
+    {
+      "epoch": 0.318592,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.869193548387097e-05,
+      "loss": 0.2062,
+      "step": 19912
+    },
+    {
+      "epoch": 0.318608,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.869032258064516e-05,
+      "loss": 0.1177,
+      "step": 19913
+    },
+    {
+      "epoch": 0.318624,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.868870967741936e-05,
+      "loss": 0.1683,
+      "step": 19914
+    },
+    {
+      "epoch": 0.31864,
+      "grad_norm": 0.859375,
+      "learning_rate": 6.868709677419355e-05,
+      "loss": 0.1764,
+      "step": 19915
+    },
+    {
+      "epoch": 0.318656,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.868548387096775e-05,
+      "loss": 0.1449,
+      "step": 19916
+    },
+    {
+      "epoch": 0.318672,
+      "grad_norm": 0.8203125,
+      "learning_rate": 6.868387096774193e-05,
+      "loss": 0.1968,
+      "step": 19917
+    },
+    {
+      "epoch": 0.318688,
+      "grad_norm": 0.82421875,
+      "learning_rate": 6.868225806451613e-05,
+      "loss": 0.1712,
+      "step": 19918
+    },
+    {
+      "epoch": 0.318704,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.868064516129033e-05,
+      "loss": 0.1757,
+      "step": 19919
+    },
+    {
+      "epoch": 0.31872,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.867903225806453e-05,
+      "loss": 0.19,
+      "step": 19920
+    },
+    {
+      "epoch": 0.318736,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.867741935483872e-05,
+      "loss": 0.1682,
+      "step": 19921
+    },
+    {
+      "epoch": 0.318752,
+      "grad_norm": 0.90234375,
+      "learning_rate": 6.867580645161292e-05,
+      "loss": 0.1485,
+      "step": 19922
+    },
+    {
+      "epoch": 0.318768,
+      "grad_norm": 0.953125,
+      "learning_rate": 6.86741935483871e-05,
+      "loss": 0.1609,
+      "step": 19923
+    },
+    {
+      "epoch": 0.318784,
+      "grad_norm": 0.74609375,
+      "learning_rate": 6.867258064516129e-05,
+      "loss": 0.2022,
+      "step": 19924
+    },
+    {
+      "epoch": 0.3188,
+      "grad_norm": 1.3515625,
+      "learning_rate": 6.867096774193549e-05,
+      "loss": 0.1826,
+      "step": 19925
+    },
+    {
+      "epoch": 0.318816,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.866935483870967e-05,
+      "loss": 0.1691,
+      "step": 19926
+    },
+    {
+      "epoch": 0.318832,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.866774193548387e-05,
+      "loss": 0.2292,
+      "step": 19927
+    },
+    {
+      "epoch": 0.318848,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.866612903225806e-05,
+      "loss": 0.1619,
+      "step": 19928
+    },
+    {
+      "epoch": 0.318864,
+      "grad_norm": 0.765625,
+      "learning_rate": 6.866451612903226e-05,
+      "loss": 0.161,
+      "step": 19929
+    },
+    {
+      "epoch": 0.31888,
+      "grad_norm": 0.84765625,
+      "learning_rate": 6.866290322580645e-05,
+      "loss": 0.1725,
+      "step": 19930
+    },
+    {
+      "epoch": 0.318896,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.866129032258064e-05,
+      "loss": 0.1625,
+      "step": 19931
+    },
+    {
+      "epoch": 0.318912,
+      "grad_norm": 0.76953125,
+      "learning_rate": 6.865967741935484e-05,
+      "loss": 0.178,
+      "step": 19932
+    },
+    {
+      "epoch": 0.318928,
+      "grad_norm": 0.7421875,
+      "learning_rate": 6.865806451612904e-05,
+      "loss": 0.1986,
+      "step": 19933
+    },
+    {
+      "epoch": 0.318944,
+      "grad_norm": 0.90234375,
+      "learning_rate": 6.865645161290323e-05,
+      "loss": 0.1357,
+      "step": 19934
+    },
+    {
+      "epoch": 0.31896,
+      "grad_norm": 0.859375,
+      "learning_rate": 6.865483870967743e-05,
+      "loss": 0.1591,
+      "step": 19935
+    },
+    {
+      "epoch": 0.318976,
+      "grad_norm": 0.49609375,
+      "learning_rate": 6.865322580645162e-05,
+      "loss": 0.127,
+      "step": 19936
+    },
+    {
+      "epoch": 0.318992,
+      "grad_norm": 0.88671875,
+      "learning_rate": 6.865161290322582e-05,
+      "loss": 0.2072,
+      "step": 19937
+    },
+    {
+      "epoch": 0.319008,
+      "grad_norm": 0.80859375,
+      "learning_rate": 6.865e-05,
+      "loss": 0.1731,
+      "step": 19938
+    },
+    {
+      "epoch": 0.319024,
+      "grad_norm": 0.80078125,
+      "learning_rate": 6.864838709677419e-05,
+      "loss": 0.1305,
+      "step": 19939
+    },
+    {
+      "epoch": 0.31904,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.864677419354839e-05,
+      "loss": 0.1596,
+      "step": 19940
+    },
+    {
+      "epoch": 0.319056,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.864516129032257e-05,
+      "loss": 0.1934,
+      "step": 19941
+    },
+    {
+      "epoch": 0.319072,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.864354838709677e-05,
+      "loss": 0.1672,
+      "step": 19942
+    },
+    {
+      "epoch": 0.319088,
+      "grad_norm": 0.8203125,
+      "learning_rate": 6.864193548387097e-05,
+      "loss": 0.1475,
+      "step": 19943
+    },
+    {
+      "epoch": 0.319104,
+      "grad_norm": 1.171875,
+      "learning_rate": 6.864032258064517e-05,
+      "loss": 0.1828,
+      "step": 19944
+    },
+    {
+      "epoch": 0.31912,
+      "grad_norm": 0.88671875,
+      "learning_rate": 6.863870967741936e-05,
+      "loss": 0.1791,
+      "step": 19945
+    },
+    {
+      "epoch": 0.319136,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.863709677419356e-05,
+      "loss": 0.1668,
+      "step": 19946
+    },
+    {
+      "epoch": 0.319152,
+      "grad_norm": 0.80078125,
+      "learning_rate": 6.863548387096774e-05,
+      "loss": 0.1541,
+      "step": 19947
+    },
+    {
+      "epoch": 0.319168,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.863387096774194e-05,
+      "loss": 0.181,
+      "step": 19948
+    },
+    {
+      "epoch": 0.319184,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.863225806451613e-05,
+      "loss": 0.1461,
+      "step": 19949
+    },
+    {
+      "epoch": 0.3192,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.863064516129033e-05,
+      "loss": 0.1961,
+      "step": 19950
+    },
+    {
+      "epoch": 0.319216,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.862903225806452e-05,
+      "loss": 0.1346,
+      "step": 19951
+    },
+    {
+      "epoch": 0.319232,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.862741935483872e-05,
+      "loss": 0.192,
+      "step": 19952
+    },
+    {
+      "epoch": 0.319248,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.86258064516129e-05,
+      "loss": 0.1708,
+      "step": 19953
+    },
+    {
+      "epoch": 0.319264,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.86241935483871e-05,
+      "loss": 0.1698,
+      "step": 19954
+    },
+    {
+      "epoch": 0.31928,
+      "grad_norm": 0.67578125,
+      "learning_rate": 6.86225806451613e-05,
+      "loss": 0.1511,
+      "step": 19955
+    },
+    {
+      "epoch": 0.319296,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.862096774193549e-05,
+      "loss": 0.1707,
+      "step": 19956
+    },
+    {
+      "epoch": 0.319312,
+      "grad_norm": 0.7421875,
+      "learning_rate": 6.861935483870969e-05,
+      "loss": 0.189,
+      "step": 19957
+    },
+    {
+      "epoch": 0.319328,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.861774193548387e-05,
+      "loss": 0.157,
+      "step": 19958
+    },
+    {
+      "epoch": 0.319344,
+      "grad_norm": 0.7421875,
+      "learning_rate": 6.861612903225807e-05,
+      "loss": 0.1752,
+      "step": 19959
+    },
+    {
+      "epoch": 0.31936,
+      "grad_norm": 0.47265625,
+      "learning_rate": 6.861451612903226e-05,
+      "loss": 0.1601,
+      "step": 19960
+    },
+    {
+      "epoch": 0.319376,
+      "grad_norm": 0.78125,
+      "learning_rate": 6.861290322580646e-05,
+      "loss": 0.1753,
+      "step": 19961
+    },
+    {
+      "epoch": 0.319392,
+      "grad_norm": 0.81640625,
+      "learning_rate": 6.861129032258064e-05,
+      "loss": 0.1541,
+      "step": 19962
+    },
+    {
+      "epoch": 0.319408,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.860967741935484e-05,
+      "loss": 0.178,
+      "step": 19963
+    },
+    {
+      "epoch": 0.319424,
+      "grad_norm": 0.7890625,
+      "learning_rate": 6.860806451612903e-05,
+      "loss": 0.1893,
+      "step": 19964
+    },
+    {
+      "epoch": 0.31944,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.860645161290323e-05,
+      "loss": 0.1717,
+      "step": 19965
+    },
+    {
+      "epoch": 0.319456,
+      "grad_norm": 1.2578125,
+      "learning_rate": 6.860483870967742e-05,
+      "loss": 0.1912,
+      "step": 19966
+    },
+    {
+      "epoch": 0.319472,
+      "grad_norm": 1.734375,
+      "learning_rate": 6.860322580645161e-05,
+      "loss": 0.2287,
+      "step": 19967
+    },
+    {
+      "epoch": 0.319488,
+      "grad_norm": 0.9375,
+      "learning_rate": 6.860161290322581e-05,
+      "loss": 0.1857,
+      "step": 19968
+    },
+    {
+      "epoch": 0.319504,
+      "grad_norm": 0.76171875,
+      "learning_rate": 6.860000000000001e-05,
+      "loss": 0.1788,
+      "step": 19969
+    },
+    {
+      "epoch": 0.31952,
+      "grad_norm": 0.83203125,
+      "learning_rate": 6.85983870967742e-05,
+      "loss": 0.1886,
+      "step": 19970
+    },
+    {
+      "epoch": 0.319536,
+      "grad_norm": 0.8359375,
+      "learning_rate": 6.859677419354839e-05,
+      "loss": 0.1574,
+      "step": 19971
+    },
+    {
+      "epoch": 0.319552,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.859516129032259e-05,
+      "loss": 0.1957,
+      "step": 19972
+    },
+    {
+      "epoch": 0.319568,
+      "grad_norm": 0.87109375,
+      "learning_rate": 6.859354838709677e-05,
+      "loss": 0.1482,
+      "step": 19973
+    },
+    {
+      "epoch": 0.319584,
+      "grad_norm": 0.88671875,
+      "learning_rate": 6.859193548387097e-05,
+      "loss": 0.1699,
+      "step": 19974
+    },
+    {
+      "epoch": 0.3196,
+      "grad_norm": 0.7890625,
+      "learning_rate": 6.859032258064516e-05,
+      "loss": 0.161,
+      "step": 19975
+    },
+    {
+      "epoch": 0.319616,
+      "grad_norm": 0.90234375,
+      "learning_rate": 6.858870967741936e-05,
+      "loss": 0.2062,
+      "step": 19976
+    },
+    {
+      "epoch": 0.319632,
+      "grad_norm": 0.8125,
+      "learning_rate": 6.858709677419354e-05,
+      "loss": 0.1603,
+      "step": 19977
+    },
+    {
+      "epoch": 0.319648,
+      "grad_norm": 0.84765625,
+      "learning_rate": 6.858548387096774e-05,
+      "loss": 0.189,
+      "step": 19978
+    },
+    {
+      "epoch": 0.319664,
+      "grad_norm": 0.74609375,
+      "learning_rate": 6.858387096774194e-05,
+      "loss": 0.1777,
+      "step": 19979
+    },
+    {
+      "epoch": 0.31968,
+      "grad_norm": 0.76953125,
+      "learning_rate": 6.858225806451614e-05,
+      "loss": 0.1754,
+      "step": 19980
+    },
+    {
+      "epoch": 0.319696,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.858064516129033e-05,
+      "loss": 0.1428,
+      "step": 19981
+    },
+    {
+      "epoch": 0.319712,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.857903225806453e-05,
+      "loss": 0.1464,
+      "step": 19982
+    },
+    {
+      "epoch": 0.319728,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.857741935483871e-05,
+      "loss": 0.1831,
+      "step": 19983
+    },
+    {
+      "epoch": 0.319744,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.857580645161291e-05,
+      "loss": 0.1583,
+      "step": 19984
+    },
+    {
+      "epoch": 0.31976,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.85741935483871e-05,
+      "loss": 0.1688,
+      "step": 19985
+    },
+    {
+      "epoch": 0.319776,
+      "grad_norm": 0.87109375,
+      "learning_rate": 6.857258064516129e-05,
+      "loss": 0.1823,
+      "step": 19986
+    },
+    {
+      "epoch": 0.319792,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.857096774193549e-05,
+      "loss": 0.1429,
+      "step": 19987
+    },
+    {
+      "epoch": 0.319808,
+      "grad_norm": 1.53125,
+      "learning_rate": 6.856935483870967e-05,
+      "loss": 0.2029,
+      "step": 19988
+    },
+    {
+      "epoch": 0.319824,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.856774193548387e-05,
+      "loss": 0.1644,
+      "step": 19989
+    },
+    {
+      "epoch": 0.31984,
+      "grad_norm": 0.447265625,
+      "learning_rate": 6.856612903225807e-05,
+      "loss": 0.139,
+      "step": 19990
+    },
+    {
+      "epoch": 0.319856,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.856451612903226e-05,
+      "loss": 0.1537,
+      "step": 19991
+    },
+    {
+      "epoch": 0.319872,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.856290322580646e-05,
+      "loss": 0.1444,
+      "step": 19992
+    },
+    {
+      "epoch": 0.319888,
+      "grad_norm": 0.9296875,
+      "learning_rate": 6.856129032258066e-05,
+      "loss": 0.1697,
+      "step": 19993
+    },
+    {
+      "epoch": 0.319904,
+      "grad_norm": 1.125,
+      "learning_rate": 6.855967741935484e-05,
+      "loss": 0.1623,
+      "step": 19994
+    },
+    {
+      "epoch": 0.31992,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.855806451612904e-05,
+      "loss": 0.1563,
+      "step": 19995
+    },
+    {
+      "epoch": 0.319936,
+      "grad_norm": 0.7421875,
+      "learning_rate": 6.855645161290323e-05,
+      "loss": 0.1543,
+      "step": 19996
+    },
+    {
+      "epoch": 0.319952,
+      "grad_norm": 0.7578125,
+      "learning_rate": 6.855483870967743e-05,
+      "loss": 0.1597,
+      "step": 19997
+    },
+    {
+      "epoch": 0.319968,
+      "grad_norm": 1.328125,
+      "learning_rate": 6.855322580645161e-05,
+      "loss": 0.1128,
+      "step": 19998
+    },
+    {
+      "epoch": 0.319984,
+      "grad_norm": 1.265625,
+      "learning_rate": 6.855161290322581e-05,
+      "loss": 0.215,
+      "step": 19999
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.855e-05,
+      "loss": 0.148,
+      "step": 20000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 62500,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 10000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.492612561421189e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}