diff --git "a/checkpoint-2618/trainer_state.json" "b/checkpoint-2618/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2618/trainer_state.json"
@@ -0,0 +1,18379 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 655,
+  "global_step": 2618,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.4369395077228546,
+      "learning_rate": 1e-05,
+      "loss": 1.3161,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "eval_loss": 1.559309482574463,
+      "eval_runtime": 17.7472,
+      "eval_samples_per_second": 153.771,
+      "eval_steps_per_second": 38.485,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5338654518127441,
+      "learning_rate": 2e-05,
+      "loss": 1.56,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.517049252986908,
+      "learning_rate": 3e-05,
+      "loss": 1.519,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6359376907348633,
+      "learning_rate": 4e-05,
+      "loss": 1.5065,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5600838661193848,
+      "learning_rate": 5e-05,
+      "loss": 1.3367,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.49170756340026855,
+      "learning_rate": 6e-05,
+      "loss": 1.4911,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5616886019706726,
+      "learning_rate": 7e-05,
+      "loss": 1.4518,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5916192531585693,
+      "learning_rate": 8e-05,
+      "loss": 1.1924,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6707382798194885,
+      "learning_rate": 9e-05,
+      "loss": 1.281,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8203673362731934,
+      "learning_rate": 0.0001,
+      "loss": 1.4793,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5819201469421387,
+      "learning_rate": 0.00011000000000000002,
+      "loss": 1.2622,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6588992476463318,
+      "learning_rate": 0.00012,
+      "loss": 1.226,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7845558524131775,
+      "learning_rate": 0.00013000000000000002,
+      "loss": 1.3695,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5720924139022827,
+      "learning_rate": 0.00014,
+      "loss": 1.3338,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5074018836021423,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 0.9675,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5563233494758606,
+      "learning_rate": 0.00016,
+      "loss": 1.3316,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5730008482933044,
+      "learning_rate": 0.00017,
+      "loss": 1.1587,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.557260274887085,
+      "learning_rate": 0.00018,
+      "loss": 1.0954,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5220329165458679,
+      "learning_rate": 0.00019,
+      "loss": 1.2881,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6541146039962769,
+      "learning_rate": 0.0002,
+      "loss": 1.1474,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.621552050113678,
+      "learning_rate": 0.0001999999954827831,
+      "loss": 1.1891,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6264808177947998,
+      "learning_rate": 0.0001999999819311328,
+      "loss": 1.2244,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.550438642501831,
+      "learning_rate": 0.00019999995934505033,
+      "loss": 1.3496,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5227052569389343,
+      "learning_rate": 0.00019999992772453778,
+      "loss": 1.1316,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4750714898109436,
+      "learning_rate": 0.0001999998870695979,
+      "loss": 1.1112,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5615024566650391,
+      "learning_rate": 0.00019999983738023446,
+      "loss": 1.2386,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4816851317882538,
+      "learning_rate": 0.00019999977865645186,
+      "loss": 1.0982,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5965765118598938,
+      "learning_rate": 0.00019999971089825549,
+      "loss": 1.2331,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5031258463859558,
+      "learning_rate": 0.00019999963410565142,
+      "loss": 1.2095,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5152090787887573,
+      "learning_rate": 0.00019999954827864663,
+      "loss": 1.1875,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5212087035179138,
+      "learning_rate": 0.0001999994534172488,
+      "loss": 1.1729,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4427972137928009,
+      "learning_rate": 0.0001999993495214666,
+      "loss": 1.129,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5280154943466187,
+      "learning_rate": 0.00019999923659130934,
+      "loss": 1.1871,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.437427282333374,
+      "learning_rate": 0.00019999911462678728,
+      "loss": 1.0607,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4341500699520111,
+      "learning_rate": 0.00019999898362791139,
+      "loss": 1.1404,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.45423945784568787,
+      "learning_rate": 0.0001999988435946935,
+      "loss": 1.1451,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.47884103655815125,
+      "learning_rate": 0.0001999986945271463,
+      "loss": 1.1615,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4309810996055603,
+      "learning_rate": 0.00019999853642528327,
+      "loss": 1.0011,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4934332072734833,
+      "learning_rate": 0.0001999983692891186,
+      "loss": 1.2428,
+      "step": 39
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5232778787612915,
+      "learning_rate": 0.0001999981931186675,
+      "loss": 1.2445,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4906657934188843,
+      "learning_rate": 0.00019999800791394582,
+      "loss": 1.112,
+      "step": 41
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4859429895877838,
+      "learning_rate": 0.00019999781367497033,
+      "loss": 1.2359,
+      "step": 42
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5400615930557251,
+      "learning_rate": 0.00019999761040175856,
+      "loss": 1.1782,
+      "step": 43
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.48445501923561096,
+      "learning_rate": 0.00019999739809432887,
+      "loss": 1.1459,
+      "step": 44
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6529949307441711,
+      "learning_rate": 0.0001999971767527004,
+      "loss": 1.2262,
+      "step": 45
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5127531886100769,
+      "learning_rate": 0.00019999694637689327,
+      "loss": 1.1846,
+      "step": 46
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5761195421218872,
+      "learning_rate": 0.00019999670696692817,
+      "loss": 1.009,
+      "step": 47
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.469578355550766,
+      "learning_rate": 0.0001999964585228268,
+      "loss": 1.0883,
+      "step": 48
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5488829612731934,
+      "learning_rate": 0.00019999620104461157,
+      "loss": 1.1774,
+      "step": 49
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5317593216896057,
+      "learning_rate": 0.00019999593453230575,
+      "loss": 1.2767,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4968358278274536,
+      "learning_rate": 0.0001999956589859334,
+      "loss": 1.2343,
+      "step": 51
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5535591244697571,
+      "learning_rate": 0.00019999537440551946,
+      "loss": 1.18,
+      "step": 52
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4718022644519806,
+      "learning_rate": 0.00019999508079108962,
+      "loss": 1.1936,
+      "step": 53
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5101296305656433,
+      "learning_rate": 0.0001999947781426704,
+      "loss": 1.3481,
+      "step": 54
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4822118282318115,
+      "learning_rate": 0.00019999446646028912,
+      "loss": 1.1209,
+      "step": 55
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5444576144218445,
+      "learning_rate": 0.00019999414574397396,
+      "loss": 1.2257,
+      "step": 56
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5353034734725952,
+      "learning_rate": 0.00019999381599375394,
+      "loss": 1.2426,
+      "step": 57
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.65821373462677,
+      "learning_rate": 0.0001999934772096588,
+      "loss": 1.1748,
+      "step": 58
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5470272898674011,
+      "learning_rate": 0.00019999312939171914,
+      "loss": 1.1972,
+      "step": 59
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5184686779975891,
+      "learning_rate": 0.00019999277253996637,
+      "loss": 1.1509,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5672779083251953,
+      "learning_rate": 0.0001999924066544328,
+      "loss": 1.1963,
+      "step": 61
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5974107980728149,
+      "learning_rate": 0.00019999203173515142,
+      "loss": 1.1709,
+      "step": 62
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5848557949066162,
+      "learning_rate": 0.00019999164778215613,
+      "loss": 1.1453,
+      "step": 63
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.48387718200683594,
+      "learning_rate": 0.00019999125479548165,
+      "loss": 0.9851,
+      "step": 64
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.44351813197135925,
+      "learning_rate": 0.0001999908527751634,
+      "loss": 1.105,
+      "step": 65
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.500687301158905,
+      "learning_rate": 0.00019999044172123775,
+      "loss": 1.1271,
+      "step": 66
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5383696556091309,
+      "learning_rate": 0.00019999002163374188,
+      "loss": 1.2929,
+      "step": 67
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4801323115825653,
+      "learning_rate": 0.00019998959251271367,
+      "loss": 0.9591,
+      "step": 68
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5076848864555359,
+      "learning_rate": 0.0001999891543581919,
+      "loss": 1.1723,
+      "step": 69
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4907875955104828,
+      "learning_rate": 0.0001999887071702162,
+      "loss": 1.126,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6520359516143799,
+      "learning_rate": 0.00019998825094882694,
+      "loss": 1.2344,
+      "step": 71
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5223540663719177,
+      "learning_rate": 0.0001999877856940653,
+      "loss": 1.2312,
+      "step": 72
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5751238465309143,
+      "learning_rate": 0.00019998731140597338,
+      "loss": 1.2276,
+      "step": 73
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.540924608707428,
+      "learning_rate": 0.000199986828084594,
+      "loss": 1.1759,
+      "step": 74
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4565652310848236,
+      "learning_rate": 0.00019998633572997085,
+      "loss": 1.0503,
+      "step": 75
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.46219968795776367,
+      "learning_rate": 0.00019998583434214834,
+      "loss": 1.1385,
+      "step": 76
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5129724740982056,
+      "learning_rate": 0.00019998532392117183,
+      "loss": 1.1054,
+      "step": 77
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.576178252696991,
+      "learning_rate": 0.00019998480446708746,
+      "loss": 1.2987,
+      "step": 78
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5178436040878296,
+      "learning_rate": 0.00019998427597994208,
+      "loss": 1.0454,
+      "step": 79
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5002140998840332,
+      "learning_rate": 0.0001999837384597835,
+      "loss": 1.1912,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5395256280899048,
+      "learning_rate": 0.00019998319190666023,
+      "loss": 1.1172,
+      "step": 81
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5830578207969666,
+      "learning_rate": 0.00019998263632062168,
+      "loss": 1.15,
+      "step": 82
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.46478208899497986,
+      "learning_rate": 0.00019998207170171805,
+      "loss": 1.1597,
+      "step": 83
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5071977376937866,
+      "learning_rate": 0.00019998149805000034,
+      "loss": 1.1847,
+      "step": 84
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.455512672662735,
+      "learning_rate": 0.00019998091536552035,
+      "loss": 1.0309,
+      "step": 85
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.48724740743637085,
+      "learning_rate": 0.00019998032364833076,
+      "loss": 1.2756,
+      "step": 86
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.48692628741264343,
+      "learning_rate": 0.00019997972289848503,
+      "loss": 1.1126,
+      "step": 87
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5381304621696472,
+      "learning_rate": 0.00019997911311603737,
+      "loss": 1.2491,
+      "step": 88
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.44811877608299255,
+      "learning_rate": 0.00019997849430104298,
+      "loss": 1.0797,
+      "step": 89
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5321200489997864,
+      "learning_rate": 0.0001999778664535577,
+      "loss": 1.2419,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5130752325057983,
+      "learning_rate": 0.00019997722957363823,
+      "loss": 1.1859,
+      "step": 91
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.49227434396743774,
+      "learning_rate": 0.00019997658366134217,
+      "loss": 1.2368,
+      "step": 92
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5907085537910461,
+      "learning_rate": 0.0001999759287167278,
+      "loss": 1.0933,
+      "step": 93
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5108592510223389,
+      "learning_rate": 0.00019997526473985438,
+      "loss": 1.214,
+      "step": 94
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5565717816352844,
+      "learning_rate": 0.0001999745917307818,
+      "loss": 1.3146,
+      "step": 95
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5041968822479248,
+      "learning_rate": 0.00019997390968957098,
+      "loss": 1.2215,
+      "step": 96
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.49414822459220886,
+      "learning_rate": 0.0001999732186162834,
+      "loss": 1.1545,
+      "step": 97
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5626671314239502,
+      "learning_rate": 0.0001999725185109816,
+      "loss": 1.3104,
+      "step": 98
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.459627628326416,
+      "learning_rate": 0.00019997180937372878,
+      "loss": 1.1461,
+      "step": 99
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.484120637178421,
+      "learning_rate": 0.00019997109120458903,
+      "loss": 1.0694,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5494018197059631,
+      "learning_rate": 0.00019997036400362723,
+      "loss": 1.1665,
+      "step": 101
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4950861632823944,
+      "learning_rate": 0.00019996962777090905,
+      "loss": 1.1021,
+      "step": 102
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4946760833263397,
+      "learning_rate": 0.00019996888250650108,
+      "loss": 1.2554,
+      "step": 103
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5194969773292542,
+      "learning_rate": 0.00019996812821047055,
+      "loss": 1.0449,
+      "step": 104
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6520335078239441,
+      "learning_rate": 0.00019996736488288568,
+      "loss": 0.9141,
+      "step": 105
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5012494921684265,
+      "learning_rate": 0.0001999665925238154,
+      "loss": 1.2185,
+      "step": 106
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.47467753291130066,
+      "learning_rate": 0.00019996581113332948,
+      "loss": 1.106,
+      "step": 107
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.45975348353385925,
+      "learning_rate": 0.00019996502071149854,
+      "loss": 1.1995,
+      "step": 108
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5025244355201721,
+      "learning_rate": 0.00019996422125839397,
+      "loss": 1.0846,
+      "step": 109
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5010673999786377,
+      "learning_rate": 0.00019996341277408802,
+      "loss": 1.1118,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5463888645172119,
+      "learning_rate": 0.0001999625952586537,
+      "loss": 1.3421,
+      "step": 111
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4908903241157532,
+      "learning_rate": 0.0001999617687121649,
+      "loss": 1.2373,
+      "step": 112
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4427969753742218,
+      "learning_rate": 0.00019996093313469628,
+      "loss": 1.0984,
+      "step": 113
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.45442789793014526,
+      "learning_rate": 0.00019996008852632332,
+      "loss": 1.1715,
+      "step": 114
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4985520541667938,
+      "learning_rate": 0.00019995923488712233,
+      "loss": 1.0816,
+      "step": 115
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5188296437263489,
+      "learning_rate": 0.00019995837221717044,
+      "loss": 1.1667,
+      "step": 116
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.547581136226654,
+      "learning_rate": 0.0001999575005165456,
+      "loss": 1.1387,
+      "step": 117
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5591850876808167,
+      "learning_rate": 0.00019995661978532652,
+      "loss": 1.2088,
+      "step": 118
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5121439695358276,
+      "learning_rate": 0.0001999557300235928,
+      "loss": 1.2097,
+      "step": 119
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4894709885120392,
+      "learning_rate": 0.00019995483123142483,
+      "loss": 1.1892,
+      "step": 120
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5119152665138245,
+      "learning_rate": 0.00019995392340890379,
+      "loss": 1.257,
+      "step": 121
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4950897991657257,
+      "learning_rate": 0.00019995300655611172,
+      "loss": 1.247,
+      "step": 122
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5071575045585632,
+      "learning_rate": 0.00019995208067313144,
+      "loss": 1.219,
+      "step": 123
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.48433318734169006,
+      "learning_rate": 0.0001999511457600466,
+      "loss": 1.1201,
+      "step": 124
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.507840096950531,
+      "learning_rate": 0.00019995020181694167,
+      "loss": 1.0267,
+      "step": 125
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.604081392288208,
+      "learning_rate": 0.00019994924884390188,
+      "loss": 1.0828,
+      "step": 126
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5288147926330566,
+      "learning_rate": 0.0001999482868410134,
+      "loss": 1.1112,
+      "step": 127
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5043405294418335,
+      "learning_rate": 0.00019994731580836312,
+      "loss": 1.1041,
+      "step": 128
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5322749614715576,
+      "learning_rate": 0.00019994633574603877,
+      "loss": 0.8943,
+      "step": 129
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5540361404418945,
+      "learning_rate": 0.00019994534665412883,
+      "loss": 1.0218,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5248005390167236,
+      "learning_rate": 0.00019994434853272272,
+      "loss": 1.0896,
+      "step": 131
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5261930227279663,
+      "learning_rate": 0.00019994334138191065,
+      "loss": 0.9431,
+      "step": 132
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5371242761611938,
+      "learning_rate": 0.00019994232520178352,
+      "loss": 1.2457,
+      "step": 133
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4499124586582184,
+      "learning_rate": 0.0001999412999924332,
+      "loss": 1.1453,
+      "step": 134
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4927183985710144,
+      "learning_rate": 0.00019994026575395225,
+      "loss": 0.9934,
+      "step": 135
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4928085505962372,
+      "learning_rate": 0.00019993922248643417,
+      "loss": 1.1859,
+      "step": 136
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5777192711830139,
+      "learning_rate": 0.00019993817018997323,
+      "loss": 1.124,
+      "step": 137
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4621788561344147,
+      "learning_rate": 0.00019993710886466443,
+      "loss": 1.0674,
+      "step": 138
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5149490237236023,
+      "learning_rate": 0.00019993603851060367,
+      "loss": 1.2451,
+      "step": 139
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.46570640802383423,
+      "learning_rate": 0.0001999349591278877,
+      "loss": 1.0202,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.48668521642684937,
+      "learning_rate": 0.000199933870716614,
+      "loss": 1.1878,
+      "step": 141
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5824366211891174,
+      "learning_rate": 0.0001999327732768809,
+      "loss": 1.0239,
+      "step": 142
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.447409451007843,
+      "learning_rate": 0.00019993166680878756,
+      "loss": 1.0978,
+      "step": 143
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.42665061354637146,
+      "learning_rate": 0.00019993055131243393,
+      "loss": 1.0087,
+      "step": 144
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4215356409549713,
+      "learning_rate": 0.00019992942678792076,
+      "loss": 0.9732,
+      "step": 145
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5241891145706177,
+      "learning_rate": 0.00019992829323534973,
+      "loss": 1.096,
+      "step": 146
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.483650803565979,
+      "learning_rate": 0.00019992715065482317,
+      "loss": 1.1588,
+      "step": 147
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5333243608474731,
+      "learning_rate": 0.00019992599904644433,
+      "loss": 1.0622,
+      "step": 148
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4784759283065796,
+      "learning_rate": 0.00019992483841031725,
+      "loss": 1.1042,
+      "step": 149
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.47385480999946594,
+      "learning_rate": 0.00019992366874654684,
+      "loss": 1.1554,
+      "step": 150
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.49824684858322144,
+      "learning_rate": 0.00019992249005523869,
+      "loss": 1.1757,
+      "step": 151
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5586369633674622,
+      "learning_rate": 0.00019992130233649933,
+      "loss": 1.0107,
+      "step": 152
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5922762751579285,
+      "learning_rate": 0.00019992010559043605,
+      "loss": 1.1978,
+      "step": 153
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4725801646709442,
+      "learning_rate": 0.00019991889981715698,
+      "loss": 0.993,
+      "step": 154
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5265770554542542,
+      "learning_rate": 0.00019991768501677104,
+      "loss": 1.0756,
+      "step": 155
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5356742143630981,
+      "learning_rate": 0.00019991646118938802,
+      "loss": 1.1144,
+      "step": 156
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.45678016543388367,
+      "learning_rate": 0.00019991522833511842,
+      "loss": 1.0504,
+      "step": 157
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.48851096630096436,
+      "learning_rate": 0.0001999139864540737,
+      "loss": 1.2359,
+      "step": 158
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4795781373977661,
+      "learning_rate": 0.00019991273554636595,
+      "loss": 1.1277,
+      "step": 159
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4898346960544586,
+      "learning_rate": 0.00019991147561210834,
+      "loss": 1.1664,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5645068883895874,
+      "learning_rate": 0.00019991020665141454,
+      "loss": 1.2044,
+      "step": 161
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.488219290971756,
+      "learning_rate": 0.00019990892866439928,
+      "loss": 1.1247,
+      "step": 162
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5142199397087097,
+      "learning_rate": 0.000199907641651178,
+      "loss": 1.178,
+      "step": 163
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5282805562019348,
+      "learning_rate": 0.00019990634561186695,
+      "loss": 1.1834,
+      "step": 164
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5011789798736572,
+      "learning_rate": 0.00019990504054658327,
+      "loss": 1.3229,
+      "step": 165
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4901934266090393,
+      "learning_rate": 0.00019990372645544482,
+      "loss": 1.1035,
+      "step": 166
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.49366825819015503,
+      "learning_rate": 0.00019990240333857035,
+      "loss": 1.3065,
+      "step": 167
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5546338558197021,
+      "learning_rate": 0.00019990107119607938,
+      "loss": 1.2996,
+      "step": 168
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4404127895832062,
+      "learning_rate": 0.00019989973002809225,
+      "loss": 1.1766,
+      "step": 169
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6145757436752319,
+      "learning_rate": 0.00019989837983473016,
+      "loss": 1.3385,
+      "step": 170
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5417330265045166,
+      "learning_rate": 0.00019989702061611506,
+      "loss": 1.1634,
+      "step": 171
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4397343397140503,
+      "learning_rate": 0.00019989565237236978,
+      "loss": 1.0259,
+      "step": 172
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.48129138350486755,
+      "learning_rate": 0.0001998942751036179,
+      "loss": 1.0549,
+      "step": 173
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4494941234588623,
+      "learning_rate": 0.00019989288880998388,
+      "loss": 0.9936,
+      "step": 174
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4411003291606903,
+      "learning_rate": 0.00019989149349159294,
+      "loss": 1.2235,
+      "step": 175
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4981510043144226,
+      "learning_rate": 0.00019989008914857116,
+      "loss": 1.0567,
+      "step": 176
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.43061527609825134,
+      "learning_rate": 0.0001998886757810454,
+      "loss": 0.8618,
+      "step": 177
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4655401408672333,
+      "learning_rate": 0.00019988725338914335,
+      "loss": 1.2116,
+      "step": 178
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.43923428654670715,
+      "learning_rate": 0.00019988582197299348,
+      "loss": 1.102,
+      "step": 179
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5278385877609253,
+      "learning_rate": 0.0001998843815327252,
+      "loss": 1.1409,
+      "step": 180
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.47227001190185547,
+      "learning_rate": 0.00019988293206846854,
+      "loss": 0.9435,
+      "step": 181
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.46571439504623413,
+      "learning_rate": 0.00019988147358035454,
+      "loss": 1.0508,
+      "step": 182
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.44509732723236084,
+      "learning_rate": 0.0001998800060685149,
+      "loss": 1.0575,
+      "step": 183
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4418330192565918,
+      "learning_rate": 0.00019987852953308227,
+      "loss": 0.9238,
+      "step": 184
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5942162275314331,
+      "learning_rate": 0.00019987704397418998,
+      "loss": 1.3404,
+      "step": 185
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.49123358726501465,
+      "learning_rate": 0.0001998755493919723,
+      "loss": 1.2518,
+      "step": 186
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.48855024576187134,
+      "learning_rate": 0.00019987404578656421,
+      "loss": 1.1701,
+      "step": 187
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.450970321893692,
+      "learning_rate": 0.00019987253315810154,
+      "loss": 1.0354,
+      "step": 188
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4818516671657562,
+      "learning_rate": 0.00019987101150672102,
+      "loss": 0.9924,
+      "step": 189
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5119654536247253,
+      "learning_rate": 0.00019986948083256003,
+      "loss": 1.157,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5616949200630188,
+      "learning_rate": 0.00019986794113575695,
+      "loss": 1.2006,
+      "step": 191
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5410425066947937,
+      "learning_rate": 0.00019986639241645083,
+      "loss": 1.0805,
+      "step": 192
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.47965753078460693,
+      "learning_rate": 0.00019986483467478158,
+      "loss": 1.0842,
+      "step": 193
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.45503634214401245,
+      "learning_rate": 0.00019986326791088997,
+      "loss": 1.1245,
+      "step": 194
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5413264632225037,
+      "learning_rate": 0.0001998616921249175,
+      "loss": 1.3534,
+      "step": 195
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.45204904675483704,
+      "learning_rate": 0.0001998601073170066,
+      "loss": 1.0567,
+      "step": 196
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.44439467787742615,
+      "learning_rate": 0.0001998585134873004,
+      "loss": 1.1653,
+      "step": 197
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5485827326774597,
+      "learning_rate": 0.00019985691063594287,
+      "loss": 1.3116,
+      "step": 198
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.466046541929245,
+      "learning_rate": 0.0001998552987630789,
+      "loss": 0.9963,
+      "step": 199
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5031853914260864,
+      "learning_rate": 0.00019985367786885404,
+      "loss": 1.184,
+      "step": 200
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5012284517288208,
+      "learning_rate": 0.00019985204795341477,
+      "loss": 1.0757,
+      "step": 201
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.48965683579444885,
+      "learning_rate": 0.0001998504090169083,
+      "loss": 1.2179,
+      "step": 202
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5679877996444702,
+      "learning_rate": 0.00019984876105948274,
+      "loss": 1.109,
+      "step": 203
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.559474766254425,
+      "learning_rate": 0.00019984710408128695,
+      "loss": 1.2473,
+      "step": 204
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4734433591365814,
+      "learning_rate": 0.00019984543808247067,
+      "loss": 1.1077,
+      "step": 205
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4242611229419708,
+      "learning_rate": 0.00019984376306318437,
+      "loss": 1.1687,
+      "step": 206
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.48564302921295166,
+      "learning_rate": 0.00019984207902357936,
+      "loss": 1.1441,
+      "step": 207
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4814341068267822,
+      "learning_rate": 0.00019984038596380788,
+      "loss": 1.0971,
+      "step": 208
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.506041407585144,
+      "learning_rate": 0.00019983868388402277,
+      "loss": 1.1421,
+      "step": 209
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4765606224536896,
+      "learning_rate": 0.00019983697278437785,
+      "loss": 0.9763,
+      "step": 210
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4573558568954468,
+      "learning_rate": 0.00019983525266502775,
+      "loss": 1.0432,
+      "step": 211
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.49923500418663025,
+      "learning_rate": 0.0001998335235261278,
+      "loss": 0.9541,
+      "step": 212
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8542819619178772,
+      "learning_rate": 0.0001998317853678343,
+      "loss": 1.2511,
+      "step": 213
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.48192134499549866,
+      "learning_rate": 0.0001998300381903042,
+      "loss": 1.037,
+      "step": 214
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5000719428062439,
+      "learning_rate": 0.00019982828199369541,
+      "loss": 1.2035,
+      "step": 215
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5127101540565491,
+      "learning_rate": 0.00019982651677816657,
+      "loss": 1.0139,
+      "step": 216
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5559672713279724,
+      "learning_rate": 0.00019982474254387716,
+      "loss": 1.123,
+      "step": 217
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.47846996784210205,
+      "learning_rate": 0.00019982295929098744,
+      "loss": 1.1112,
+      "step": 218
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5220378041267395,
+      "learning_rate": 0.00019982116701965853,
+      "loss": 1.1898,
+      "step": 219
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4617263376712799,
+      "learning_rate": 0.0001998193657300524,
+      "loss": 1.0178,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5010495781898499,
+      "learning_rate": 0.00019981755542233177,
+      "loss": 1.0739,
+      "step": 221
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.47030919790267944,
+      "learning_rate": 0.00019981573609666013,
+      "loss": 0.943,
+      "step": 222
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.48498865962028503,
+      "learning_rate": 0.00019981390775320192,
+      "loss": 1.0353,
+      "step": 223
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.48837897181510925,
+      "learning_rate": 0.00019981207039212226,
+      "loss": 1.0481,
+      "step": 224
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6528398394584656,
+      "learning_rate": 0.0001998102240135872,
+      "loss": 1.1928,
+      "step": 225
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.46565020084381104,
+      "learning_rate": 0.0001998083686177635,
+      "loss": 1.1219,
+      "step": 226
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.48760753870010376,
+      "learning_rate": 0.00019980650420481884,
+      "loss": 1.1069,
+      "step": 227
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.45117929577827454,
+      "learning_rate": 0.0001998046307749216,
+      "loss": 1.1878,
+      "step": 228
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5408883690834045,
+      "learning_rate": 0.00019980274832824107,
+      "loss": 1.2435,
+      "step": 229
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5185354351997375,
+      "learning_rate": 0.00019980085686494733,
+      "loss": 1.1932,
+      "step": 230
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4900776445865631,
+      "learning_rate": 0.00019979895638521123,
+      "loss": 1.2784,
+      "step": 231
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.47951623797416687,
+      "learning_rate": 0.00019979704688920447,
+      "loss": 1.1653,
+      "step": 232
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.47839614748954773,
+      "learning_rate": 0.00019979512837709957,
+      "loss": 1.1518,
+      "step": 233
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5080764293670654,
+      "learning_rate": 0.00019979320084906988,
+      "loss": 1.2718,
+      "step": 234
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4477003812789917,
+      "learning_rate": 0.00019979126430528949,
+      "loss": 1.1909,
+      "step": 235
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5127712488174438,
+      "learning_rate": 0.00019978931874593342,
+      "loss": 1.0702,
+      "step": 236
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4719958007335663,
+      "learning_rate": 0.0001997873641711774,
+      "loss": 1.1527,
+      "step": 237
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.571456789970398,
+      "learning_rate": 0.000199785400581198,
+      "loss": 1.1834,
+      "step": 238
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5032457113265991,
+      "learning_rate": 0.00019978342797617266,
+      "loss": 1.0674,
+      "step": 239
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.48442238569259644,
+      "learning_rate": 0.00019978144635627958,
+      "loss": 1.0571,
+      "step": 240
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4788631498813629,
+      "learning_rate": 0.0001997794557216978,
+      "loss": 1.1184,
+      "step": 241
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.598183274269104,
+      "learning_rate": 0.00019977745607260708,
+      "loss": 1.0265,
+      "step": 242
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5404005646705627,
+      "learning_rate": 0.0001997754474091882,
+      "loss": 1.2397,
+      "step": 243
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4596356749534607,
+      "learning_rate": 0.00019977342973162253,
+      "loss": 1.1055,
+      "step": 244
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5855688452720642,
+      "learning_rate": 0.00019977140304009244,
+      "loss": 1.2011,
+      "step": 245
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5521079897880554,
+      "learning_rate": 0.00019976936733478098,
+      "loss": 1.2093,
+      "step": 246
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.623518168926239,
+      "learning_rate": 0.00019976732261587206,
+      "loss": 1.1398,
+      "step": 247
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4470531940460205,
+      "learning_rate": 0.00019976526888355043,
+      "loss": 1.0628,
+      "step": 248
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5066975951194763,
+      "learning_rate": 0.00019976320613800162,
+      "loss": 1.0332,
+      "step": 249
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5067146420478821,
+      "learning_rate": 0.00019976113437941202,
+      "loss": 1.2509,
+      "step": 250
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.475010484457016,
+      "learning_rate": 0.00019975905360796877,
+      "loss": 1.1128,
+      "step": 251
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5198237895965576,
+      "learning_rate": 0.00019975696382385985,
+      "loss": 0.9448,
+      "step": 252
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.42087745666503906,
+      "learning_rate": 0.00019975486502727405,
+      "loss": 1.0128,
+      "step": 253
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.45363667607307434,
+      "learning_rate": 0.00019975275721840103,
+      "loss": 1.1145,
+      "step": 254
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5228469967842102,
+      "learning_rate": 0.0001997506403974312,
+      "loss": 1.0619,
+      "step": 255
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5582649111747742,
+      "learning_rate": 0.00019974851456455576,
+      "loss": 1.2683,
+      "step": 256
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.507866382598877,
+      "learning_rate": 0.00019974637971996684,
+      "loss": 1.1629,
+      "step": 257
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4675002694129944,
+      "learning_rate": 0.00019974423586385724,
+      "loss": 1.172,
+      "step": 258
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.508798360824585,
+      "learning_rate": 0.00019974208299642073,
+      "loss": 1.2427,
+      "step": 259
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5536615252494812,
+      "learning_rate": 0.00019973992111785172,
+      "loss": 0.9203,
+      "step": 260
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5231752991676331,
+      "learning_rate": 0.00019973775022834555,
+      "loss": 1.0862,
+      "step": 261
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.45515871047973633,
+      "learning_rate": 0.0001997355703280984,
+      "loss": 1.1443,
+      "step": 262
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6466176509857178,
+      "learning_rate": 0.00019973338141730714,
+      "loss": 1.3494,
+      "step": 263
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4798348844051361,
+      "learning_rate": 0.0001997311834961696,
+      "loss": 1.065,
+      "step": 264
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5107787251472473,
+      "learning_rate": 0.00019972897656488428,
+      "loss": 1.153,
+      "step": 265
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.44574907422065735,
+      "learning_rate": 0.00019972676062365058,
+      "loss": 0.8468,
+      "step": 266
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5033632516860962,
+      "learning_rate": 0.00019972453567266874,
+      "loss": 1.2483,
+      "step": 267
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5271995663642883,
+      "learning_rate": 0.00019972230171213972,
+      "loss": 0.977,
+      "step": 268
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.48854905366897583,
+      "learning_rate": 0.00019972005874226536,
+      "loss": 1.2457,
+      "step": 269
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5067196488380432,
+      "learning_rate": 0.00019971780676324833,
+      "loss": 1.0715,
+      "step": 270
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4790126383304596,
+      "learning_rate": 0.00019971554577529203,
+      "loss": 1.1117,
+      "step": 271
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4900168478488922,
+      "learning_rate": 0.00019971327577860077,
+      "loss": 1.1876,
+      "step": 272
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.654327392578125,
+      "learning_rate": 0.00019971099677337963,
+      "loss": 1.2282,
+      "step": 273
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4578112065792084,
+      "learning_rate": 0.00019970870875983448,
+      "loss": 1.1386,
+      "step": 274
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5260718464851379,
+      "learning_rate": 0.00019970641173817203,
+      "loss": 1.1163,
+      "step": 275
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5048907995223999,
+      "learning_rate": 0.00019970410570859984,
+      "loss": 1.0529,
+      "step": 276
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4529321789741516,
+      "learning_rate": 0.00019970179067132623,
+      "loss": 1.188,
+      "step": 277
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.490434467792511,
+      "learning_rate": 0.00019969946662656032,
+      "loss": 1.127,
+      "step": 278
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4846895635128021,
+      "learning_rate": 0.00019969713357451208,
+      "loss": 1.2213,
+      "step": 279
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4657130837440491,
+      "learning_rate": 0.00019969479151539236,
+      "loss": 1.1398,
+      "step": 280
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.44968271255493164,
+      "learning_rate": 0.00019969244044941266,
+      "loss": 1.0703,
+      "step": 281
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.49765828251838684,
+      "learning_rate": 0.00019969008037678542,
+      "loss": 0.955,
+      "step": 282
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4985436797142029,
+      "learning_rate": 0.0001996877112977239,
+      "loss": 1.1547,
+      "step": 283
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4842534363269806,
+      "learning_rate": 0.00019968533321244208,
+      "loss": 1.1838,
+      "step": 284
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4974116086959839,
+      "learning_rate": 0.0001996829461211548,
+      "loss": 1.0741,
+      "step": 285
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5230697989463806,
+      "learning_rate": 0.00019968055002407776,
+      "loss": 1.2932,
+      "step": 286
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4365893304347992,
+      "learning_rate": 0.00019967814492142738,
+      "loss": 1.0626,
+      "step": 287
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5909698009490967,
+      "learning_rate": 0.00019967573081342103,
+      "loss": 1.1187,
+      "step": 288
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.48496195673942566,
+      "learning_rate": 0.00019967330770027673,
+      "loss": 1.3553,
+      "step": 289
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.46772319078445435,
+      "learning_rate": 0.0001996708755822134,
+      "loss": 0.9889,
+      "step": 290
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5091903805732727,
+      "learning_rate": 0.00019966843445945088,
+      "loss": 1.1878,
+      "step": 291
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.500097930431366,
+      "learning_rate": 0.00019966598433220957,
+      "loss": 1.107,
+      "step": 292
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3994102478027344,
+      "learning_rate": 0.0001996635252007109,
+      "loss": 0.861,
+      "step": 293
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5007700324058533,
+      "learning_rate": 0.000199661057065177,
+      "loss": 1.0658,
+      "step": 294
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.48105570673942566,
+      "learning_rate": 0.0001996585799258309,
+      "loss": 1.2075,
+      "step": 295
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4781433343887329,
+      "learning_rate": 0.00019965609378289637,
+      "loss": 1.2312,
+      "step": 296
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4458458721637726,
+      "learning_rate": 0.000199653598636598,
+      "loss": 0.9215,
+      "step": 297
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5045272707939148,
+      "learning_rate": 0.00019965109448716125,
+      "loss": 1.2269,
+      "step": 298
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.606421947479248,
+      "learning_rate": 0.0001996485813348123,
+      "loss": 1.074,
+      "step": 299
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4829641580581665,
+      "learning_rate": 0.00019964605917977828,
+      "loss": 1.0225,
+      "step": 300
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5274330377578735,
+      "learning_rate": 0.00019964352802228697,
+      "loss": 1.1151,
+      "step": 301
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5542184710502625,
+      "learning_rate": 0.0001996409878625671,
+      "loss": 1.1317,
+      "step": 302
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5314948558807373,
+      "learning_rate": 0.00019963843870084814,
+      "loss": 1.004,
+      "step": 303
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5083348751068115,
+      "learning_rate": 0.0001996358805373604,
+      "loss": 1.0661,
+      "step": 304
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4306085407733917,
+      "learning_rate": 0.00019963331337233498,
+      "loss": 0.8742,
+      "step": 305
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4597809314727783,
+      "learning_rate": 0.00019963073720600383,
+      "loss": 1.0761,
+      "step": 306
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4719105362892151,
+      "learning_rate": 0.00019962815203859968,
+      "loss": 0.8865,
+      "step": 307
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.45550569891929626,
+      "learning_rate": 0.00019962555787035607,
+      "loss": 0.9979,
+      "step": 308
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4660949409008026,
+      "learning_rate": 0.00019962295470150743,
+      "loss": 1.1004,
+      "step": 309
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.48242247104644775,
+      "learning_rate": 0.00019962034253228885,
+      "loss": 1.0824,
+      "step": 310
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5680294036865234,
+      "learning_rate": 0.0001996177213629364,
+      "loss": 1.2334,
+      "step": 311
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.47598588466644287,
+      "learning_rate": 0.00019961509119368684,
+      "loss": 0.9472,
+      "step": 312
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4807590842247009,
+      "learning_rate": 0.0001996124520247778,
+      "loss": 1.0742,
+      "step": 313
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.45941096544265747,
+      "learning_rate": 0.00019960980385644775,
+      "loss": 1.1298,
+      "step": 314
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.49606096744537354,
+      "learning_rate": 0.0001996071466889359,
+      "loss": 1.1194,
+      "step": 315
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5158990621566772,
+      "learning_rate": 0.00019960448052248233,
+      "loss": 1.0912,
+      "step": 316
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5971901416778564,
+      "learning_rate": 0.00019960180535732788,
+      "loss": 1.1296,
+      "step": 317
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5912845730781555,
+      "learning_rate": 0.00019959912119371426,
+      "loss": 1.1678,
+      "step": 318
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.47174957394599915,
+      "learning_rate": 0.00019959642803188402,
+      "loss": 1.1544,
+      "step": 319
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5198231935501099,
+      "learning_rate": 0.00019959372587208035,
+      "loss": 1.1442,
+      "step": 320
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.49466976523399353,
+      "learning_rate": 0.0001995910147145475,
+      "loss": 1.1183,
+      "step": 321
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5019547343254089,
+      "learning_rate": 0.00019958829455953033,
+      "loss": 1.2528,
+      "step": 322
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4621056914329529,
+      "learning_rate": 0.0001995855654072746,
+      "loss": 0.9278,
+      "step": 323
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4333496689796448,
+      "learning_rate": 0.0001995828272580269,
+      "loss": 1.1729,
+      "step": 324
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.45768311619758606,
+      "learning_rate": 0.0001995800801120346,
+      "loss": 1.1136,
+      "step": 325
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4838506281375885,
+      "learning_rate": 0.0001995773239695459,
+      "loss": 1.17,
+      "step": 326
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5374295115470886,
+      "learning_rate": 0.00019957455883080978,
+      "loss": 1.2547,
+      "step": 327
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4597136378288269,
+      "learning_rate": 0.00019957178469607602,
+      "loss": 1.1141,
+      "step": 328
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4528232514858246,
+      "learning_rate": 0.00019956900156559534,
+      "loss": 1.0884,
+      "step": 329
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4871821403503418,
+      "learning_rate": 0.00019956620943961908,
+      "loss": 1.2074,
+      "step": 330
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4409606456756592,
+      "learning_rate": 0.00019956340831839956,
+      "loss": 0.9346,
+      "step": 331
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.498367041349411,
+      "learning_rate": 0.00019956059820218982,
+      "loss": 1.1141,
+      "step": 332
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5050720572471619,
+      "learning_rate": 0.00019955777909124372,
+      "loss": 1.2472,
+      "step": 333
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5095576047897339,
+      "learning_rate": 0.00019955495098581604,
+      "loss": 0.9069,
+      "step": 334
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5386484265327454,
+      "learning_rate": 0.00019955211388616214,
+      "loss": 1.087,
+      "step": 335
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6390095353126526,
+      "learning_rate": 0.00019954926779253848,
+      "loss": 1.3301,
+      "step": 336
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5223362445831299,
+      "learning_rate": 0.00019954641270520206,
+      "loss": 1.3201,
+      "step": 337
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5296879410743713,
+      "learning_rate": 0.00019954354862441093,
+      "loss": 1.2484,
+      "step": 338
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4566580653190613,
+      "learning_rate": 0.0001995406755504238,
+      "loss": 0.9785,
+      "step": 339
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5283248424530029,
+      "learning_rate": 0.0001995377934835002,
+      "loss": 1.1248,
+      "step": 340
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4475520849227905,
+      "learning_rate": 0.00019953490242390057,
+      "loss": 1.0337,
+      "step": 341
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5099059343338013,
+      "learning_rate": 0.00019953200237188604,
+      "loss": 1.059,
+      "step": 342
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4882862865924835,
+      "learning_rate": 0.00019952909332771865,
+      "loss": 1.1237,
+      "step": 343
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4969572424888611,
+      "learning_rate": 0.0001995261752916612,
+      "loss": 1.194,
+      "step": 344
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4979659914970398,
+      "learning_rate": 0.00019952324826397736,
+      "loss": 1.22,
+      "step": 345
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5434485077857971,
+      "learning_rate": 0.00019952031224493153,
+      "loss": 1.3075,
+      "step": 346
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4933587908744812,
+      "learning_rate": 0.00019951736723478896,
+      "loss": 1.0579,
+      "step": 347
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4828581213951111,
+      "learning_rate": 0.0001995144132338157,
+      "loss": 1.2079,
+      "step": 348
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5049371123313904,
+      "learning_rate": 0.00019951145024227872,
+      "loss": 1.115,
+      "step": 349
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.47797635197639465,
+      "learning_rate": 0.0001995084782604456,
+      "loss": 0.9859,
+      "step": 350
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5296621322631836,
+      "learning_rate": 0.0001995054972885849,
+      "loss": 1.0991,
+      "step": 351
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4989735186100006,
+      "learning_rate": 0.0001995025073269659,
+      "loss": 1.0342,
+      "step": 352
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5142637491226196,
+      "learning_rate": 0.00019949950837585875,
+      "loss": 1.2478,
+      "step": 353
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.48040661215782166,
+      "learning_rate": 0.0001994965004355344,
+      "loss": 1.0811,
+      "step": 354
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4938506484031677,
+      "learning_rate": 0.00019949348350626456,
+      "loss": 1.279,
+      "step": 355
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5298872590065002,
+      "learning_rate": 0.00019949045758832186,
+      "loss": 1.0503,
+      "step": 356
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5765920281410217,
+      "learning_rate": 0.00019948742268197957,
+      "loss": 1.2411,
+      "step": 357
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5472105145454407,
+      "learning_rate": 0.00019948437878751198,
+      "loss": 1.296,
+      "step": 358
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5123063921928406,
+      "learning_rate": 0.00019948132590519402,
+      "loss": 1.0406,
+      "step": 359
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.512622058391571,
+      "learning_rate": 0.00019947826403530155,
+      "loss": 1.0446,
+      "step": 360
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5731431841850281,
+      "learning_rate": 0.00019947519317811113,
+      "loss": 1.1033,
+      "step": 361
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5514195561408997,
+      "learning_rate": 0.00019947211333390028,
+      "loss": 1.2534,
+      "step": 362
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.48427775502204895,
+      "learning_rate": 0.00019946902450294717,
+      "loss": 1.2724,
+      "step": 363
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4817625880241394,
+      "learning_rate": 0.0001994659266855309,
+      "loss": 0.935,
+      "step": 364
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5593578219413757,
+      "learning_rate": 0.00019946281988193133,
+      "loss": 1.1755,
+      "step": 365
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.48790112137794495,
+      "learning_rate": 0.00019945970409242914,
+      "loss": 1.2172,
+      "step": 366
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5469039082527161,
+      "learning_rate": 0.0001994565793173058,
+      "loss": 1.2309,
+      "step": 367
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5210196375846863,
+      "learning_rate": 0.00019945344555684366,
+      "loss": 1.2343,
+      "step": 368
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5936497449874878,
+      "learning_rate": 0.00019945030281132584,
+      "loss": 1.3347,
+      "step": 369
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5588001608848572,
+      "learning_rate": 0.00019944715108103622,
+      "loss": 1.226,
+      "step": 370
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4977017343044281,
+      "learning_rate": 0.0001994439903662596,
+      "loss": 1.1088,
+      "step": 371
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5470244884490967,
+      "learning_rate": 0.00019944082066728145,
+      "loss": 1.2318,
+      "step": 372
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.48215651512145996,
+      "learning_rate": 0.0001994376419843882,
+      "loss": 1.0203,
+      "step": 373
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.473414808511734,
+      "learning_rate": 0.00019943445431786706,
+      "loss": 0.9304,
+      "step": 374
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.47687965631484985,
+      "learning_rate": 0.00019943125766800593,
+      "loss": 1.1128,
+      "step": 375
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5194618105888367,
+      "learning_rate": 0.00019942805203509363,
+      "loss": 1.0648,
+      "step": 376
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6038587689399719,
+      "learning_rate": 0.0001994248374194198,
+      "loss": 0.9894,
+      "step": 377
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6201353073120117,
+      "learning_rate": 0.00019942161382127487,
+      "loss": 1.1983,
+      "step": 378
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5684635639190674,
+      "learning_rate": 0.00019941838124095003,
+      "loss": 1.1072,
+      "step": 379
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.518798828125,
+      "learning_rate": 0.00019941513967873737,
+      "loss": 1.0759,
+      "step": 380
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5163238644599915,
+      "learning_rate": 0.00019941188913492972,
+      "loss": 1.1764,
+      "step": 381
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.47596409916877747,
+      "learning_rate": 0.00019940862960982075,
+      "loss": 1.0686,
+      "step": 382
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5515662431716919,
+      "learning_rate": 0.00019940536110370495,
+      "loss": 1.274,
+      "step": 383
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4495806396007538,
+      "learning_rate": 0.0001994020836168776,
+      "loss": 0.9764,
+      "step": 384
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5222442150115967,
+      "learning_rate": 0.00019939879714963484,
+      "loss": 1.108,
+      "step": 385
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.47913652658462524,
+      "learning_rate": 0.0001993955017022735,
+      "loss": 1.1575,
+      "step": 386
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4735942482948303,
+      "learning_rate": 0.0001993921972750914,
+      "loss": 1.1684,
+      "step": 387
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.44809603691101074,
+      "learning_rate": 0.000199388883868387,
+      "loss": 1.0301,
+      "step": 388
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5486351847648621,
+      "learning_rate": 0.0001993855614824597,
+      "loss": 1.2701,
+      "step": 389
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4721758961677551,
+      "learning_rate": 0.00019938223011760963,
+      "loss": 1.1417,
+      "step": 390
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.561992883682251,
+      "learning_rate": 0.0001993788897741378,
+      "loss": 1.1319,
+      "step": 391
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5158048272132874,
+      "learning_rate": 0.00019937554045234594,
+      "loss": 0.9899,
+      "step": 392
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5148165225982666,
+      "learning_rate": 0.00019937218215253667,
+      "loss": 1.2607,
+      "step": 393
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4346942901611328,
+      "learning_rate": 0.00019936881487501337,
+      "loss": 0.8729,
+      "step": 394
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5264867544174194,
+      "learning_rate": 0.0001993654386200803,
+      "loss": 1.0844,
+      "step": 395
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5341372489929199,
+      "learning_rate": 0.0001993620533880424,
+      "loss": 1.055,
+      "step": 396
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5435947775840759,
+      "learning_rate": 0.0001993586591792056,
+      "loss": 1.109,
+      "step": 397
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4812977612018585,
+      "learning_rate": 0.0001993552559938765,
+      "loss": 1.1092,
+      "step": 398
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5616835951805115,
+      "learning_rate": 0.00019935184383236258,
+      "loss": 1.1535,
+      "step": 399
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4366762638092041,
+      "learning_rate": 0.00019934842269497209,
+      "loss": 0.9751,
+      "step": 400
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5996217131614685,
+      "learning_rate": 0.00019934499258201415,
+      "loss": 1.2012,
+      "step": 401
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.49099141359329224,
+      "learning_rate": 0.00019934155349379858,
+      "loss": 1.0873,
+      "step": 402
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4674723744392395,
+      "learning_rate": 0.00019933810543063617,
+      "loss": 1.1367,
+      "step": 403
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5626590847969055,
+      "learning_rate": 0.00019933464839283832,
+      "loss": 1.3648,
+      "step": 404
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5492469072341919,
+      "learning_rate": 0.00019933118238071743,
+      "loss": 1.249,
+      "step": 405
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5024725198745728,
+      "learning_rate": 0.00019932770739458668,
+      "loss": 1.0049,
+      "step": 406
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.48954451084136963,
+      "learning_rate": 0.00019932422343475992,
+      "loss": 1.1437,
+      "step": 407
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5946091413497925,
+      "learning_rate": 0.00019932073050155194,
+      "loss": 1.2738,
+      "step": 408
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.504777729511261,
+      "learning_rate": 0.00019931722859527833,
+      "loss": 1.2256,
+      "step": 409
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5623805522918701,
+      "learning_rate": 0.00019931371771625544,
+      "loss": 1.0886,
+      "step": 410
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.545048177242279,
+      "learning_rate": 0.00019931019786480046,
+      "loss": 1.0154,
+      "step": 411
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4837050139904022,
+      "learning_rate": 0.00019930666904123138,
+      "loss": 1.1486,
+      "step": 412
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5394037365913391,
+      "learning_rate": 0.00019930313124586705,
+      "loss": 1.211,
+      "step": 413
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.45422598719596863,
+      "learning_rate": 0.00019929958447902708,
+      "loss": 1.1847,
+      "step": 414
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5187952518463135,
+      "learning_rate": 0.00019929602874103186,
+      "loss": 1.1617,
+      "step": 415
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5281522870063782,
+      "learning_rate": 0.00019929246403220267,
+      "loss": 1.1458,
+      "step": 416
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5519776344299316,
+      "learning_rate": 0.00019928889035286153,
+      "loss": 1.0696,
+      "step": 417
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4418583810329437,
+      "learning_rate": 0.00019928530770333135,
+      "loss": 0.8771,
+      "step": 418
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4728032648563385,
+      "learning_rate": 0.00019928171608393574,
+      "loss": 0.9559,
+      "step": 419
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.49322378635406494,
+      "learning_rate": 0.0001992781154949992,
+      "loss": 1.1025,
+      "step": 420
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5575322508811951,
+      "learning_rate": 0.00019927450593684707,
+      "loss": 1.2714,
+      "step": 421
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.46672430634498596,
+      "learning_rate": 0.0001992708874098054,
+      "loss": 1.1076,
+      "step": 422
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5349050164222717,
+      "learning_rate": 0.00019926725991420115,
+      "loss": 1.1568,
+      "step": 423
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6394595503807068,
+      "learning_rate": 0.00019926362345036198,
+      "loss": 1.1381,
+      "step": 424
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5520142316818237,
+      "learning_rate": 0.00019925997801861648,
+      "loss": 0.9789,
+      "step": 425
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.48990902304649353,
+      "learning_rate": 0.000199256323619294,
+      "loss": 1.0684,
+      "step": 426
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4744899570941925,
+      "learning_rate": 0.00019925266025272461,
+      "loss": 1.0202,
+      "step": 427
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5531217455863953,
+      "learning_rate": 0.00019924898791923935,
+      "loss": 1.135,
+      "step": 428
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4650912880897522,
+      "learning_rate": 0.00019924530661917,
+      "loss": 1.3101,
+      "step": 429
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4258342385292053,
+      "learning_rate": 0.00019924161635284912,
+      "loss": 1.0714,
+      "step": 430
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4997932016849518,
+      "learning_rate": 0.00019923791712061012,
+      "loss": 1.1918,
+      "step": 431
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4654412269592285,
+      "learning_rate": 0.00019923420892278718,
+      "loss": 0.8791,
+      "step": 432
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5107722282409668,
+      "learning_rate": 0.0001992304917597153,
+      "loss": 1.199,
+      "step": 433
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5670316815376282,
+      "learning_rate": 0.00019922676563173037,
+      "loss": 1.1498,
+      "step": 434
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5506672263145447,
+      "learning_rate": 0.00019922303053916896,
+      "loss": 1.0911,
+      "step": 435
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5242930054664612,
+      "learning_rate": 0.00019921928648236853,
+      "loss": 1.0317,
+      "step": 436
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5948449373245239,
+      "learning_rate": 0.00019921553346166737,
+      "loss": 1.2627,
+      "step": 437
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5378965139389038,
+      "learning_rate": 0.0001992117714774045,
+      "loss": 1.1651,
+      "step": 438
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.614194929599762,
+      "learning_rate": 0.00019920800052991984,
+      "loss": 1.1672,
+      "step": 439
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5187473893165588,
+      "learning_rate": 0.00019920422061955404,
+      "loss": 1.1762,
+      "step": 440
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.47316837310791016,
+      "learning_rate": 0.00019920043174664857,
+      "loss": 1.0654,
+      "step": 441
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5558268427848816,
+      "learning_rate": 0.00019919663391154577,
+      "loss": 1.0374,
+      "step": 442
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5064995884895325,
+      "learning_rate": 0.0001991928271145888,
+      "loss": 1.205,
+      "step": 443
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5428379774093628,
+      "learning_rate": 0.00019918901135612145,
+      "loss": 1.0972,
+      "step": 444
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.500322163105011,
+      "learning_rate": 0.00019918518663648857,
+      "loss": 1.1213,
+      "step": 445
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.47069665789604187,
+      "learning_rate": 0.00019918135295603567,
+      "loss": 1.0122,
+      "step": 446
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5628748536109924,
+      "learning_rate": 0.00019917751031510908,
+      "loss": 1.1794,
+      "step": 447
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.47065627574920654,
+      "learning_rate": 0.00019917365871405598,
+      "loss": 1.0724,
+      "step": 448
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5735258460044861,
+      "learning_rate": 0.00019916979815322433,
+      "loss": 1.0173,
+      "step": 449
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6087215542793274,
+      "learning_rate": 0.00019916592863296293,
+      "loss": 1.1557,
+      "step": 450
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4948890209197998,
+      "learning_rate": 0.00019916205015362132,
+      "loss": 1.1302,
+      "step": 451
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5734955668449402,
+      "learning_rate": 0.00019915816271554997,
+      "loss": 1.0741,
+      "step": 452
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5332045555114746,
+      "learning_rate": 0.00019915426631910004,
+      "loss": 1.1055,
+      "step": 453
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5224483609199524,
+      "learning_rate": 0.00019915036096462355,
+      "loss": 1.1148,
+      "step": 454
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5145097374916077,
+      "learning_rate": 0.00019914644665247332,
+      "loss": 1.1198,
+      "step": 455
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5348233580589294,
+      "learning_rate": 0.000199142523383003,
+      "loss": 1.2597,
+      "step": 456
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5970399975776672,
+      "learning_rate": 0.00019913859115656707,
+      "loss": 1.3203,
+      "step": 457
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.514270544052124,
+      "learning_rate": 0.00019913464997352073,
+      "loss": 1.2183,
+      "step": 458
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5271024107933044,
+      "learning_rate": 0.0001991306998342201,
+      "loss": 0.9667,
+      "step": 459
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.507999837398529,
+      "learning_rate": 0.00019912674073902195,
+      "loss": 1.1014,
+      "step": 460
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.48636719584465027,
+      "learning_rate": 0.00019912277268828408,
+      "loss": 1.1854,
+      "step": 461
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5034084916114807,
+      "learning_rate": 0.0001991187956823649,
+      "loss": 0.9449,
+      "step": 462
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5856225490570068,
+      "learning_rate": 0.00019911480972162378,
+      "loss": 1.2222,
+      "step": 463
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.516998827457428,
+      "learning_rate": 0.00019911081480642072,
+      "loss": 1.0252,
+      "step": 464
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4869126081466675,
+      "learning_rate": 0.00019910681093711675,
+      "loss": 1.1473,
+      "step": 465
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5495402812957764,
+      "learning_rate": 0.00019910279811407357,
+      "loss": 1.131,
+      "step": 466
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4950717091560364,
+      "learning_rate": 0.00019909877633765368,
+      "loss": 1.0872,
+      "step": 467
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.609627366065979,
+      "learning_rate": 0.00019909474560822043,
+      "loss": 1.1313,
+      "step": 468
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4924461841583252,
+      "learning_rate": 0.00019909070592613799,
+      "loss": 1.1424,
+      "step": 469
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.48707911372184753,
+      "learning_rate": 0.00019908665729177134,
+      "loss": 1.1168,
+      "step": 470
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5481781363487244,
+      "learning_rate": 0.00019908259970548623,
+      "loss": 1.2636,
+      "step": 471
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4931468963623047,
+      "learning_rate": 0.0001990785331676492,
+      "loss": 0.9433,
+      "step": 472
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5407423377037048,
+      "learning_rate": 0.00019907445767862773,
+      "loss": 1.0628,
+      "step": 473
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4662988483905792,
+      "learning_rate": 0.00019907037323878995,
+      "loss": 1.0792,
+      "step": 474
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4906497299671173,
+      "learning_rate": 0.00019906627984850488,
+      "loss": 1.0281,
+      "step": 475
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4651735723018646,
+      "learning_rate": 0.00019906217750814234,
+      "loss": 0.9659,
+      "step": 476
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.485735684633255,
+      "learning_rate": 0.00019905806621807294,
+      "loss": 1.1931,
+      "step": 477
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4991046190261841,
+      "learning_rate": 0.00019905394597866813,
+      "loss": 1.2245,
+      "step": 478
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5731810927391052,
+      "learning_rate": 0.00019904981679030016,
+      "loss": 1.185,
+      "step": 479
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5500648021697998,
+      "learning_rate": 0.00019904567865334204,
+      "loss": 1.0929,
+      "step": 480
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.47128331661224365,
+      "learning_rate": 0.00019904153156816764,
+      "loss": 0.7726,
+      "step": 481
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5505924820899963,
+      "learning_rate": 0.00019903737553515164,
+      "loss": 1.1731,
+      "step": 482
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5098860263824463,
+      "learning_rate": 0.00019903321055466953,
+      "loss": 1.0776,
+      "step": 483
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.54964280128479,
+      "learning_rate": 0.00019902903662709755,
+      "loss": 1.1415,
+      "step": 484
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5043061971664429,
+      "learning_rate": 0.00019902485375281281,
+      "loss": 1.0319,
+      "step": 485
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5406997203826904,
+      "learning_rate": 0.00019902066193219322,
+      "loss": 1.1259,
+      "step": 486
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6080590486526489,
+      "learning_rate": 0.0001990164611656175,
+      "loss": 1.0707,
+      "step": 487
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.46147873997688293,
+      "learning_rate": 0.0001990122514534651,
+      "loss": 0.8235,
+      "step": 488
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5430011749267578,
+      "learning_rate": 0.0001990080327961164,
+      "loss": 1.1288,
+      "step": 489
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4875696003437042,
+      "learning_rate": 0.00019900380519395253,
+      "loss": 1.0249,
+      "step": 490
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.46323084831237793,
+      "learning_rate": 0.00019899956864735544,
+      "loss": 0.9812,
+      "step": 491
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4611752927303314,
+      "learning_rate": 0.00019899532315670782,
+      "loss": 0.8467,
+      "step": 492
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5283900499343872,
+      "learning_rate": 0.00019899106872239326,
+      "loss": 1.2849,
+      "step": 493
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.47522106766700745,
+      "learning_rate": 0.00019898680534479618,
+      "loss": 0.9322,
+      "step": 494
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4241219758987427,
+      "learning_rate": 0.00019898253302430164,
+      "loss": 1.1607,
+      "step": 495
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5026375651359558,
+      "learning_rate": 0.0001989782517612957,
+      "loss": 1.1405,
+      "step": 496
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5204796195030212,
+      "learning_rate": 0.00019897396155616512,
+      "loss": 1.1251,
+      "step": 497
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.612561821937561,
+      "learning_rate": 0.0001989696624092975,
+      "loss": 0.9414,
+      "step": 498
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6452274322509766,
+      "learning_rate": 0.00019896535432108127,
+      "loss": 1.1394,
+      "step": 499
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5177410244941711,
+      "learning_rate": 0.00019896103729190562,
+      "loss": 1.1647,
+      "step": 500
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5609557628631592,
+      "learning_rate": 0.0001989567113221605,
+      "loss": 1.1083,
+      "step": 501
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4865688383579254,
+      "learning_rate": 0.00019895237641223687,
+      "loss": 0.8964,
+      "step": 502
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5484577417373657,
+      "learning_rate": 0.00019894803256252628,
+      "loss": 1.1321,
+      "step": 503
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5816463232040405,
+      "learning_rate": 0.0001989436797734212,
+      "loss": 1.1356,
+      "step": 504
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5487164855003357,
+      "learning_rate": 0.00019893931804531488,
+      "loss": 1.3929,
+      "step": 505
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5809823870658875,
+      "learning_rate": 0.00019893494737860135,
+      "loss": 1.2339,
+      "step": 506
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.601434051990509,
+      "learning_rate": 0.00019893056777367548,
+      "loss": 1.1482,
+      "step": 507
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.48192596435546875,
+      "learning_rate": 0.00019892617923093298,
+      "loss": 1.0692,
+      "step": 508
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6076588034629822,
+      "learning_rate": 0.00019892178175077028,
+      "loss": 1.229,
+      "step": 509
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5927091836929321,
+      "learning_rate": 0.0001989173753335847,
+      "loss": 1.0472,
+      "step": 510
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.48782506585121155,
+      "learning_rate": 0.00019891295997977434,
+      "loss": 1.1462,
+      "step": 511
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.48741236329078674,
+      "learning_rate": 0.00019890853568973807,
+      "loss": 1.1603,
+      "step": 512
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5580458045005798,
+      "learning_rate": 0.00019890410246387562,
+      "loss": 1.1697,
+      "step": 513
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5367544889450073,
+      "learning_rate": 0.00019889966030258753,
+      "loss": 1.118,
+      "step": 514
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4435345232486725,
+      "learning_rate": 0.00019889520920627506,
+      "loss": 1.0305,
+      "step": 515
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5683806538581848,
+      "learning_rate": 0.0001988907491753404,
+      "loss": 1.0101,
+      "step": 516
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5692442059516907,
+      "learning_rate": 0.00019888628021018652,
+      "loss": 1.0924,
+      "step": 517
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5492920279502869,
+      "learning_rate": 0.00019888180231121707,
+      "loss": 1.1479,
+      "step": 518
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5364216566085815,
+      "learning_rate": 0.00019887731547883663,
+      "loss": 1.0438,
+      "step": 519
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4900088310241699,
+      "learning_rate": 0.0001988728197134506,
+      "loss": 1.1337,
+      "step": 520
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5172410011291504,
+      "learning_rate": 0.00019886831501546512,
+      "loss": 1.2197,
+      "step": 521
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5062763690948486,
+      "learning_rate": 0.00019886380138528717,
+      "loss": 1.1406,
+      "step": 522
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4935275912284851,
+      "learning_rate": 0.00019885927882332455,
+      "loss": 1.1764,
+      "step": 523
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5174946188926697,
+      "learning_rate": 0.0001988547473299858,
+      "loss": 1.131,
+      "step": 524
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.45522478222846985,
+      "learning_rate": 0.00019885020690568034,
+      "loss": 0.9077,
+      "step": 525
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5257841944694519,
+      "learning_rate": 0.0001988456575508184,
+      "loss": 0.9448,
+      "step": 526
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.484298974275589,
+      "learning_rate": 0.00019884109926581096,
+      "loss": 1.1791,
+      "step": 527
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.536929190158844,
+      "learning_rate": 0.00019883653205106985,
+      "loss": 1.2124,
+      "step": 528
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5551151037216187,
+      "learning_rate": 0.00019883195590700764,
+      "loss": 1.1383,
+      "step": 529
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.49238866567611694,
+      "learning_rate": 0.00019882737083403785,
+      "loss": 1.0841,
+      "step": 530
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.49737808108329773,
+      "learning_rate": 0.0001988227768325746,
+      "loss": 1.0843,
+      "step": 531
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.529952883720398,
+      "learning_rate": 0.00019881817390303303,
+      "loss": 1.1796,
+      "step": 532
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5074145197868347,
+      "learning_rate": 0.00019881356204582895,
+      "loss": 1.1394,
+      "step": 533
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.553284227848053,
+      "learning_rate": 0.00019880894126137903,
+      "loss": 1.1341,
+      "step": 534
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4974454641342163,
+      "learning_rate": 0.00019880431155010073,
+      "loss": 1.0177,
+      "step": 535
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5199590921401978,
+      "learning_rate": 0.00019879967291241228,
+      "loss": 1.2168,
+      "step": 536
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5003868341445923,
+      "learning_rate": 0.00019879502534873283,
+      "loss": 1.0574,
+      "step": 537
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.49978846311569214,
+      "learning_rate": 0.00019879036885948217,
+      "loss": 1.268,
+      "step": 538
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5444263219833374,
+      "learning_rate": 0.00019878570344508106,
+      "loss": 1.147,
+      "step": 539
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.540688693523407,
+      "learning_rate": 0.00019878102910595095,
+      "loss": 1.2374,
+      "step": 540
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4839583933353424,
+      "learning_rate": 0.00019877634584251417,
+      "loss": 0.9215,
+      "step": 541
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5708199143409729,
+      "learning_rate": 0.00019877165365519379,
+      "loss": 1.2743,
+      "step": 542
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.496097594499588,
+      "learning_rate": 0.00019876695254441378,
+      "loss": 1.1455,
+      "step": 543
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4960477948188782,
+      "learning_rate": 0.0001987622425105988,
+      "loss": 1.0126,
+      "step": 544
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5258330702781677,
+      "learning_rate": 0.00019875752355417441,
+      "loss": 1.1997,
+      "step": 545
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5457305312156677,
+      "learning_rate": 0.00019875279567556692,
+      "loss": 1.2039,
+      "step": 546
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5998676419258118,
+      "learning_rate": 0.0001987480588752035,
+      "loss": 0.9432,
+      "step": 547
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5208805799484253,
+      "learning_rate": 0.00019874331315351205,
+      "loss": 1.0473,
+      "step": 548
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6455830335617065,
+      "learning_rate": 0.00019873855851092133,
+      "loss": 1.0942,
+      "step": 549
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5795512199401855,
+      "learning_rate": 0.00019873379494786094,
+      "loss": 1.1694,
+      "step": 550
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5360243320465088,
+      "learning_rate": 0.0001987290224647612,
+      "loss": 0.9945,
+      "step": 551
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.48501020669937134,
+      "learning_rate": 0.00019872424106205324,
+      "loss": 1.1037,
+      "step": 552
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5583041310310364,
+      "learning_rate": 0.0001987194507401691,
+      "loss": 0.9302,
+      "step": 553
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.539480447769165,
+      "learning_rate": 0.00019871465149954155,
+      "loss": 1.0732,
+      "step": 554
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4358266294002533,
+      "learning_rate": 0.00019870984334060415,
+      "loss": 0.8591,
+      "step": 555
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5443989634513855,
+      "learning_rate": 0.00019870502626379127,
+      "loss": 1.0836,
+      "step": 556
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5532322525978088,
+      "learning_rate": 0.00019870020026953818,
+      "loss": 1.0821,
+      "step": 557
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5223169922828674,
+      "learning_rate": 0.00019869536535828083,
+      "loss": 1.0802,
+      "step": 558
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4703574478626251,
+      "learning_rate": 0.00019869052153045598,
+      "loss": 0.9764,
+      "step": 559
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.542907178401947,
+      "learning_rate": 0.00019868566878650134,
+      "loss": 1.2281,
+      "step": 560
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5914562344551086,
+      "learning_rate": 0.00019868080712685528,
+      "loss": 1.2142,
+      "step": 561
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5440067052841187,
+      "learning_rate": 0.000198675936551957,
+      "loss": 1.177,
+      "step": 562
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4753040075302124,
+      "learning_rate": 0.0001986710570622466,
+      "loss": 1.1859,
+      "step": 563
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4693201780319214,
+      "learning_rate": 0.00019866616865816484,
+      "loss": 0.9495,
+      "step": 564
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.48725131154060364,
+      "learning_rate": 0.00019866127134015335,
+      "loss": 1.1822,
+      "step": 565
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4840669631958008,
+      "learning_rate": 0.00019865636510865467,
+      "loss": 1.0614,
+      "step": 566
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4818721413612366,
+      "learning_rate": 0.00019865144996411197,
+      "loss": 1.0566,
+      "step": 567
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5293144583702087,
+      "learning_rate": 0.00019864652590696934,
+      "loss": 1.1449,
+      "step": 568
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5687733292579651,
+      "learning_rate": 0.00019864159293767162,
+      "loss": 1.2512,
+      "step": 569
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5350620150566101,
+      "learning_rate": 0.00019863665105666447,
+      "loss": 1.176,
+      "step": 570
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5288693904876709,
+      "learning_rate": 0.0001986317002643944,
+      "loss": 0.9136,
+      "step": 571
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4852888584136963,
+      "learning_rate": 0.00019862674056130865,
+      "loss": 1.0506,
+      "step": 572
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5876291394233704,
+      "learning_rate": 0.0001986217719478553,
+      "loss": 1.2772,
+      "step": 573
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5367861390113831,
+      "learning_rate": 0.00019861679442448327,
+      "loss": 1.1672,
+      "step": 574
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5490912199020386,
+      "learning_rate": 0.00019861180799164222,
+      "loss": 1.0972,
+      "step": 575
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5793784260749817,
+      "learning_rate": 0.0001986068126497827,
+      "loss": 1.1316,
+      "step": 576
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.502665638923645,
+      "learning_rate": 0.0001986018083993559,
+      "loss": 1.1683,
+      "step": 577
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5572852492332458,
+      "learning_rate": 0.00019859679524081406,
+      "loss": 1.3337,
+      "step": 578
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.49824976921081543,
+      "learning_rate": 0.00019859177317460994,
+      "loss": 1.2149,
+      "step": 579
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5626625418663025,
+      "learning_rate": 0.00019858674220119742,
+      "loss": 1.1765,
+      "step": 580
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5086746215820312,
+      "learning_rate": 0.00019858170232103087,
+      "loss": 1.293,
+      "step": 581
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5596989989280701,
+      "learning_rate": 0.00019857665353456577,
+      "loss": 1.084,
+      "step": 582
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5050382018089294,
+      "learning_rate": 0.00019857159584225808,
+      "loss": 1.0266,
+      "step": 583
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4970535635948181,
+      "learning_rate": 0.00019856652924456488,
+      "loss": 1.2084,
+      "step": 584
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5188227891921997,
+      "learning_rate": 0.0001985614537419438,
+      "loss": 1.2503,
+      "step": 585
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5552204251289368,
+      "learning_rate": 0.00019855636933485343,
+      "loss": 1.2152,
+      "step": 586
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7056576013565063,
+      "learning_rate": 0.0001985512760237531,
+      "loss": 1.0611,
+      "step": 587
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.499446839094162,
+      "learning_rate": 0.00019854617380910303,
+      "loss": 1.0997,
+      "step": 588
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6023534536361694,
+      "learning_rate": 0.0001985410626913641,
+      "loss": 1.102,
+      "step": 589
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5384407043457031,
+      "learning_rate": 0.0001985359426709981,
+      "loss": 1.2433,
+      "step": 590
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5116797685623169,
+      "learning_rate": 0.00019853081374846755,
+      "loss": 0.9768,
+      "step": 591
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5307533144950867,
+      "learning_rate": 0.0001985256759242359,
+      "loss": 1.1518,
+      "step": 592
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4904177784919739,
+      "learning_rate": 0.00019852052919876726,
+      "loss": 1.0781,
+      "step": 593
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.48693302273750305,
+      "learning_rate": 0.00019851537357252667,
+      "loss": 1.1335,
+      "step": 594
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.557604193687439,
+      "learning_rate": 0.00019851020904597982,
+      "loss": 1.0688,
+      "step": 595
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5276851058006287,
+      "learning_rate": 0.00019850503561959335,
+      "loss": 1.0065,
+      "step": 596
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.46584829688072205,
+      "learning_rate": 0.00019849985329383466,
+      "loss": 1.0136,
+      "step": 597
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5401173830032349,
+      "learning_rate": 0.00019849466206917193,
+      "loss": 1.1503,
+      "step": 598
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5231667160987854,
+      "learning_rate": 0.00019848946194607416,
+      "loss": 1.079,
+      "step": 599
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.46920567750930786,
+      "learning_rate": 0.00019848425292501114,
+      "loss": 1.0505,
+      "step": 600
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4620935022830963,
+      "learning_rate": 0.00019847903500645351,
+      "loss": 1.1187,
+      "step": 601
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5359715223312378,
+      "learning_rate": 0.00019847380819087263,
+      "loss": 0.9545,
+      "step": 602
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5216434001922607,
+      "learning_rate": 0.00019846857247874072,
+      "loss": 0.958,
+      "step": 603
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4918813109397888,
+      "learning_rate": 0.00019846332787053084,
+      "loss": 1.1468,
+      "step": 604
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5258176922798157,
+      "learning_rate": 0.0001984580743667168,
+      "loss": 0.9893,
+      "step": 605
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5433478355407715,
+      "learning_rate": 0.00019845281196777319,
+      "loss": 1.365,
+      "step": 606
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5165151357650757,
+      "learning_rate": 0.00019844754067417544,
+      "loss": 0.9692,
+      "step": 607
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.562602698802948,
+      "learning_rate": 0.00019844226048639982,
+      "loss": 1.0073,
+      "step": 608
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.47087547183036804,
+      "learning_rate": 0.00019843697140492332,
+      "loss": 1.178,
+      "step": 609
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.633965790271759,
+      "learning_rate": 0.00019843167343022383,
+      "loss": 1.0784,
+      "step": 610
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5291125774383545,
+      "learning_rate": 0.00019842636656277994,
+      "loss": 1.164,
+      "step": 611
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.46963217854499817,
+      "learning_rate": 0.00019842105080307117,
+      "loss": 1.1386,
+      "step": 612
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4947594404220581,
+      "learning_rate": 0.0001984157261515777,
+      "loss": 1.1135,
+      "step": 613
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.47310054302215576,
+      "learning_rate": 0.00019841039260878055,
+      "loss": 1.0398,
+      "step": 614
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5538829565048218,
+      "learning_rate": 0.00019840505017516166,
+      "loss": 1.31,
+      "step": 615
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5245307683944702,
+      "learning_rate": 0.00019839969885120367,
+      "loss": 1.1665,
+      "step": 616
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.500907301902771,
+      "learning_rate": 0.00019839433863739,
+      "loss": 1.0136,
+      "step": 617
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5003906488418579,
+      "learning_rate": 0.00019838896953420493,
+      "loss": 1.2467,
+      "step": 618
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.48780718445777893,
+      "learning_rate": 0.00019838359154213358,
+      "loss": 0.8645,
+      "step": 619
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.520158588886261,
+      "learning_rate": 0.00019837820466166175,
+      "loss": 0.8612,
+      "step": 620
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5978065133094788,
+      "learning_rate": 0.0001983728088932762,
+      "loss": 1.1102,
+      "step": 621
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5158056020736694,
+      "learning_rate": 0.00019836740423746427,
+      "loss": 1.0781,
+      "step": 622
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5026460886001587,
+      "learning_rate": 0.00019836199069471437,
+      "loss": 1.0797,
+      "step": 623
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5634730458259583,
+      "learning_rate": 0.00019835656826551552,
+      "loss": 0.9508,
+      "step": 624
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6410563588142395,
+      "learning_rate": 0.00019835113695035764,
+      "loss": 0.931,
+      "step": 625
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.539527952671051,
+      "learning_rate": 0.00019834569674973137,
+      "loss": 1.0981,
+      "step": 626
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6172979474067688,
+      "learning_rate": 0.00019834024766412825,
+      "loss": 1.131,
+      "step": 627
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5615272521972656,
+      "learning_rate": 0.00019833478969404054,
+      "loss": 1.246,
+      "step": 628
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6126092076301575,
+      "learning_rate": 0.00019832932283996138,
+      "loss": 1.0175,
+      "step": 629
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4889993369579315,
+      "learning_rate": 0.00019832384710238464,
+      "loss": 1.0162,
+      "step": 630
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.49586522579193115,
+      "learning_rate": 0.00019831836248180497,
+      "loss": 1.0558,
+      "step": 631
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5401427745819092,
+      "learning_rate": 0.00019831286897871797,
+      "loss": 1.1391,
+      "step": 632
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6986876130104065,
+      "learning_rate": 0.00019830736659361988,
+      "loss": 1.2645,
+      "step": 633
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.49258509278297424,
+      "learning_rate": 0.00019830185532700788,
+      "loss": 1.0886,
+      "step": 634
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4943321943283081,
+      "learning_rate": 0.0001982963351793798,
+      "loss": 1.2191,
+      "step": 635
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.46854934096336365,
+      "learning_rate": 0.0001982908061512344,
+      "loss": 1.0435,
+      "step": 636
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5639446377754211,
+      "learning_rate": 0.00019828526824307115,
+      "loss": 1.132,
+      "step": 637
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5033563375473022,
+      "learning_rate": 0.00019827972145539043,
+      "loss": 1.1469,
+      "step": 638
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5321240425109863,
+      "learning_rate": 0.00019827416578869334,
+      "loss": 1.2356,
+      "step": 639
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5140905380249023,
+      "learning_rate": 0.0001982686012434818,
+      "loss": 1.0041,
+      "step": 640
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5813565850257874,
+      "learning_rate": 0.00019826302782025853,
+      "loss": 1.2285,
+      "step": 641
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4850314259529114,
+      "learning_rate": 0.00019825744551952706,
+      "loss": 1.1132,
+      "step": 642
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5613282322883606,
+      "learning_rate": 0.0001982518543417917,
+      "loss": 1.1257,
+      "step": 643
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5611051917076111,
+      "learning_rate": 0.0001982462542875576,
+      "loss": 1.1526,
+      "step": 644
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.48500093817710876,
+      "learning_rate": 0.00019824064535733073,
+      "loss": 1.0102,
+      "step": 645
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5718590617179871,
+      "learning_rate": 0.00019823502755161778,
+      "loss": 1.2613,
+      "step": 646
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5485739707946777,
+      "learning_rate": 0.00019822940087092629,
+      "loss": 0.9491,
+      "step": 647
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4833703339099884,
+      "learning_rate": 0.0001982237653157646,
+      "loss": 0.9716,
+      "step": 648
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5101345777511597,
+      "learning_rate": 0.00019821812088664185,
+      "loss": 1.0698,
+      "step": 649
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5286256074905396,
+      "learning_rate": 0.000198212467584068,
+      "loss": 1.019,
+      "step": 650
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5590866208076477,
+      "learning_rate": 0.00019820680540855377,
+      "loss": 1.2636,
+      "step": 651
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5393713712692261,
+      "learning_rate": 0.00019820113436061073,
+      "loss": 1.2037,
+      "step": 652
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.495000958442688,
+      "learning_rate": 0.0001981954544407512,
+      "loss": 1.0951,
+      "step": 653
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5339246988296509,
+      "learning_rate": 0.00019818976564948835,
+      "loss": 1.0815,
+      "step": 654
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5487880110740662,
+      "learning_rate": 0.00019818406798733612,
+      "loss": 1.1396,
+      "step": 655
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 1.0153948068618774,
+      "eval_runtime": 17.9627,
+      "eval_samples_per_second": 151.926,
+      "eval_steps_per_second": 38.023,
+      "step": 655
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.52154541015625,
+      "learning_rate": 0.00019817836145480925,
+      "loss": 1.1395,
+      "step": 656
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4636710286140442,
+      "learning_rate": 0.00019817264605242332,
+      "loss": 1.071,
+      "step": 657
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5892826914787292,
+      "learning_rate": 0.0001981669217806947,
+      "loss": 1.0462,
+      "step": 658
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5976261496543884,
+      "learning_rate": 0.00019816118864014048,
+      "loss": 1.2249,
+      "step": 659
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6425426602363586,
+      "learning_rate": 0.00019815544663127866,
+      "loss": 1.2933,
+      "step": 660
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6284853219985962,
+      "learning_rate": 0.000198149695754628,
+      "loss": 0.9808,
+      "step": 661
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5053337216377258,
+      "learning_rate": 0.00019814393601070803,
+      "loss": 0.9719,
+      "step": 662
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5163286924362183,
+      "learning_rate": 0.00019813816740003914,
+      "loss": 1.1396,
+      "step": 663
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.54429692029953,
+      "learning_rate": 0.00019813238992314253,
+      "loss": 1.0073,
+      "step": 664
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5420398116111755,
+      "learning_rate": 0.00019812660358054006,
+      "loss": 1.1474,
+      "step": 665
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5848464369773865,
+      "learning_rate": 0.00019812080837275454,
+      "loss": 1.0693,
+      "step": 666
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5412850379943848,
+      "learning_rate": 0.00019811500430030958,
+      "loss": 1.2033,
+      "step": 667
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5231608152389526,
+      "learning_rate": 0.0001981091913637295,
+      "loss": 0.9086,
+      "step": 668
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.517072319984436,
+      "learning_rate": 0.00019810336956353942,
+      "loss": 1.1337,
+      "step": 669
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6114885210990906,
+      "learning_rate": 0.00019809753890026542,
+      "loss": 1.1743,
+      "step": 670
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.48399075865745544,
+      "learning_rate": 0.0001980916993744342,
+      "loss": 0.9902,
+      "step": 671
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5298494696617126,
+      "learning_rate": 0.00019808585098657333,
+      "loss": 1.0834,
+      "step": 672
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5229631662368774,
+      "learning_rate": 0.00019807999373721118,
+      "loss": 1.1433,
+      "step": 673
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5524067282676697,
+      "learning_rate": 0.0001980741276268769,
+      "loss": 1.1291,
+      "step": 674
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4828319847583771,
+      "learning_rate": 0.00019806825265610052,
+      "loss": 1.0705,
+      "step": 675
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5148788690567017,
+      "learning_rate": 0.00019806236882541275,
+      "loss": 1.0661,
+      "step": 676
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5624847412109375,
+      "learning_rate": 0.0001980564761353452,
+      "loss": 1.2352,
+      "step": 677
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4777878522872925,
+      "learning_rate": 0.00019805057458643024,
+      "loss": 1.0932,
+      "step": 678
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5480254888534546,
+      "learning_rate": 0.000198044664179201,
+      "loss": 1.1319,
+      "step": 679
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5098835229873657,
+      "learning_rate": 0.00019803874491419146,
+      "loss": 1.0556,
+      "step": 680
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5364381074905396,
+      "learning_rate": 0.00019803281679193645,
+      "loss": 1.2167,
+      "step": 681
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5941902995109558,
+      "learning_rate": 0.0001980268798129715,
+      "loss": 1.4031,
+      "step": 682
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.47746074199676514,
+      "learning_rate": 0.00019802093397783296,
+      "loss": 0.9778,
+      "step": 683
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5172235369682312,
+      "learning_rate": 0.00019801497928705804,
+      "loss": 1.1598,
+      "step": 684
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6679323315620422,
+      "learning_rate": 0.00019800901574118472,
+      "loss": 1.2064,
+      "step": 685
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5311746597290039,
+      "learning_rate": 0.00019800304334075173,
+      "loss": 1.1862,
+      "step": 686
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5657793283462524,
+      "learning_rate": 0.00019799706208629866,
+      "loss": 1.1143,
+      "step": 687
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.47132763266563416,
+      "learning_rate": 0.00019799107197836593,
+      "loss": 1.1225,
+      "step": 688
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5872971415519714,
+      "learning_rate": 0.00019798507301749463,
+      "loss": 1.1489,
+      "step": 689
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.511029839515686,
+      "learning_rate": 0.00019797906520422677,
+      "loss": 1.0488,
+      "step": 690
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5370973944664001,
+      "learning_rate": 0.00019797304853910513,
+      "loss": 1.1974,
+      "step": 691
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5594046711921692,
+      "learning_rate": 0.0001979670230226733,
+      "loss": 1.0964,
+      "step": 692
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6002780795097351,
+      "learning_rate": 0.00019796098865547564,
+      "loss": 1.0355,
+      "step": 693
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5207370519638062,
+      "learning_rate": 0.00019795494543805727,
+      "loss": 1.0801,
+      "step": 694
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6425591707229614,
+      "learning_rate": 0.0001979488933709642,
+      "loss": 1.0904,
+      "step": 695
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5756493210792542,
+      "learning_rate": 0.0001979428324547432,
+      "loss": 1.0362,
+      "step": 696
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6055238842964172,
+      "learning_rate": 0.00019793676268994187,
+      "loss": 1.2432,
+      "step": 697
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.48325762152671814,
+      "learning_rate": 0.00019793068407710851,
+      "loss": 1.0837,
+      "step": 698
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5550029873847961,
+      "learning_rate": 0.00019792459661679236,
+      "loss": 1.2134,
+      "step": 699
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.49292463064193726,
+      "learning_rate": 0.00019791850030954333,
+      "loss": 1.0565,
+      "step": 700
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.48105907440185547,
+      "learning_rate": 0.00019791239515591224,
+      "loss": 1.0579,
+      "step": 701
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5209401845932007,
+      "learning_rate": 0.00019790628115645058,
+      "loss": 1.1013,
+      "step": 702
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5278286337852478,
+      "learning_rate": 0.00019790015831171079,
+      "loss": 1.0529,
+      "step": 703
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5473965406417847,
+      "learning_rate": 0.000197894026622246,
+      "loss": 1.121,
+      "step": 704
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5992923974990845,
+      "learning_rate": 0.00019788788608861017,
+      "loss": 1.3566,
+      "step": 705
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6117520928382874,
+      "learning_rate": 0.00019788173671135808,
+      "loss": 1.1498,
+      "step": 706
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5273234248161316,
+      "learning_rate": 0.00019787557849104527,
+      "loss": 1.0953,
+      "step": 707
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.49864324927330017,
+      "learning_rate": 0.00019786941142822814,
+      "loss": 1.0837,
+      "step": 708
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6347833871841431,
+      "learning_rate": 0.0001978632355234638,
+      "loss": 0.9678,
+      "step": 709
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5046628713607788,
+      "learning_rate": 0.00019785705077731024,
+      "loss": 1.0924,
+      "step": 710
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4987553656101227,
+      "learning_rate": 0.00019785085719032622,
+      "loss": 0.953,
+      "step": 711
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5188940167427063,
+      "learning_rate": 0.0001978446547630713,
+      "loss": 0.9737,
+      "step": 712
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.585609495639801,
+      "learning_rate": 0.00019783844349610576,
+      "loss": 0.9819,
+      "step": 713
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5059577226638794,
+      "learning_rate": 0.00019783222338999083,
+      "loss": 1.0045,
+      "step": 714
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5237625241279602,
+      "learning_rate": 0.00019782599444528846,
+      "loss": 0.8314,
+      "step": 715
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.49594178795814514,
+      "learning_rate": 0.00019781975666256137,
+      "loss": 0.9754,
+      "step": 716
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6107966899871826,
+      "learning_rate": 0.0001978135100423731,
+      "loss": 1.2708,
+      "step": 717
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5323823094367981,
+      "learning_rate": 0.00019780725458528803,
+      "loss": 1.1506,
+      "step": 718
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5096708536148071,
+      "learning_rate": 0.0001978009902918713,
+      "loss": 0.9683,
+      "step": 719
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.48245328664779663,
+      "learning_rate": 0.00019779471716268885,
+      "loss": 1.0692,
+      "step": 720
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5705918073654175,
+      "learning_rate": 0.00019778843519830742,
+      "loss": 1.1573,
+      "step": 721
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5196608901023865,
+      "learning_rate": 0.00019778214439929452,
+      "loss": 1.0694,
+      "step": 722
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5451333522796631,
+      "learning_rate": 0.00019777584476621853,
+      "loss": 1.1756,
+      "step": 723
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6256030797958374,
+      "learning_rate": 0.00019776953629964855,
+      "loss": 1.1507,
+      "step": 724
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5393909811973572,
+      "learning_rate": 0.00019776321900015457,
+      "loss": 0.992,
+      "step": 725
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.563827395439148,
+      "learning_rate": 0.0001977568928683073,
+      "loss": 1.111,
+      "step": 726
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.55351722240448,
+      "learning_rate": 0.0001977505579046782,
+      "loss": 1.1057,
+      "step": 727
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5464900135993958,
+      "learning_rate": 0.0001977442141098397,
+      "loss": 1.1258,
+      "step": 728
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6698980927467346,
+      "learning_rate": 0.00019773786148436485,
+      "loss": 1.1193,
+      "step": 729
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5508808493614197,
+      "learning_rate": 0.00019773150002882762,
+      "loss": 0.9323,
+      "step": 730
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5508869886398315,
+      "learning_rate": 0.00019772512974380275,
+      "loss": 1.0364,
+      "step": 731
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5076518654823303,
+      "learning_rate": 0.00019771875062986566,
+      "loss": 1.0878,
+      "step": 732
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4498327374458313,
+      "learning_rate": 0.0001977123626875928,
+      "loss": 1.0504,
+      "step": 733
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.45657438039779663,
+      "learning_rate": 0.00019770596591756116,
+      "loss": 0.9122,
+      "step": 734
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5719609260559082,
+      "learning_rate": 0.00019769956032034873,
+      "loss": 1.3045,
+      "step": 735
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6293899416923523,
+      "learning_rate": 0.00019769314589653423,
+      "loss": 1.1325,
+      "step": 736
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5307220220565796,
+      "learning_rate": 0.0001976867226466971,
+      "loss": 1.0942,
+      "step": 737
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5085244178771973,
+      "learning_rate": 0.0001976802905714177,
+      "loss": 1.1687,
+      "step": 738
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6277353167533875,
+      "learning_rate": 0.00019767384967127707,
+      "loss": 1.0866,
+      "step": 739
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5884594917297363,
+      "learning_rate": 0.00019766739994685718,
+      "loss": 1.175,
+      "step": 740
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5401700139045715,
+      "learning_rate": 0.0001976609413987407,
+      "loss": 1.1499,
+      "step": 741
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5686537623405457,
+      "learning_rate": 0.0001976544740275111,
+      "loss": 1.0965,
+      "step": 742
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5871611833572388,
+      "learning_rate": 0.00019764799783375273,
+      "loss": 1.2375,
+      "step": 743
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5987391471862793,
+      "learning_rate": 0.0001976415128180506,
+      "loss": 1.2463,
+      "step": 744
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4921867251396179,
+      "learning_rate": 0.00019763501898099065,
+      "loss": 1.0619,
+      "step": 745
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5578277707099915,
+      "learning_rate": 0.00019762851632315953,
+      "loss": 1.0181,
+      "step": 746
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5684673190116882,
+      "learning_rate": 0.00019762200484514474,
+      "loss": 1.2507,
+      "step": 747
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5005682110786438,
+      "learning_rate": 0.00019761548454753453,
+      "loss": 1.1111,
+      "step": 748
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5176371335983276,
+      "learning_rate": 0.000197608955430918,
+      "loss": 1.0868,
+      "step": 749
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6516520977020264,
+      "learning_rate": 0.00019760241749588501,
+      "loss": 0.8422,
+      "step": 750
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5634924173355103,
+      "learning_rate": 0.00019759587074302625,
+      "loss": 1.184,
+      "step": 751
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4741862416267395,
+      "learning_rate": 0.0001975893151729331,
+      "loss": 1.0444,
+      "step": 752
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5085967183113098,
+      "learning_rate": 0.0001975827507861979,
+      "loss": 0.9326,
+      "step": 753
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5261965394020081,
+      "learning_rate": 0.0001975761775834137,
+      "loss": 1.0514,
+      "step": 754
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.558593213558197,
+      "learning_rate": 0.0001975695955651743,
+      "loss": 0.9558,
+      "step": 755
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4968149960041046,
+      "learning_rate": 0.0001975630047320744,
+      "loss": 0.9564,
+      "step": 756
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5512811541557312,
+      "learning_rate": 0.00019755640508470942,
+      "loss": 1.0315,
+      "step": 757
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5796666145324707,
+      "learning_rate": 0.00019754979662367557,
+      "loss": 1.1686,
+      "step": 758
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.48021426796913147,
+      "learning_rate": 0.00019754317934956995,
+      "loss": 0.9551,
+      "step": 759
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5672682523727417,
+      "learning_rate": 0.00019753655326299035,
+      "loss": 1.0394,
+      "step": 760
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5781568288803101,
+      "learning_rate": 0.00019752991836453543,
+      "loss": 1.0976,
+      "step": 761
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.566172182559967,
+      "learning_rate": 0.00019752327465480458,
+      "loss": 1.0958,
+      "step": 762
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5664011240005493,
+      "learning_rate": 0.00019751662213439805,
+      "loss": 1.28,
+      "step": 763
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5427090525627136,
+      "learning_rate": 0.00019750996080391685,
+      "loss": 1.1989,
+      "step": 764
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5100749731063843,
+      "learning_rate": 0.0001975032906639628,
+      "loss": 0.9632,
+      "step": 765
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5142964720726013,
+      "learning_rate": 0.0001974966117151385,
+      "loss": 1.0154,
+      "step": 766
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.48683270812034607,
+      "learning_rate": 0.00019748992395804733,
+      "loss": 1.1774,
+      "step": 767
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4964374601840973,
+      "learning_rate": 0.00019748322739329352,
+      "loss": 1.0984,
+      "step": 768
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5703089833259583,
+      "learning_rate": 0.0001974765220214821,
+      "loss": 1.263,
+      "step": 769
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5038565993309021,
+      "learning_rate": 0.0001974698078432188,
+      "loss": 0.9689,
+      "step": 770
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5699312090873718,
+      "learning_rate": 0.00019746308485911023,
+      "loss": 1.2384,
+      "step": 771
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.47143155336380005,
+      "learning_rate": 0.0001974563530697638,
+      "loss": 1.0539,
+      "step": 772
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5221191644668579,
+      "learning_rate": 0.00019744961247578763,
+      "loss": 1.0732,
+      "step": 773
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5209429264068604,
+      "learning_rate": 0.00019744286307779075,
+      "loss": 1.0056,
+      "step": 774
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5356734395027161,
+      "learning_rate": 0.0001974361048763829,
+      "loss": 0.9635,
+      "step": 775
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.512253999710083,
+      "learning_rate": 0.00019742933787217468,
+      "loss": 1.1979,
+      "step": 776
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5092757344245911,
+      "learning_rate": 0.0001974225620657774,
+      "loss": 1.1399,
+      "step": 777
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6500785946846008,
+      "learning_rate": 0.00019741577745780326,
+      "loss": 1.0413,
+      "step": 778
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5108067393302917,
+      "learning_rate": 0.0001974089840488652,
+      "loss": 0.8789,
+      "step": 779
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5151072144508362,
+      "learning_rate": 0.00019740218183957696,
+      "loss": 1.0146,
+      "step": 780
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.49716851115226746,
+      "learning_rate": 0.00019739537083055306,
+      "loss": 1.0114,
+      "step": 781
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.48336997628211975,
+      "learning_rate": 0.00019738855102240885,
+      "loss": 0.8086,
+      "step": 782
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5495067238807678,
+      "learning_rate": 0.00019738172241576049,
+      "loss": 1.1508,
+      "step": 783
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6326724886894226,
+      "learning_rate": 0.00019737488501122487,
+      "loss": 1.0295,
+      "step": 784
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5587719082832336,
+      "learning_rate": 0.00019736803880941977,
+      "loss": 1.0741,
+      "step": 785
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6244709491729736,
+      "learning_rate": 0.0001973611838109636,
+      "loss": 1.1393,
+      "step": 786
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5473697185516357,
+      "learning_rate": 0.00019735432001647576,
+      "loss": 1.0927,
+      "step": 787
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.531061053276062,
+      "learning_rate": 0.00019734744742657632,
+      "loss": 1.1514,
+      "step": 788
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.48232513666152954,
+      "learning_rate": 0.00019734056604188616,
+      "loss": 1.1516,
+      "step": 789
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6108174324035645,
+      "learning_rate": 0.00019733367586302703,
+      "loss": 1.1386,
+      "step": 790
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6374059319496155,
+      "learning_rate": 0.0001973267768906214,
+      "loss": 0.9623,
+      "step": 791
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5877413153648376,
+      "learning_rate": 0.0001973198691252925,
+      "loss": 1.0466,
+      "step": 792
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5468994975090027,
+      "learning_rate": 0.00019731295256766447,
+      "loss": 1.1942,
+      "step": 793
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5458383560180664,
+      "learning_rate": 0.00019730602721836218,
+      "loss": 1.1622,
+      "step": 794
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4972197115421295,
+      "learning_rate": 0.00019729909307801124,
+      "loss": 1.027,
+      "step": 795
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5629715919494629,
+      "learning_rate": 0.00019729215014723816,
+      "loss": 1.3745,
+      "step": 796
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5757743120193481,
+      "learning_rate": 0.0001972851984266702,
+      "loss": 1.0184,
+      "step": 797
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.572532594203949,
+      "learning_rate": 0.00019727823791693537,
+      "loss": 1.0389,
+      "step": 798
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5449635982513428,
+      "learning_rate": 0.00019727126861866253,
+      "loss": 1.0972,
+      "step": 799
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5162309408187866,
+      "learning_rate": 0.0001972642905324813,
+      "loss": 1.167,
+      "step": 800
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.611807644367218,
+      "learning_rate": 0.0001972573036590222,
+      "loss": 1.3319,
+      "step": 801
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5340973138809204,
+      "learning_rate": 0.0001972503079989163,
+      "loss": 1.1356,
+      "step": 802
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.56681889295578,
+      "learning_rate": 0.00019724330355279574,
+      "loss": 1.2657,
+      "step": 803
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.49857208132743835,
+      "learning_rate": 0.00019723629032129329,
+      "loss": 1.1107,
+      "step": 804
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5250954627990723,
+      "learning_rate": 0.00019722926830504256,
+      "loss": 1.1725,
+      "step": 805
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5133506655693054,
+      "learning_rate": 0.00019722223750467794,
+      "loss": 0.9294,
+      "step": 806
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5176982879638672,
+      "learning_rate": 0.00019721519792083468,
+      "loss": 1.0949,
+      "step": 807
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5122258067131042,
+      "learning_rate": 0.00019720814955414865,
+      "loss": 1.0317,
+      "step": 808
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6034652590751648,
+      "learning_rate": 0.00019720109240525674,
+      "loss": 0.9448,
+      "step": 809
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5805848240852356,
+      "learning_rate": 0.00019719402647479649,
+      "loss": 1.1572,
+      "step": 810
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5414599180221558,
+      "learning_rate": 0.00019718695176340623,
+      "loss": 1.0706,
+      "step": 811
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5518433451652527,
+      "learning_rate": 0.00019717986827172515,
+      "loss": 1.1705,
+      "step": 812
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5469940304756165,
+      "learning_rate": 0.00019717277600039325,
+      "loss": 1.1299,
+      "step": 813
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5368144512176514,
+      "learning_rate": 0.00019716567495005116,
+      "loss": 1.1,
+      "step": 814
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5745230317115784,
+      "learning_rate": 0.00019715856512134052,
+      "loss": 1.052,
+      "step": 815
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.466328501701355,
+      "learning_rate": 0.00019715144651490362,
+      "loss": 0.9903,
+      "step": 816
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6098025441169739,
+      "learning_rate": 0.00019714431913138363,
+      "loss": 1.0933,
+      "step": 817
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5240216851234436,
+      "learning_rate": 0.0001971371829714244,
+      "loss": 1.1298,
+      "step": 818
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.510958731174469,
+      "learning_rate": 0.00019713003803567073,
+      "loss": 1.1109,
+      "step": 819
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.505074679851532,
+      "learning_rate": 0.00019712288432476802,
+      "loss": 1.1258,
+      "step": 820
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5658806562423706,
+      "learning_rate": 0.00019711572183936266,
+      "loss": 1.0631,
+      "step": 821
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4546443521976471,
+      "learning_rate": 0.00019710855058010168,
+      "loss": 0.916,
+      "step": 822
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.526270866394043,
+      "learning_rate": 0.00019710137054763295,
+      "loss": 1.1089,
+      "step": 823
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5158497095108032,
+      "learning_rate": 0.0001970941817426052,
+      "loss": 1.0486,
+      "step": 824
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5480288863182068,
+      "learning_rate": 0.0001970869841656679,
+      "loss": 1.0551,
+      "step": 825
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.46271073818206787,
+      "learning_rate": 0.00019707977781747125,
+      "loss": 1.0294,
+      "step": 826
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5485731363296509,
+      "learning_rate": 0.00019707256269866638,
+      "loss": 1.1623,
+      "step": 827
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5255643725395203,
+      "learning_rate": 0.00019706533880990506,
+      "loss": 0.9108,
+      "step": 828
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5513526201248169,
+      "learning_rate": 0.00019705810615183996,
+      "loss": 1.1178,
+      "step": 829
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5589191317558289,
+      "learning_rate": 0.00019705086472512453,
+      "loss": 1.2384,
+      "step": 830
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.48324573040008545,
+      "learning_rate": 0.00019704361453041296,
+      "loss": 0.9521,
+      "step": 831
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5757077932357788,
+      "learning_rate": 0.00019703635556836028,
+      "loss": 1.1034,
+      "step": 832
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6256049871444702,
+      "learning_rate": 0.00019702908783962228,
+      "loss": 1.2153,
+      "step": 833
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5107579827308655,
+      "learning_rate": 0.0001970218113448556,
+      "loss": 1.0975,
+      "step": 834
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.47654515504837036,
+      "learning_rate": 0.00019701452608471759,
+      "loss": 1.0845,
+      "step": 835
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.557918906211853,
+      "learning_rate": 0.00019700723205986643,
+      "loss": 1.1108,
+      "step": 836
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5833942294120789,
+      "learning_rate": 0.00019699992927096113,
+      "loss": 1.124,
+      "step": 837
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5856698751449585,
+      "learning_rate": 0.0001969926177186614,
+      "loss": 1.2941,
+      "step": 838
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6649505496025085,
+      "learning_rate": 0.00019698529740362785,
+      "loss": 1.1847,
+      "step": 839
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5869125723838806,
+      "learning_rate": 0.0001969779683265218,
+      "loss": 1.1469,
+      "step": 840
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.47669145464897156,
+      "learning_rate": 0.00019697063048800542,
+      "loss": 1.0475,
+      "step": 841
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5173715949058533,
+      "learning_rate": 0.00019696328388874162,
+      "loss": 1.1664,
+      "step": 842
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4949846863746643,
+      "learning_rate": 0.0001969559285293941,
+      "loss": 1.2023,
+      "step": 843
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5999616384506226,
+      "learning_rate": 0.0001969485644106274,
+      "loss": 1.1492,
+      "step": 844
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5564090013504028,
+      "learning_rate": 0.00019694119153310685,
+      "loss": 0.9729,
+      "step": 845
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5422832369804382,
+      "learning_rate": 0.00019693380989749853,
+      "loss": 1.1642,
+      "step": 846
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5115847587585449,
+      "learning_rate": 0.0001969264195044693,
+      "loss": 1.1102,
+      "step": 847
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5365723967552185,
+      "learning_rate": 0.00019691902035468687,
+      "loss": 1.1249,
+      "step": 848
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5202025175094604,
+      "learning_rate": 0.0001969116124488197,
+      "loss": 1.1992,
+      "step": 849
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5052026510238647,
+      "learning_rate": 0.00019690419578753706,
+      "loss": 0.9967,
+      "step": 850
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.552277147769928,
+      "learning_rate": 0.00019689677037150902,
+      "loss": 1.161,
+      "step": 851
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5004220604896545,
+      "learning_rate": 0.00019688933620140637,
+      "loss": 1.2874,
+      "step": 852
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.596388578414917,
+      "learning_rate": 0.0001968818932779008,
+      "loss": 1.1453,
+      "step": 853
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5080670714378357,
+      "learning_rate": 0.00019687444160166473,
+      "loss": 0.9901,
+      "step": 854
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5395559072494507,
+      "learning_rate": 0.00019686698117337133,
+      "loss": 1.2054,
+      "step": 855
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.592228353023529,
+      "learning_rate": 0.00019685951199369467,
+      "loss": 1.1201,
+      "step": 856
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5213476419448853,
+      "learning_rate": 0.0001968520340633095,
+      "loss": 1.1273,
+      "step": 857
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5358402729034424,
+      "learning_rate": 0.00019684454738289143,
+      "loss": 1.1751,
+      "step": 858
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6148902177810669,
+      "learning_rate": 0.00019683705195311686,
+      "loss": 1.2059,
+      "step": 859
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6230486035346985,
+      "learning_rate": 0.00019682954777466293,
+      "loss": 1.1166,
+      "step": 860
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4527413547039032,
+      "learning_rate": 0.00019682203484820758,
+      "loss": 1.0781,
+      "step": 861
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5335724949836731,
+      "learning_rate": 0.00019681451317442962,
+      "loss": 1.1133,
+      "step": 862
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5677633881568909,
+      "learning_rate": 0.00019680698275400854,
+      "loss": 1.0835,
+      "step": 863
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5778499245643616,
+      "learning_rate": 0.00019679944358762466,
+      "loss": 1.2469,
+      "step": 864
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5224189162254333,
+      "learning_rate": 0.00019679189567595916,
+      "loss": 1.2731,
+      "step": 865
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6404327154159546,
+      "learning_rate": 0.00019678433901969391,
+      "loss": 1.2088,
+      "step": 866
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5300662517547607,
+      "learning_rate": 0.00019677677361951165,
+      "loss": 1.3882,
+      "step": 867
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.47733646631240845,
+      "learning_rate": 0.0001967691994760958,
+      "loss": 0.9524,
+      "step": 868
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5740486979484558,
+      "learning_rate": 0.0001967616165901307,
+      "loss": 1.1168,
+      "step": 869
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5322950482368469,
+      "learning_rate": 0.0001967540249623014,
+      "loss": 1.176,
+      "step": 870
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4464534819126129,
+      "learning_rate": 0.00019674642459329377,
+      "loss": 0.888,
+      "step": 871
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5296953916549683,
+      "learning_rate": 0.00019673881548379444,
+      "loss": 1.2192,
+      "step": 872
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5391892194747925,
+      "learning_rate": 0.0001967311976344909,
+      "loss": 1.1632,
+      "step": 873
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.47413426637649536,
+      "learning_rate": 0.00019672357104607132,
+      "loss": 1.0121,
+      "step": 874
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5288782119750977,
+      "learning_rate": 0.00019671593571922476,
+      "loss": 0.895,
+      "step": 875
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5563179850578308,
+      "learning_rate": 0.000196708291654641,
+      "loss": 1.306,
+      "step": 876
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6332594752311707,
+      "learning_rate": 0.00019670063885301065,
+      "loss": 1.2006,
+      "step": 877
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5800545811653137,
+      "learning_rate": 0.00019669297731502507,
+      "loss": 1.1349,
+      "step": 878
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.553862452507019,
+      "learning_rate": 0.0001966853070413765,
+      "loss": 1.1195,
+      "step": 879
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5305919647216797,
+      "learning_rate": 0.00019667762803275785,
+      "loss": 1.1277,
+      "step": 880
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6028212308883667,
+      "learning_rate": 0.0001966699402898629,
+      "loss": 1.3267,
+      "step": 881
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5439375042915344,
+      "learning_rate": 0.00019666224381338618,
+      "loss": 1.0961,
+      "step": 882
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5040217638015747,
+      "learning_rate": 0.00019665453860402305,
+      "loss": 0.9993,
+      "step": 883
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6264601945877075,
+      "learning_rate": 0.00019664682466246963,
+      "loss": 1.2248,
+      "step": 884
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.48470181226730347,
+      "learning_rate": 0.00019663910198942277,
+      "loss": 1.0676,
+      "step": 885
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4559854567050934,
+      "learning_rate": 0.00019663137058558024,
+      "loss": 0.9664,
+      "step": 886
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6404067277908325,
+      "learning_rate": 0.00019662363045164048,
+      "loss": 1.1033,
+      "step": 887
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5953955054283142,
+      "learning_rate": 0.00019661588158830285,
+      "loss": 1.2204,
+      "step": 888
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5667288899421692,
+      "learning_rate": 0.0001966081239962673,
+      "loss": 0.9428,
+      "step": 889
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5570257902145386,
+      "learning_rate": 0.00019660035767623475,
+      "loss": 1.1523,
+      "step": 890
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6536972522735596,
+      "learning_rate": 0.00019659258262890683,
+      "loss": 1.2999,
+      "step": 891
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5308727622032166,
+      "learning_rate": 0.000196584798854986,
+      "loss": 1.1709,
+      "step": 892
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4792231321334839,
+      "learning_rate": 0.00019657700635517542,
+      "loss": 0.9741,
+      "step": 893
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5758273005485535,
+      "learning_rate": 0.00019656920513017915,
+      "loss": 1.2811,
+      "step": 894
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5072007179260254,
+      "learning_rate": 0.00019656139518070198,
+      "loss": 1.0564,
+      "step": 895
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5229262709617615,
+      "learning_rate": 0.0001965535765074495,
+      "loss": 0.9786,
+      "step": 896
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.46316370368003845,
+      "learning_rate": 0.00019654574911112804,
+      "loss": 1.1107,
+      "step": 897
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5957727432250977,
+      "learning_rate": 0.00019653791299244477,
+      "loss": 1.1218,
+      "step": 898
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.534303605556488,
+      "learning_rate": 0.0001965300681521077,
+      "loss": 1.2024,
+      "step": 899
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5468682050704956,
+      "learning_rate": 0.00019652221459082552,
+      "loss": 0.936,
+      "step": 900
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5230331420898438,
+      "learning_rate": 0.0001965143523093077,
+      "loss": 1.0716,
+      "step": 901
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4797626733779907,
+      "learning_rate": 0.00019650648130826467,
+      "loss": 1.002,
+      "step": 902
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5666486024856567,
+      "learning_rate": 0.00019649860158840748,
+      "loss": 1.0306,
+      "step": 903
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5790895223617554,
+      "learning_rate": 0.00019649071315044797,
+      "loss": 1.0271,
+      "step": 904
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5381376147270203,
+      "learning_rate": 0.00019648281599509888,
+      "loss": 1.1928,
+      "step": 905
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5734236836433411,
+      "learning_rate": 0.00019647491012307364,
+      "loss": 1.2453,
+      "step": 906
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4927200376987457,
+      "learning_rate": 0.0001964669955350865,
+      "loss": 0.98,
+      "step": 907
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5980042219161987,
+      "learning_rate": 0.00019645907223185254,
+      "loss": 1.0942,
+      "step": 908
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4530969560146332,
+      "learning_rate": 0.0001964511402140875,
+      "loss": 0.9979,
+      "step": 909
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6606910228729248,
+      "learning_rate": 0.00019644319948250812,
+      "loss": 1.1149,
+      "step": 910
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5825684666633606,
+      "learning_rate": 0.0001964352500378317,
+      "loss": 1.094,
+      "step": 911
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5432309508323669,
+      "learning_rate": 0.00019642729188077643,
+      "loss": 1.0084,
+      "step": 912
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5725202560424805,
+      "learning_rate": 0.0001964193250120613,
+      "loss": 1.1022,
+      "step": 913
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6069723963737488,
+      "learning_rate": 0.00019641134943240612,
+      "loss": 1.1993,
+      "step": 914
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5283558964729309,
+      "learning_rate": 0.00019640336514253136,
+      "loss": 0.8801,
+      "step": 915
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6143441796302795,
+      "learning_rate": 0.00019639537214315844,
+      "loss": 1.0242,
+      "step": 916
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5614748597145081,
+      "learning_rate": 0.0001963873704350094,
+      "loss": 1.1471,
+      "step": 917
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5619175434112549,
+      "learning_rate": 0.0001963793600188072,
+      "loss": 1.0781,
+      "step": 918
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5835575461387634,
+      "learning_rate": 0.0001963713408952755,
+      "loss": 1.0833,
+      "step": 919
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6437380909919739,
+      "learning_rate": 0.00019636331306513883,
+      "loss": 1.0159,
+      "step": 920
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5471232533454895,
+      "learning_rate": 0.00019635527652912244,
+      "loss": 0.9727,
+      "step": 921
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5375336408615112,
+      "learning_rate": 0.00019634723128795231,
+      "loss": 1.15,
+      "step": 922
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5501574277877808,
+      "learning_rate": 0.00019633917734235542,
+      "loss": 1.0332,
+      "step": 923
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5777191519737244,
+      "learning_rate": 0.0001963311146930593,
+      "loss": 1.0671,
+      "step": 924
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.558753490447998,
+      "learning_rate": 0.00019632304334079242,
+      "loss": 1.0862,
+      "step": 925
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.49828171730041504,
+      "learning_rate": 0.00019631496328628391,
+      "loss": 1.1407,
+      "step": 926
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.538738489151001,
+      "learning_rate": 0.0001963068745302638,
+      "loss": 1.0637,
+      "step": 927
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.609881579875946,
+      "learning_rate": 0.00019629877707346292,
+      "loss": 1.1749,
+      "step": 928
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5410135388374329,
+      "learning_rate": 0.00019629067091661272,
+      "loss": 1.0109,
+      "step": 929
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4779724180698395,
+      "learning_rate": 0.00019628255606044564,
+      "loss": 1.1007,
+      "step": 930
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.603225827217102,
+      "learning_rate": 0.00019627443250569475,
+      "loss": 1.0987,
+      "step": 931
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5400119423866272,
+      "learning_rate": 0.000196266300253094,
+      "loss": 1.1858,
+      "step": 932
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5057092905044556,
+      "learning_rate": 0.00019625815930337804,
+      "loss": 1.1432,
+      "step": 933
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.545569658279419,
+      "learning_rate": 0.00019625000965728243,
+      "loss": 1.2376,
+      "step": 934
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.47009384632110596,
+      "learning_rate": 0.00019624185131554342,
+      "loss": 0.9036,
+      "step": 935
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5348869562149048,
+      "learning_rate": 0.00019623368427889806,
+      "loss": 1.052,
+      "step": 936
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.45622172951698303,
+      "learning_rate": 0.0001962255085480842,
+      "loss": 0.8479,
+      "step": 937
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5504438281059265,
+      "learning_rate": 0.00019621732412384045,
+      "loss": 1.1362,
+      "step": 938
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.48980462551116943,
+      "learning_rate": 0.00019620913100690626,
+      "loss": 1.009,
+      "step": 939
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5361881256103516,
+      "learning_rate": 0.00019620092919802183,
+      "loss": 1.1752,
+      "step": 940
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.54579758644104,
+      "learning_rate": 0.00019619271869792814,
+      "loss": 1.0558,
+      "step": 941
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.549178421497345,
+      "learning_rate": 0.0001961844995073669,
+      "loss": 1.0308,
+      "step": 942
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5896941423416138,
+      "learning_rate": 0.0001961762716270808,
+      "loss": 1.2261,
+      "step": 943
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5419191122055054,
+      "learning_rate": 0.00019616803505781305,
+      "loss": 1.152,
+      "step": 944
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5686753392219543,
+      "learning_rate": 0.00019615978980030784,
+      "loss": 1.0636,
+      "step": 945
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5907273888587952,
+      "learning_rate": 0.0001961515358553101,
+      "loss": 1.0999,
+      "step": 946
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5179197788238525,
+      "learning_rate": 0.0001961432732235655,
+      "loss": 1.0976,
+      "step": 947
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5165078639984131,
+      "learning_rate": 0.00019613500190582053,
+      "loss": 1.1532,
+      "step": 948
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5007051229476929,
+      "learning_rate": 0.00019612672190282243,
+      "loss": 1.0341,
+      "step": 949
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5154484510421753,
+      "learning_rate": 0.0001961184332153193,
+      "loss": 1.0931,
+      "step": 950
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6172532439231873,
+      "learning_rate": 0.00019611013584405995,
+      "loss": 1.2503,
+      "step": 951
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5256145000457764,
+      "learning_rate": 0.000196101829789794,
+      "loss": 1.1124,
+      "step": 952
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.49367621541023254,
+      "learning_rate": 0.00019609351505327184,
+      "loss": 1.0179,
+      "step": 953
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5787032246589661,
+      "learning_rate": 0.00019608519163524468,
+      "loss": 1.0986,
+      "step": 954
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5206260085105896,
+      "learning_rate": 0.00019607685953646452,
+      "loss": 0.9906,
+      "step": 955
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5005407929420471,
+      "learning_rate": 0.000196068518757684,
+      "loss": 1.0735,
+      "step": 956
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5073145031929016,
+      "learning_rate": 0.00019606016929965683,
+      "loss": 1.0736,
+      "step": 957
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5678821206092834,
+      "learning_rate": 0.00019605181116313724,
+      "loss": 1.1991,
+      "step": 958
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5657750964164734,
+      "learning_rate": 0.00019604344434888033,
+      "loss": 1.0046,
+      "step": 959
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5822930335998535,
+      "learning_rate": 0.00019603506885764206,
+      "loss": 1.2042,
+      "step": 960
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5306090116500854,
+      "learning_rate": 0.00019602668469017903,
+      "loss": 1.0183,
+      "step": 961
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6258050799369812,
+      "learning_rate": 0.00019601829184724877,
+      "loss": 1.2029,
+      "step": 962
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5240448713302612,
+      "learning_rate": 0.0001960098903296095,
+      "loss": 0.9021,
+      "step": 963
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5050828456878662,
+      "learning_rate": 0.0001960014801380202,
+      "loss": 0.9737,
+      "step": 964
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4429282546043396,
+      "learning_rate": 0.00019599306127324076,
+      "loss": 0.8778,
+      "step": 965
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5414524078369141,
+      "learning_rate": 0.00019598463373603174,
+      "loss": 1.1454,
+      "step": 966
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5213294625282288,
+      "learning_rate": 0.0001959761975271545,
+      "loss": 1.171,
+      "step": 967
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5557194352149963,
+      "learning_rate": 0.00019596775264737126,
+      "loss": 1.2047,
+      "step": 968
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5759382843971252,
+      "learning_rate": 0.00019595929909744494,
+      "loss": 1.1048,
+      "step": 969
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5198878645896912,
+      "learning_rate": 0.00019595083687813926,
+      "loss": 1.044,
+      "step": 970
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5648888349533081,
+      "learning_rate": 0.00019594236599021874,
+      "loss": 1.1299,
+      "step": 971
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5540794730186462,
+      "learning_rate": 0.00019593388643444865,
+      "loss": 1.1495,
+      "step": 972
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5652869939804077,
+      "learning_rate": 0.00019592539821159513,
+      "loss": 0.9845,
+      "step": 973
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6016077399253845,
+      "learning_rate": 0.000195916901322425,
+      "loss": 1.0299,
+      "step": 974
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6304303407669067,
+      "learning_rate": 0.00019590839576770592,
+      "loss": 1.1451,
+      "step": 975
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5486611127853394,
+      "learning_rate": 0.00019589988154820632,
+      "loss": 1.1388,
+      "step": 976
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5753059983253479,
+      "learning_rate": 0.0001958913586646954,
+      "loss": 1.0921,
+      "step": 977
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.49512913823127747,
+      "learning_rate": 0.00019588282711794315,
+      "loss": 1.0276,
+      "step": 978
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5401121377944946,
+      "learning_rate": 0.00019587428690872035,
+      "loss": 0.9792,
+      "step": 979
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5886608362197876,
+      "learning_rate": 0.00019586573803779858,
+      "loss": 1.0626,
+      "step": 980
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4967743158340454,
+      "learning_rate": 0.00019585718050595019,
+      "loss": 1.1254,
+      "step": 981
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.49494069814682007,
+      "learning_rate": 0.00019584861431394822,
+      "loss": 1.1964,
+      "step": 982
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5207960605621338,
+      "learning_rate": 0.0001958400394625667,
+      "loss": 1.1442,
+      "step": 983
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5062996745109558,
+      "learning_rate": 0.00019583145595258023,
+      "loss": 1.1793,
+      "step": 984
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5631035566329956,
+      "learning_rate": 0.00019582286378476432,
+      "loss": 1.4415,
+      "step": 985
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.568055272102356,
+      "learning_rate": 0.00019581426295989525,
+      "loss": 1.2234,
+      "step": 986
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5561823844909668,
+      "learning_rate": 0.00019580565347874997,
+      "loss": 1.2309,
+      "step": 987
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5685033202171326,
+      "learning_rate": 0.00019579703534210635,
+      "loss": 1.3636,
+      "step": 988
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5274617671966553,
+      "learning_rate": 0.00019578840855074303,
+      "loss": 1.0681,
+      "step": 989
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5587525963783264,
+      "learning_rate": 0.00019577977310543932,
+      "loss": 1.1667,
+      "step": 990
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5266395211219788,
+      "learning_rate": 0.00019577112900697542,
+      "loss": 1.2668,
+      "step": 991
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5371447801589966,
+      "learning_rate": 0.00019576247625613226,
+      "loss": 1.2036,
+      "step": 992
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6547426581382751,
+      "learning_rate": 0.00019575381485369158,
+      "loss": 1.1855,
+      "step": 993
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4874076843261719,
+      "learning_rate": 0.00019574514480043587,
+      "loss": 1.1199,
+      "step": 994
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5545238852500916,
+      "learning_rate": 0.0001957364660971485,
+      "loss": 1.1382,
+      "step": 995
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5526460409164429,
+      "learning_rate": 0.00019572777874461342,
+      "loss": 1.0697,
+      "step": 996
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6076409816741943,
+      "learning_rate": 0.00019571908274361555,
+      "loss": 1.3189,
+      "step": 997
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5631090998649597,
+      "learning_rate": 0.00019571037809494048,
+      "loss": 1.2412,
+      "step": 998
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5299128890037537,
+      "learning_rate": 0.0001957016647993747,
+      "loss": 1.0527,
+      "step": 999
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5710862874984741,
+      "learning_rate": 0.00019569294285770537,
+      "loss": 1.0326,
+      "step": 1000
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5347865223884583,
+      "learning_rate": 0.00019568421227072046,
+      "loss": 1.0002,
+      "step": 1001
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6362932324409485,
+      "learning_rate": 0.0001956754730392087,
+      "loss": 1.1981,
+      "step": 1002
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5342056155204773,
+      "learning_rate": 0.00019566672516395972,
+      "loss": 1.1775,
+      "step": 1003
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6062530279159546,
+      "learning_rate": 0.00019565796864576377,
+      "loss": 1.1539,
+      "step": 1004
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6102681159973145,
+      "learning_rate": 0.00019564920348541192,
+      "loss": 1.0816,
+      "step": 1005
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5745230913162231,
+      "learning_rate": 0.00019564042968369614,
+      "loss": 0.9836,
+      "step": 1006
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5198560953140259,
+      "learning_rate": 0.00019563164724140903,
+      "loss": 1.003,
+      "step": 1007
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6608436703681946,
+      "learning_rate": 0.0001956228561593441,
+      "loss": 1.2068,
+      "step": 1008
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6578373908996582,
+      "learning_rate": 0.00019561405643829549,
+      "loss": 1.2482,
+      "step": 1009
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5678254961967468,
+      "learning_rate": 0.00019560524807905828,
+      "loss": 1.0239,
+      "step": 1010
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5507716536521912,
+      "learning_rate": 0.00019559643108242817,
+      "loss": 1.0791,
+      "step": 1011
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5153545141220093,
+      "learning_rate": 0.00019558760544920184,
+      "loss": 1.1079,
+      "step": 1012
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5246258974075317,
+      "learning_rate": 0.0001955787711801765,
+      "loss": 1.0968,
+      "step": 1013
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5086669325828552,
+      "learning_rate": 0.00019556992827615043,
+      "loss": 1.0082,
+      "step": 1014
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5282449722290039,
+      "learning_rate": 0.00019556107673792238,
+      "loss": 1.04,
+      "step": 1015
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5100098848342896,
+      "learning_rate": 0.00019555221656629218,
+      "loss": 1.106,
+      "step": 1016
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5582842826843262,
+      "learning_rate": 0.0001955433477620602,
+      "loss": 1.1453,
+      "step": 1017
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5469565391540527,
+      "learning_rate": 0.0001955344703260277,
+      "loss": 1.0879,
+      "step": 1018
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5286270976066589,
+      "learning_rate": 0.00019552558425899673,
+      "loss": 1.0567,
+      "step": 1019
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5991113185882568,
+      "learning_rate": 0.0001955166895617701,
+      "loss": 1.1594,
+      "step": 1020
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.514499843120575,
+      "learning_rate": 0.00019550778623515135,
+      "loss": 1.0381,
+      "step": 1021
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5724852085113525,
+      "learning_rate": 0.00019549887427994494,
+      "loss": 1.1264,
+      "step": 1022
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5247836112976074,
+      "learning_rate": 0.00019548995369695588,
+      "loss": 1.0393,
+      "step": 1023
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5243775248527527,
+      "learning_rate": 0.0001954810244869902,
+      "loss": 0.9849,
+      "step": 1024
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5602267980575562,
+      "learning_rate": 0.00019547208665085457,
+      "loss": 1.1284,
+      "step": 1025
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5384296178817749,
+      "learning_rate": 0.00019546314018935648,
+      "loss": 1.1361,
+      "step": 1026
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5097274780273438,
+      "learning_rate": 0.00019545418510330417,
+      "loss": 0.7718,
+      "step": 1027
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.529334306716919,
+      "learning_rate": 0.0001954452213935067,
+      "loss": 0.9913,
+      "step": 1028
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6140390634536743,
+      "learning_rate": 0.00019543624906077388,
+      "loss": 1.1762,
+      "step": 1029
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5565348863601685,
+      "learning_rate": 0.00019542726810591634,
+      "loss": 1.0107,
+      "step": 1030
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6203579306602478,
+      "learning_rate": 0.0001954182785297454,
+      "loss": 0.9787,
+      "step": 1031
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.541043758392334,
+      "learning_rate": 0.00019540928033307328,
+      "loss": 1.0197,
+      "step": 1032
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5186806917190552,
+      "learning_rate": 0.00019540027351671287,
+      "loss": 1.0029,
+      "step": 1033
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5269144177436829,
+      "learning_rate": 0.0001953912580814779,
+      "loss": 1.1005,
+      "step": 1034
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5349538326263428,
+      "learning_rate": 0.00019538223402818287,
+      "loss": 1.1029,
+      "step": 1035
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5993475914001465,
+      "learning_rate": 0.00019537320135764308,
+      "loss": 1.1197,
+      "step": 1036
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.525951623916626,
+      "learning_rate": 0.0001953641600706745,
+      "loss": 0.9929,
+      "step": 1037
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6319278478622437,
+      "learning_rate": 0.00019535511016809405,
+      "loss": 1.053,
+      "step": 1038
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5484718680381775,
+      "learning_rate": 0.0001953460516507193,
+      "loss": 1.0309,
+      "step": 1039
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.649655818939209,
+      "learning_rate": 0.00019533698451936856,
+      "loss": 1.1151,
+      "step": 1040
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5016194581985474,
+      "learning_rate": 0.00019532790877486113,
+      "loss": 1.0153,
+      "step": 1041
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6027005910873413,
+      "learning_rate": 0.0001953188244180169,
+      "loss": 0.9985,
+      "step": 1042
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.526719331741333,
+      "learning_rate": 0.00019530973144965652,
+      "loss": 0.9503,
+      "step": 1043
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6107224822044373,
+      "learning_rate": 0.0001953006298706016,
+      "loss": 1.0747,
+      "step": 1044
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5502063035964966,
+      "learning_rate": 0.0001952915196816743,
+      "loss": 1.0741,
+      "step": 1045
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5427907109260559,
+      "learning_rate": 0.00019528240088369778,
+      "loss": 0.8883,
+      "step": 1046
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5658277869224548,
+      "learning_rate": 0.00019527327347749582,
+      "loss": 1.1569,
+      "step": 1047
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5371255874633789,
+      "learning_rate": 0.00019526413746389306,
+      "loss": 1.4064,
+      "step": 1048
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.572701632976532,
+      "learning_rate": 0.00019525499284371483,
+      "loss": 1.3042,
+      "step": 1049
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6183245182037354,
+      "learning_rate": 0.00019524583961778733,
+      "loss": 1.0634,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5503761172294617,
+      "learning_rate": 0.00019523667778693752,
+      "loss": 0.9387,
+      "step": 1051
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.553078293800354,
+      "learning_rate": 0.0001952275073519931,
+      "loss": 0.9584,
+      "step": 1052
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6011327505111694,
+      "learning_rate": 0.00019521832831378256,
+      "loss": 1.0545,
+      "step": 1053
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5534068942070007,
+      "learning_rate": 0.00019520914067313515,
+      "loss": 0.9852,
+      "step": 1054
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5138425230979919,
+      "learning_rate": 0.000195199944430881,
+      "loss": 0.9858,
+      "step": 1055
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5652523040771484,
+      "learning_rate": 0.00019519073958785088,
+      "loss": 1.0748,
+      "step": 1056
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5925012230873108,
+      "learning_rate": 0.0001951815261448764,
+      "loss": 1.0663,
+      "step": 1057
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6615341305732727,
+      "learning_rate": 0.00019517230410278995,
+      "loss": 1.2507,
+      "step": 1058
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5900757908821106,
+      "learning_rate": 0.0001951630734624247,
+      "loss": 1.1842,
+      "step": 1059
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6076449155807495,
+      "learning_rate": 0.00019515383422461454,
+      "loss": 1.0514,
+      "step": 1060
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5401352643966675,
+      "learning_rate": 0.00019514458639019426,
+      "loss": 1.036,
+      "step": 1061
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5284707546234131,
+      "learning_rate": 0.00019513532995999927,
+      "loss": 1.1178,
+      "step": 1062
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.606998860836029,
+      "learning_rate": 0.0001951260649348659,
+      "loss": 1.1377,
+      "step": 1063
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5934026837348938,
+      "learning_rate": 0.00019511679131563114,
+      "loss": 1.1728,
+      "step": 1064
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5732269883155823,
+      "learning_rate": 0.00019510750910313286,
+      "loss": 0.9816,
+      "step": 1065
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.492980033159256,
+      "learning_rate": 0.00019509821829820964,
+      "loss": 1.1011,
+      "step": 1066
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6220079064369202,
+      "learning_rate": 0.0001950889189017008,
+      "loss": 1.1738,
+      "step": 1067
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5315749645233154,
+      "learning_rate": 0.00019507961091444655,
+      "loss": 1.1742,
+      "step": 1068
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5375043153762817,
+      "learning_rate": 0.00019507029433728782,
+      "loss": 0.9013,
+      "step": 1069
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5389807820320129,
+      "learning_rate": 0.00019506096917106626,
+      "loss": 1.1154,
+      "step": 1070
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5907564759254456,
+      "learning_rate": 0.00019505163541662433,
+      "loss": 1.2198,
+      "step": 1071
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5677747130393982,
+      "learning_rate": 0.0001950422930748054,
+      "loss": 1.0026,
+      "step": 1072
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4908526539802551,
+      "learning_rate": 0.00019503294214645337,
+      "loss": 1.1098,
+      "step": 1073
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6064789295196533,
+      "learning_rate": 0.00019502358263241308,
+      "loss": 1.1671,
+      "step": 1074
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5308496356010437,
+      "learning_rate": 0.00019501421453353016,
+      "loss": 1.1727,
+      "step": 1075
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6154651045799255,
+      "learning_rate": 0.0001950048378506509,
+      "loss": 1.2673,
+      "step": 1076
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5188403725624084,
+      "learning_rate": 0.00019499545258462245,
+      "loss": 1.0785,
+      "step": 1077
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5489109754562378,
+      "learning_rate": 0.00019498605873629273,
+      "loss": 1.0982,
+      "step": 1078
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5472975373268127,
+      "learning_rate": 0.00019497665630651044,
+      "loss": 0.9751,
+      "step": 1079
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5919507145881653,
+      "learning_rate": 0.00019496724529612498,
+      "loss": 1.0141,
+      "step": 1080
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6334194540977478,
+      "learning_rate": 0.00019495782570598663,
+      "loss": 1.381,
+      "step": 1081
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5649259090423584,
+      "learning_rate": 0.00019494839753694633,
+      "loss": 1.2031,
+      "step": 1082
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.532267153263092,
+      "learning_rate": 0.00019493896078985596,
+      "loss": 1.1422,
+      "step": 1083
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5439337491989136,
+      "learning_rate": 0.000194929515465568,
+      "loss": 1.0401,
+      "step": 1084
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.59439617395401,
+      "learning_rate": 0.00019492006156493583,
+      "loss": 1.0084,
+      "step": 1085
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5659463405609131,
+      "learning_rate": 0.0001949105990888135,
+      "loss": 1.194,
+      "step": 1086
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.504042387008667,
+      "learning_rate": 0.00019490112803805597,
+      "loss": 1.1048,
+      "step": 1087
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4883994460105896,
+      "learning_rate": 0.0001948916484135188,
+      "loss": 0.938,
+      "step": 1088
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5720723867416382,
+      "learning_rate": 0.00019488216021605853,
+      "loss": 1.1989,
+      "step": 1089
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5613930821418762,
+      "learning_rate": 0.00019487266344653224,
+      "loss": 1.0445,
+      "step": 1090
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5611872673034668,
+      "learning_rate": 0.00019486315810579802,
+      "loss": 1.0526,
+      "step": 1091
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5774508714675903,
+      "learning_rate": 0.00019485364419471454,
+      "loss": 1.299,
+      "step": 1092
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5790034532546997,
+      "learning_rate": 0.0001948441217141414,
+      "loss": 1.0856,
+      "step": 1093
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6689309477806091,
+      "learning_rate": 0.00019483459066493887,
+      "loss": 1.1007,
+      "step": 1094
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5516353249549866,
+      "learning_rate": 0.00019482505104796802,
+      "loss": 1.1348,
+      "step": 1095
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5340696573257446,
+      "learning_rate": 0.00019481550286409068,
+      "loss": 1.1177,
+      "step": 1096
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5577419400215149,
+      "learning_rate": 0.00019480594611416953,
+      "loss": 1.0439,
+      "step": 1097
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.536791980266571,
+      "learning_rate": 0.00019479638079906793,
+      "loss": 1.0458,
+      "step": 1098
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6502873301506042,
+      "learning_rate": 0.0001947868069196501,
+      "loss": 1.0867,
+      "step": 1099
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5620083212852478,
+      "learning_rate": 0.0001947772244767809,
+      "loss": 1.0497,
+      "step": 1100
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5127768516540527,
+      "learning_rate": 0.00019476763347132608,
+      "loss": 1.0542,
+      "step": 1101
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.659235417842865,
+      "learning_rate": 0.0001947580339041522,
+      "loss": 1.2275,
+      "step": 1102
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6155188083648682,
+      "learning_rate": 0.00019474842577612645,
+      "loss": 1.0007,
+      "step": 1103
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5258010029792786,
+      "learning_rate": 0.0001947388090881169,
+      "loss": 1.0907,
+      "step": 1104
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5436065793037415,
+      "learning_rate": 0.00019472918384099233,
+      "loss": 0.8972,
+      "step": 1105
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5574349164962769,
+      "learning_rate": 0.0001947195500356224,
+      "loss": 1.0799,
+      "step": 1106
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5468463897705078,
+      "learning_rate": 0.00019470990767287737,
+      "loss": 1.0913,
+      "step": 1107
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5177785754203796,
+      "learning_rate": 0.00019470025675362847,
+      "loss": 1.0303,
+      "step": 1108
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4689900279045105,
+      "learning_rate": 0.00019469059727874756,
+      "loss": 1.0365,
+      "step": 1109
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5466490983963013,
+      "learning_rate": 0.00019468092924910732,
+      "loss": 1.1891,
+      "step": 1110
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5339245796203613,
+      "learning_rate": 0.00019467125266558119,
+      "loss": 1.0907,
+      "step": 1111
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5850642323493958,
+      "learning_rate": 0.00019466156752904343,
+      "loss": 1.1258,
+      "step": 1112
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5640799403190613,
+      "learning_rate": 0.000194651873840369,
+      "loss": 1.0536,
+      "step": 1113
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5763959884643555,
+      "learning_rate": 0.0001946421716004337,
+      "loss": 1.1633,
+      "step": 1114
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5479110479354858,
+      "learning_rate": 0.00019463246081011403,
+      "loss": 1.1711,
+      "step": 1115
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5160706043243408,
+      "learning_rate": 0.00019462274147028738,
+      "loss": 1.0581,
+      "step": 1116
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.620171844959259,
+      "learning_rate": 0.00019461301358183176,
+      "loss": 1.0797,
+      "step": 1117
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6158525347709656,
+      "learning_rate": 0.00019460327714562606,
+      "loss": 1.1704,
+      "step": 1118
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5333216190338135,
+      "learning_rate": 0.0001945935321625499,
+      "loss": 1.1178,
+      "step": 1119
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4954620599746704,
+      "learning_rate": 0.00019458377863348374,
+      "loss": 1.2292,
+      "step": 1120
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.573711633682251,
+      "learning_rate": 0.00019457401655930868,
+      "loss": 1.0373,
+      "step": 1121
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.513664186000824,
+      "learning_rate": 0.0001945642459409067,
+      "loss": 0.8973,
+      "step": 1122
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5909994840621948,
+      "learning_rate": 0.0001945544667791605,
+      "loss": 1.196,
+      "step": 1123
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6800118088722229,
+      "learning_rate": 0.00019454467907495363,
+      "loss": 1.1782,
+      "step": 1124
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6476559042930603,
+      "learning_rate": 0.00019453488282917031,
+      "loss": 1.1332,
+      "step": 1125
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5529746413230896,
+      "learning_rate": 0.00019452507804269557,
+      "loss": 1.1508,
+      "step": 1126
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5596948266029358,
+      "learning_rate": 0.00019451526471641527,
+      "loss": 1.1156,
+      "step": 1127
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6035382151603699,
+      "learning_rate": 0.0001945054428512159,
+      "loss": 1.1644,
+      "step": 1128
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5915977954864502,
+      "learning_rate": 0.00019449561244798486,
+      "loss": 1.1096,
+      "step": 1129
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.511471688747406,
+      "learning_rate": 0.00019448577350761028,
+      "loss": 0.9777,
+      "step": 1130
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5073311924934387,
+      "learning_rate": 0.00019447592603098103,
+      "loss": 0.9382,
+      "step": 1131
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5543911457061768,
+      "learning_rate": 0.0001944660700189868,
+      "loss": 0.9666,
+      "step": 1132
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5058267712593079,
+      "learning_rate": 0.00019445620547251798,
+      "loss": 1.109,
+      "step": 1133
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5585498809814453,
+      "learning_rate": 0.00019444633239246585,
+      "loss": 1.2121,
+      "step": 1134
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5616839528083801,
+      "learning_rate": 0.0001944364507797223,
+      "loss": 1.052,
+      "step": 1135
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5528122782707214,
+      "learning_rate": 0.00019442656063518014,
+      "loss": 1.1647,
+      "step": 1136
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5317437648773193,
+      "learning_rate": 0.00019441666195973287,
+      "loss": 1.2549,
+      "step": 1137
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5605677366256714,
+      "learning_rate": 0.00019440675475427478,
+      "loss": 1.1905,
+      "step": 1138
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6127129197120667,
+      "learning_rate": 0.00019439683901970094,
+      "loss": 1.1418,
+      "step": 1139
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6533770561218262,
+      "learning_rate": 0.00019438691475690714,
+      "loss": 1.1403,
+      "step": 1140
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5641628503799438,
+      "learning_rate": 0.00019437698196679,
+      "loss": 1.134,
+      "step": 1141
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6538822054862976,
+      "learning_rate": 0.00019436704065024693,
+      "loss": 1.1597,
+      "step": 1142
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5727140307426453,
+      "learning_rate": 0.00019435709080817604,
+      "loss": 0.9952,
+      "step": 1143
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5956428647041321,
+      "learning_rate": 0.00019434713244147623,
+      "loss": 1.0015,
+      "step": 1144
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6197925806045532,
+      "learning_rate": 0.00019433716555104723,
+      "loss": 1.1185,
+      "step": 1145
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5413385033607483,
+      "learning_rate": 0.00019432719013778938,
+      "loss": 1.1774,
+      "step": 1146
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5331758260726929,
+      "learning_rate": 0.00019431720620260404,
+      "loss": 1.222,
+      "step": 1147
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5929687023162842,
+      "learning_rate": 0.00019430721374639312,
+      "loss": 0.971,
+      "step": 1148
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6004317402839661,
+      "learning_rate": 0.00019429721277005942,
+      "loss": 1.0906,
+      "step": 1149
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5036609768867493,
+      "learning_rate": 0.00019428720327450649,
+      "loss": 1.0425,
+      "step": 1150
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5406267046928406,
+      "learning_rate": 0.00019427718526063856,
+      "loss": 1.0313,
+      "step": 1151
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5303625464439392,
+      "learning_rate": 0.0001942671587293607,
+      "loss": 0.9847,
+      "step": 1152
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4866414964199066,
+      "learning_rate": 0.00019425712368157885,
+      "loss": 1.0592,
+      "step": 1153
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4721148908138275,
+      "learning_rate": 0.00019424708011819952,
+      "loss": 0.9647,
+      "step": 1154
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.551730215549469,
+      "learning_rate": 0.00019423702804013013,
+      "loss": 1.1571,
+      "step": 1155
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5729138851165771,
+      "learning_rate": 0.00019422696744827883,
+      "loss": 0.969,
+      "step": 1156
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5567776560783386,
+      "learning_rate": 0.00019421689834355456,
+      "loss": 0.9652,
+      "step": 1157
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5598708391189575,
+      "learning_rate": 0.00019420682072686696,
+      "loss": 1.1677,
+      "step": 1158
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5344229340553284,
+      "learning_rate": 0.0001941967345991265,
+      "loss": 1.0504,
+      "step": 1159
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5083239078521729,
+      "learning_rate": 0.00019418663996124443,
+      "loss": 1.1487,
+      "step": 1160
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5367764234542847,
+      "learning_rate": 0.00019417653681413267,
+      "loss": 1.0572,
+      "step": 1161
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5423426032066345,
+      "learning_rate": 0.0001941664251587041,
+      "loss": 1.1104,
+      "step": 1162
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6547486782073975,
+      "learning_rate": 0.0001941563049958721,
+      "loss": 1.0926,
+      "step": 1163
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.535735011100769,
+      "learning_rate": 0.00019414617632655115,
+      "loss": 0.9657,
+      "step": 1164
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5491278171539307,
+      "learning_rate": 0.0001941360391516562,
+      "loss": 1.0038,
+      "step": 1165
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6106647253036499,
+      "learning_rate": 0.0001941258934721031,
+      "loss": 1.1907,
+      "step": 1166
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5045748949050903,
+      "learning_rate": 0.00019411573928880844,
+      "loss": 0.9276,
+      "step": 1167
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5820029377937317,
+      "learning_rate": 0.0001941055766026896,
+      "loss": 1.0474,
+      "step": 1168
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5021175146102905,
+      "learning_rate": 0.0001940954054146648,
+      "loss": 0.9492,
+      "step": 1169
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5358083844184875,
+      "learning_rate": 0.00019408522572565286,
+      "loss": 0.9982,
+      "step": 1170
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.667941153049469,
+      "learning_rate": 0.00019407503753657345,
+      "loss": 1.1492,
+      "step": 1171
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5163488984107971,
+      "learning_rate": 0.00019406484084834712,
+      "loss": 0.9993,
+      "step": 1172
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5500897169113159,
+      "learning_rate": 0.00019405463566189496,
+      "loss": 0.9547,
+      "step": 1173
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4756630063056946,
+      "learning_rate": 0.00019404442197813902,
+      "loss": 0.778,
+      "step": 1174
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6951075196266174,
+      "learning_rate": 0.00019403419979800203,
+      "loss": 1.1922,
+      "step": 1175
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6551975011825562,
+      "learning_rate": 0.00019402396912240752,
+      "loss": 1.141,
+      "step": 1176
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5312485694885254,
+      "learning_rate": 0.00019401372995227976,
+      "loss": 1.0688,
+      "step": 1177
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.572547197341919,
+      "learning_rate": 0.00019400348228854378,
+      "loss": 1.0993,
+      "step": 1178
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5445220470428467,
+      "learning_rate": 0.00019399322613212545,
+      "loss": 1.0595,
+      "step": 1179
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.49804481863975525,
+      "learning_rate": 0.0001939829614839513,
+      "loss": 1.1169,
+      "step": 1180
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5678784251213074,
+      "learning_rate": 0.00019397268834494874,
+      "loss": 1.1402,
+      "step": 1181
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.511607825756073,
+      "learning_rate": 0.00019396240671604585,
+      "loss": 1.1218,
+      "step": 1182
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5115561485290527,
+      "learning_rate": 0.00019395211659817152,
+      "loss": 0.9857,
+      "step": 1183
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4995981454849243,
+      "learning_rate": 0.0001939418179922554,
+      "loss": 1.1512,
+      "step": 1184
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5507586598396301,
+      "learning_rate": 0.00019393151089922795,
+      "loss": 1.0724,
+      "step": 1185
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5785311460494995,
+      "learning_rate": 0.0001939211953200203,
+      "loss": 1.1254,
+      "step": 1186
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5976409316062927,
+      "learning_rate": 0.00019391087125556446,
+      "loss": 1.191,
+      "step": 1187
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6236246824264526,
+      "learning_rate": 0.00019390053870679312,
+      "loss": 1.2168,
+      "step": 1188
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5489307641983032,
+      "learning_rate": 0.00019389019767463976,
+      "loss": 1.1693,
+      "step": 1189
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5228361487388611,
+      "learning_rate": 0.00019387984816003867,
+      "loss": 0.9666,
+      "step": 1190
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6135544180870056,
+      "learning_rate": 0.0001938694901639248,
+      "loss": 0.9748,
+      "step": 1191
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5711922645568848,
+      "learning_rate": 0.00019385912368723401,
+      "loss": 1.129,
+      "step": 1192
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5178403258323669,
+      "learning_rate": 0.00019384874873090282,
+      "loss": 1.0801,
+      "step": 1193
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5309805870056152,
+      "learning_rate": 0.00019383836529586853,
+      "loss": 1.2445,
+      "step": 1194
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5460125207901001,
+      "learning_rate": 0.00019382797338306928,
+      "loss": 0.9797,
+      "step": 1195
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5426638126373291,
+      "learning_rate": 0.00019381757299344386,
+      "loss": 1.181,
+      "step": 1196
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4527784585952759,
+      "learning_rate": 0.00019380716412793194,
+      "loss": 0.7509,
+      "step": 1197
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.49321427941322327,
+      "learning_rate": 0.00019379674678747386,
+      "loss": 0.9967,
+      "step": 1198
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5492578744888306,
+      "learning_rate": 0.0001937863209730108,
+      "loss": 0.9134,
+      "step": 1199
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5656059384346008,
+      "learning_rate": 0.00019377588668548462,
+      "loss": 1.2179,
+      "step": 1200
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5511481761932373,
+      "learning_rate": 0.00019376544392583808,
+      "loss": 1.2019,
+      "step": 1201
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5833784341812134,
+      "learning_rate": 0.00019375499269501457,
+      "loss": 1.2725,
+      "step": 1202
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5577043890953064,
+      "learning_rate": 0.0001937445329939583,
+      "loss": 1.0354,
+      "step": 1203
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5270592570304871,
+      "learning_rate": 0.00019373406482361424,
+      "loss": 1.083,
+      "step": 1204
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.652744472026825,
+      "learning_rate": 0.00019372358818492815,
+      "loss": 1.2333,
+      "step": 1205
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5374244451522827,
+      "learning_rate": 0.00019371310307884654,
+      "loss": 1.1342,
+      "step": 1206
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5085197687149048,
+      "learning_rate": 0.00019370260950631667,
+      "loss": 1.0782,
+      "step": 1207
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5113154053688049,
+      "learning_rate": 0.00019369210746828658,
+      "loss": 1.0251,
+      "step": 1208
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5512399673461914,
+      "learning_rate": 0.00019368159696570505,
+      "loss": 1.0595,
+      "step": 1209
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.533384382724762,
+      "learning_rate": 0.00019367107799952166,
+      "loss": 1.2031,
+      "step": 1210
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4948982894420624,
+      "learning_rate": 0.00019366055057068674,
+      "loss": 1.0376,
+      "step": 1211
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5904738306999207,
+      "learning_rate": 0.00019365001468015137,
+      "loss": 1.1022,
+      "step": 1212
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6015186309814453,
+      "learning_rate": 0.00019363947032886744,
+      "loss": 1.2163,
+      "step": 1213
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5421590209007263,
+      "learning_rate": 0.00019362891751778755,
+      "loss": 1.0019,
+      "step": 1214
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5972522497177124,
+      "learning_rate": 0.00019361835624786506,
+      "loss": 1.0744,
+      "step": 1215
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5512943267822266,
+      "learning_rate": 0.00019360778652005416,
+      "loss": 1.0013,
+      "step": 1216
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6016600728034973,
+      "learning_rate": 0.00019359720833530977,
+      "loss": 1.1387,
+      "step": 1217
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5378842353820801,
+      "learning_rate": 0.00019358662169458753,
+      "loss": 1.0681,
+      "step": 1218
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5843987464904785,
+      "learning_rate": 0.00019357602659884394,
+      "loss": 1.0471,
+      "step": 1219
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5850900411605835,
+      "learning_rate": 0.00019356542304903613,
+      "loss": 1.0736,
+      "step": 1220
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5504021644592285,
+      "learning_rate": 0.00019355481104612213,
+      "loss": 1.06,
+      "step": 1221
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7328195571899414,
+      "learning_rate": 0.00019354419059106065,
+      "loss": 1.2731,
+      "step": 1222
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5825977921485901,
+      "learning_rate": 0.00019353356168481123,
+      "loss": 1.0629,
+      "step": 1223
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.572723388671875,
+      "learning_rate": 0.00019352292432833407,
+      "loss": 0.8973,
+      "step": 1224
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.526881217956543,
+      "learning_rate": 0.00019351227852259023,
+      "loss": 0.9208,
+      "step": 1225
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5763203501701355,
+      "learning_rate": 0.0001935016242685415,
+      "loss": 1.1306,
+      "step": 1226
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5341783761978149,
+      "learning_rate": 0.00019349096156715039,
+      "loss": 1.143,
+      "step": 1227
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5604948997497559,
+      "learning_rate": 0.0001934802904193803,
+      "loss": 1.2042,
+      "step": 1228
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5423800349235535,
+      "learning_rate": 0.00019346961082619522,
+      "loss": 0.9875,
+      "step": 1229
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5558487176895142,
+      "learning_rate": 0.00019345892278856,
+      "loss": 0.8975,
+      "step": 1230
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.48725664615631104,
+      "learning_rate": 0.0001934482263074403,
+      "loss": 0.9847,
+      "step": 1231
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7110666036605835,
+      "learning_rate": 0.00019343752138380245,
+      "loss": 1.2409,
+      "step": 1232
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5547020435333252,
+      "learning_rate": 0.0001934268080186136,
+      "loss": 1.0696,
+      "step": 1233
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6257910132408142,
+      "learning_rate": 0.00019341608621284163,
+      "loss": 1.1821,
+      "step": 1234
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4753723740577698,
+      "learning_rate": 0.00019340535596745515,
+      "loss": 0.9012,
+      "step": 1235
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5675914883613586,
+      "learning_rate": 0.00019339461728342365,
+      "loss": 1.1259,
+      "step": 1236
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.47405683994293213,
+      "learning_rate": 0.00019338387016171727,
+      "loss": 1.0401,
+      "step": 1237
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6489463448524475,
+      "learning_rate": 0.00019337311460330695,
+      "loss": 1.1318,
+      "step": 1238
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.610980212688446,
+      "learning_rate": 0.0001933623506091644,
+      "loss": 1.273,
+      "step": 1239
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.622636616230011,
+      "learning_rate": 0.00019335157818026213,
+      "loss": 0.9558,
+      "step": 1240
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5120880603790283,
+      "learning_rate": 0.00019334079731757328,
+      "loss": 1.0607,
+      "step": 1241
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5910044312477112,
+      "learning_rate": 0.0001933300080220719,
+      "loss": 1.182,
+      "step": 1242
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6695639491081238,
+      "learning_rate": 0.00019331921029473276,
+      "loss": 1.2554,
+      "step": 1243
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5685967206954956,
+      "learning_rate": 0.00019330840413653131,
+      "loss": 1.0963,
+      "step": 1244
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5170540809631348,
+      "learning_rate": 0.00019329758954844385,
+      "loss": 1.0808,
+      "step": 1245
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.561914324760437,
+      "learning_rate": 0.00019328676653144743,
+      "loss": 1.1171,
+      "step": 1246
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5721147656440735,
+      "learning_rate": 0.00019327593508651985,
+      "loss": 1.0033,
+      "step": 1247
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.616117000579834,
+      "learning_rate": 0.00019326509521463967,
+      "loss": 1.2112,
+      "step": 1248
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.590916633605957,
+      "learning_rate": 0.0001932542469167862,
+      "loss": 1.2606,
+      "step": 1249
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5014984011650085,
+      "learning_rate": 0.00019324339019393948,
+      "loss": 1.0264,
+      "step": 1250
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.536072313785553,
+      "learning_rate": 0.00019323252504708046,
+      "loss": 0.925,
+      "step": 1251
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6966559290885925,
+      "learning_rate": 0.00019322165147719065,
+      "loss": 1.2475,
+      "step": 1252
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5732153058052063,
+      "learning_rate": 0.00019321076948525248,
+      "loss": 1.2821,
+      "step": 1253
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5714349150657654,
+      "learning_rate": 0.00019319987907224899,
+      "loss": 0.9969,
+      "step": 1254
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.49318331480026245,
+      "learning_rate": 0.00019318898023916415,
+      "loss": 1.1031,
+      "step": 1255
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5395241975784302,
+      "learning_rate": 0.00019317807298698257,
+      "loss": 1.1954,
+      "step": 1256
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5213454961776733,
+      "learning_rate": 0.00019316715731668973,
+      "loss": 1.0166,
+      "step": 1257
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5459577441215515,
+      "learning_rate": 0.0001931562332292717,
+      "loss": 1.0798,
+      "step": 1258
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5936806201934814,
+      "learning_rate": 0.0001931453007257154,
+      "loss": 1.1566,
+      "step": 1259
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5617766976356506,
+      "learning_rate": 0.0001931343598070086,
+      "loss": 1.2553,
+      "step": 1260
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6054962277412415,
+      "learning_rate": 0.00019312341047413975,
+      "loss": 1.1497,
+      "step": 1261
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5415417551994324,
+      "learning_rate": 0.000193112452728098,
+      "loss": 1.0297,
+      "step": 1262
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5991696119308472,
+      "learning_rate": 0.00019310148656987333,
+      "loss": 1.0435,
+      "step": 1263
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6084467768669128,
+      "learning_rate": 0.00019309051200045653,
+      "loss": 1.1188,
+      "step": 1264
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6937548518180847,
+      "learning_rate": 0.000193079529020839,
+      "loss": 1.1313,
+      "step": 1265
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6291505098342896,
+      "learning_rate": 0.00019306853763201303,
+      "loss": 1.1127,
+      "step": 1266
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5899692177772522,
+      "learning_rate": 0.0001930575378349717,
+      "loss": 1.0954,
+      "step": 1267
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5272820591926575,
+      "learning_rate": 0.0001930465296307087,
+      "loss": 1.2213,
+      "step": 1268
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5450857877731323,
+      "learning_rate": 0.00019303551302021852,
+      "loss": 1.1929,
+      "step": 1269
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.527804434299469,
+      "learning_rate": 0.00019302448800449654,
+      "loss": 1.12,
+      "step": 1270
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5344571471214294,
+      "learning_rate": 0.00019301345458453878,
+      "loss": 1.1377,
+      "step": 1271
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5595630407333374,
+      "learning_rate": 0.00019300241276134204,
+      "loss": 1.1037,
+      "step": 1272
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5389091968536377,
+      "learning_rate": 0.00019299136253590388,
+      "loss": 0.9875,
+      "step": 1273
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4910154640674591,
+      "learning_rate": 0.00019298030390922264,
+      "loss": 1.0903,
+      "step": 1274
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5613997578620911,
+      "learning_rate": 0.0001929692368822974,
+      "loss": 1.0653,
+      "step": 1275
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5709872245788574,
+      "learning_rate": 0.00019295816145612795,
+      "loss": 1.0943,
+      "step": 1276
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6014252305030823,
+      "learning_rate": 0.000192947077631715,
+      "loss": 1.1494,
+      "step": 1277
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6745042204856873,
+      "learning_rate": 0.00019293598541005982,
+      "loss": 1.0209,
+      "step": 1278
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5821531414985657,
+      "learning_rate": 0.00019292488479216458,
+      "loss": 1.1824,
+      "step": 1279
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5291196703910828,
+      "learning_rate": 0.00019291377577903212,
+      "loss": 1.1749,
+      "step": 1280
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6035329699516296,
+      "learning_rate": 0.00019290265837166608,
+      "loss": 1.1746,
+      "step": 1281
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.582727313041687,
+      "learning_rate": 0.0001928915325710709,
+      "loss": 1.1319,
+      "step": 1282
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5894931554794312,
+      "learning_rate": 0.00019288039837825168,
+      "loss": 1.1792,
+      "step": 1283
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5747166275978088,
+      "learning_rate": 0.00019286925579421434,
+      "loss": 1.1439,
+      "step": 1284
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5466634631156921,
+      "learning_rate": 0.00019285810481996563,
+      "loss": 1.0759,
+      "step": 1285
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5804960131645203,
+      "learning_rate": 0.00019284694545651285,
+      "loss": 1.1445,
+      "step": 1286
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5688068270683289,
+      "learning_rate": 0.00019283577770486425,
+      "loss": 1.2065,
+      "step": 1287
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5299245715141296,
+      "learning_rate": 0.00019282460156602878,
+      "loss": 1.2183,
+      "step": 1288
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5362908244132996,
+      "learning_rate": 0.00019281341704101615,
+      "loss": 1.1709,
+      "step": 1289
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5553381443023682,
+      "learning_rate": 0.0001928022241308368,
+      "loss": 1.0166,
+      "step": 1290
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5688846111297607,
+      "learning_rate": 0.0001927910228365019,
+      "loss": 1.1638,
+      "step": 1291
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5960063934326172,
+      "learning_rate": 0.00019277981315902348,
+      "loss": 1.1656,
+      "step": 1292
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5050798058509827,
+      "learning_rate": 0.0001927685950994143,
+      "loss": 1.0812,
+      "step": 1293
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6014881730079651,
+      "learning_rate": 0.00019275736865868778,
+      "loss": 1.0961,
+      "step": 1294
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5374804139137268,
+      "learning_rate": 0.00019274613383785818,
+      "loss": 1.181,
+      "step": 1295
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5485164523124695,
+      "learning_rate": 0.00019273489063794056,
+      "loss": 1.2842,
+      "step": 1296
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6008142232894897,
+      "learning_rate": 0.00019272363905995058,
+      "loss": 0.9858,
+      "step": 1297
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5489171743392944,
+      "learning_rate": 0.00019271237910490485,
+      "loss": 0.9488,
+      "step": 1298
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5818604826927185,
+      "learning_rate": 0.00019270111077382057,
+      "loss": 1.0722,
+      "step": 1299
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5716251730918884,
+      "learning_rate": 0.0001926898340677158,
+      "loss": 1.103,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5115184783935547,
+      "learning_rate": 0.0001926785489876094,
+      "loss": 1.1094,
+      "step": 1301
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.571029007434845,
+      "learning_rate": 0.00019266725553452075,
+      "loss": 1.0811,
+      "step": 1302
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5410216450691223,
+      "learning_rate": 0.00019265595370947033,
+      "loss": 1.0502,
+      "step": 1303
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.567905068397522,
+      "learning_rate": 0.00019264464351347903,
+      "loss": 1.1384,
+      "step": 1304
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5957036018371582,
+      "learning_rate": 0.0001926333249475688,
+      "loss": 0.9977,
+      "step": 1305
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5629892945289612,
+      "learning_rate": 0.00019262199801276214,
+      "loss": 1.1238,
+      "step": 1306
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5770018100738525,
+      "learning_rate": 0.00019261066271008235,
+      "loss": 1.205,
+      "step": 1307
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5584204792976379,
+      "learning_rate": 0.00019259931904055358,
+      "loss": 1.3062,
+      "step": 1308
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5548158288002014,
+      "learning_rate": 0.0001925879670052006,
+      "loss": 1.1269,
+      "step": 1309
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5640186071395874,
+      "learning_rate": 0.00019257660660504906,
+      "loss": 1.107,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 0.9923078417778015,
+      "eval_runtime": 17.9573,
+      "eval_samples_per_second": 151.972,
+      "eval_steps_per_second": 38.035,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5832816362380981,
+      "learning_rate": 0.00019256523784112525,
+      "loss": 1.0619,
+      "step": 1311
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5452430844306946,
+      "learning_rate": 0.00019255386071445637,
+      "loss": 1.253,
+      "step": 1312
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5174736976623535,
+      "learning_rate": 0.00019254247522607014,
+      "loss": 1.0906,
+      "step": 1313
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5985592007637024,
+      "learning_rate": 0.0001925310813769953,
+      "loss": 1.2273,
+      "step": 1314
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5438377261161804,
+      "learning_rate": 0.00019251967916826112,
+      "loss": 1.0868,
+      "step": 1315
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5200532674789429,
+      "learning_rate": 0.0001925082686008978,
+      "loss": 1.142,
+      "step": 1316
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5317777395248413,
+      "learning_rate": 0.0001924968496759362,
+      "loss": 1.0676,
+      "step": 1317
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6042464971542358,
+      "learning_rate": 0.00019248542239440791,
+      "loss": 1.1256,
+      "step": 1318
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5626322031021118,
+      "learning_rate": 0.00019247398675734537,
+      "loss": 1.0569,
+      "step": 1319
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.58858323097229,
+      "learning_rate": 0.00019246254276578174,
+      "loss": 1.2068,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5406404733657837,
+      "learning_rate": 0.00019245109042075086,
+      "loss": 1.074,
+      "step": 1321
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6096088290214539,
+      "learning_rate": 0.00019243962972328746,
+      "loss": 1.1028,
+      "step": 1322
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.606028139591217,
+      "learning_rate": 0.00019242816067442688,
+      "loss": 1.0781,
+      "step": 1323
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5860372185707092,
+      "learning_rate": 0.00019241668327520532,
+      "loss": 1.0578,
+      "step": 1324
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5592902302742004,
+      "learning_rate": 0.00019240519752665967,
+      "loss": 1.2027,
+      "step": 1325
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5435681939125061,
+      "learning_rate": 0.00019239370342982764,
+      "loss": 0.9271,
+      "step": 1326
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5579676032066345,
+      "learning_rate": 0.0001923822009857476,
+      "loss": 1.0776,
+      "step": 1327
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6110450029373169,
+      "learning_rate": 0.0001923706901954588,
+      "loss": 1.1934,
+      "step": 1328
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5760442018508911,
+      "learning_rate": 0.00019235917106000114,
+      "loss": 1.1367,
+      "step": 1329
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6102426648139954,
+      "learning_rate": 0.00019234764358041533,
+      "loss": 1.161,
+      "step": 1330
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5880422592163086,
+      "learning_rate": 0.00019233610775774276,
+      "loss": 1.2247,
+      "step": 1331
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5817332863807678,
+      "learning_rate": 0.00019232456359302567,
+      "loss": 1.2372,
+      "step": 1332
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5220525860786438,
+      "learning_rate": 0.000192313011087307,
+      "loss": 1.181,
+      "step": 1333
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5194857716560364,
+      "learning_rate": 0.00019230145024163044,
+      "loss": 0.9651,
+      "step": 1334
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5529868602752686,
+      "learning_rate": 0.00019228988105704048,
+      "loss": 1.0857,
+      "step": 1335
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5783809423446655,
+      "learning_rate": 0.0001922783035345823,
+      "loss": 1.1625,
+      "step": 1336
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6270204186439514,
+      "learning_rate": 0.00019226671767530186,
+      "loss": 1.2083,
+      "step": 1337
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5166578888893127,
+      "learning_rate": 0.00019225512348024592,
+      "loss": 1.1125,
+      "step": 1338
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5585746765136719,
+      "learning_rate": 0.00019224352095046192,
+      "loss": 1.2018,
+      "step": 1339
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5650426149368286,
+      "learning_rate": 0.00019223191008699807,
+      "loss": 0.9949,
+      "step": 1340
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.536048412322998,
+      "learning_rate": 0.00019222029089090335,
+      "loss": 1.1063,
+      "step": 1341
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5900925993919373,
+      "learning_rate": 0.00019220866336322752,
+      "loss": 1.1525,
+      "step": 1342
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.48170042037963867,
+      "learning_rate": 0.000192197027505021,
+      "loss": 1.0483,
+      "step": 1343
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6003119349479675,
+      "learning_rate": 0.00019218538331733514,
+      "loss": 1.0138,
+      "step": 1344
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5208913683891296,
+      "learning_rate": 0.00019217373080122177,
+      "loss": 1.0371,
+      "step": 1345
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5866532325744629,
+      "learning_rate": 0.00019216206995773373,
+      "loss": 1.1965,
+      "step": 1346
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5789579153060913,
+      "learning_rate": 0.00019215040078792453,
+      "loss": 1.1518,
+      "step": 1347
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5461020469665527,
+      "learning_rate": 0.00019213872329284834,
+      "loss": 1.1143,
+      "step": 1348
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6083654761314392,
+      "learning_rate": 0.00019212703747356017,
+      "loss": 1.1624,
+      "step": 1349
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.48819878697395325,
+      "learning_rate": 0.00019211534333111582,
+      "loss": 0.8658,
+      "step": 1350
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.614565372467041,
+      "learning_rate": 0.00019210364086657172,
+      "loss": 1.0509,
+      "step": 1351
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6060966849327087,
+      "learning_rate": 0.0001920919300809852,
+      "loss": 1.0878,
+      "step": 1352
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4852564036846161,
+      "learning_rate": 0.00019208021097541419,
+      "loss": 0.9189,
+      "step": 1353
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6118038892745972,
+      "learning_rate": 0.00019206848355091748,
+      "loss": 1.1956,
+      "step": 1354
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5745032429695129,
+      "learning_rate": 0.0001920567478085546,
+      "loss": 1.1221,
+      "step": 1355
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6071743965148926,
+      "learning_rate": 0.00019204500374938577,
+      "loss": 1.0735,
+      "step": 1356
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.48076263070106506,
+      "learning_rate": 0.000192033251374472,
+      "loss": 1.0431,
+      "step": 1357
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6090218424797058,
+      "learning_rate": 0.00019202149068487504,
+      "loss": 1.0099,
+      "step": 1358
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5336397886276245,
+      "learning_rate": 0.00019200972168165742,
+      "loss": 1.0205,
+      "step": 1359
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5435861945152283,
+      "learning_rate": 0.00019199794436588243,
+      "loss": 1.0421,
+      "step": 1360
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6408600211143494,
+      "learning_rate": 0.00019198615873861405,
+      "loss": 1.1427,
+      "step": 1361
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5759676098823547,
+      "learning_rate": 0.00019197436480091704,
+      "loss": 1.0883,
+      "step": 1362
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5735880732536316,
+      "learning_rate": 0.00019196256255385693,
+      "loss": 1.0093,
+      "step": 1363
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5333439111709595,
+      "learning_rate": 0.0001919507519985,
+      "loss": 1.1154,
+      "step": 1364
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5859337449073792,
+      "learning_rate": 0.00019193893313591324,
+      "loss": 1.1154,
+      "step": 1365
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5673715472221375,
+      "learning_rate": 0.00019192710596716443,
+      "loss": 0.9514,
+      "step": 1366
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6896955370903015,
+      "learning_rate": 0.00019191527049332207,
+      "loss": 1.1815,
+      "step": 1367
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5384703278541565,
+      "learning_rate": 0.0001919034267154555,
+      "loss": 0.9545,
+      "step": 1368
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5984772443771362,
+      "learning_rate": 0.00019189157463463463,
+      "loss": 0.9701,
+      "step": 1369
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.49715912342071533,
+      "learning_rate": 0.00019187971425193028,
+      "loss": 1.038,
+      "step": 1370
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6104459166526794,
+      "learning_rate": 0.00019186784556841398,
+      "loss": 0.9182,
+      "step": 1371
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5620800256729126,
+      "learning_rate": 0.000191855968585158,
+      "loss": 1.1721,
+      "step": 1372
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5518665909767151,
+      "learning_rate": 0.00019184408330323534,
+      "loss": 0.9945,
+      "step": 1373
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.596023440361023,
+      "learning_rate": 0.00019183218972371974,
+      "loss": 1.173,
+      "step": 1374
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5669458508491516,
+      "learning_rate": 0.0001918202878476858,
+      "loss": 1.0848,
+      "step": 1375
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.543956458568573,
+      "learning_rate": 0.00019180837767620868,
+      "loss": 1.1878,
+      "step": 1376
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.509422242641449,
+      "learning_rate": 0.00019179645921036452,
+      "loss": 1.0451,
+      "step": 1377
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5754774212837219,
+      "learning_rate": 0.00019178453245123,
+      "loss": 1.0036,
+      "step": 1378
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6185034513473511,
+      "learning_rate": 0.00019177259739988262,
+      "loss": 1.2736,
+      "step": 1379
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5642620325088501,
+      "learning_rate": 0.0001917606540574007,
+      "loss": 1.2801,
+      "step": 1380
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5372741222381592,
+      "learning_rate": 0.00019174870242486325,
+      "loss": 1.2769,
+      "step": 1381
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.575202465057373,
+      "learning_rate": 0.00019173674250335,
+      "loss": 1.1626,
+      "step": 1382
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5753564238548279,
+      "learning_rate": 0.0001917247742939415,
+      "loss": 1.0149,
+      "step": 1383
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5552440881729126,
+      "learning_rate": 0.00019171279779771897,
+      "loss": 1.0113,
+      "step": 1384
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6205223798751831,
+      "learning_rate": 0.00019170081301576444,
+      "loss": 1.2073,
+      "step": 1385
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6425143480300903,
+      "learning_rate": 0.00019168881994916065,
+      "loss": 1.1119,
+      "step": 1386
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6084074378013611,
+      "learning_rate": 0.00019167681859899114,
+      "loss": 1.1926,
+      "step": 1387
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4909365475177765,
+      "learning_rate": 0.00019166480896634012,
+      "loss": 1.0389,
+      "step": 1388
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5355455875396729,
+      "learning_rate": 0.0001916527910522926,
+      "loss": 1.0243,
+      "step": 1389
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5883185267448425,
+      "learning_rate": 0.00019164076485793437,
+      "loss": 1.3136,
+      "step": 1390
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.536249577999115,
+      "learning_rate": 0.0001916287303843519,
+      "loss": 0.9112,
+      "step": 1391
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5794864892959595,
+      "learning_rate": 0.00019161668763263245,
+      "loss": 1.204,
+      "step": 1392
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.577343761920929,
+      "learning_rate": 0.000191604636603864,
+      "loss": 1.1141,
+      "step": 1393
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4846179485321045,
+      "learning_rate": 0.00019159257729913529,
+      "loss": 1.0279,
+      "step": 1394
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5625943541526794,
+      "learning_rate": 0.00019158050971953582,
+      "loss": 0.931,
+      "step": 1395
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6180777549743652,
+      "learning_rate": 0.00019156843386615582,
+      "loss": 1.2183,
+      "step": 1396
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5064194798469543,
+      "learning_rate": 0.0001915563497400863,
+      "loss": 1.0858,
+      "step": 1397
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5856994390487671,
+      "learning_rate": 0.00019154425734241895,
+      "loss": 1.0773,
+      "step": 1398
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5765694975852966,
+      "learning_rate": 0.00019153215667424628,
+      "loss": 1.0104,
+      "step": 1399
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6079264283180237,
+      "learning_rate": 0.0001915200477366615,
+      "loss": 1.0882,
+      "step": 1400
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5938848853111267,
+      "learning_rate": 0.0001915079305307586,
+      "loss": 1.1615,
+      "step": 1401
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6206392049789429,
+      "learning_rate": 0.00019149580505763226,
+      "loss": 1.1099,
+      "step": 1402
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5335953831672668,
+      "learning_rate": 0.00019148367131837803,
+      "loss": 0.9998,
+      "step": 1403
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6532598733901978,
+      "learning_rate": 0.00019147152931409207,
+      "loss": 1.1109,
+      "step": 1404
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6051193475723267,
+      "learning_rate": 0.00019145937904587132,
+      "loss": 1.0338,
+      "step": 1405
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6793563365936279,
+      "learning_rate": 0.00019144722051481355,
+      "loss": 1.1387,
+      "step": 1406
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5717383623123169,
+      "learning_rate": 0.00019143505372201712,
+      "loss": 1.167,
+      "step": 1407
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5962218046188354,
+      "learning_rate": 0.00019142287866858135,
+      "loss": 1.1216,
+      "step": 1408
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5622634887695312,
+      "learning_rate": 0.00019141069535560612,
+      "loss": 0.702,
+      "step": 1409
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6235326528549194,
+      "learning_rate": 0.0001913985037841921,
+      "loss": 1.0167,
+      "step": 1410
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5555641055107117,
+      "learning_rate": 0.00019138630395544077,
+      "loss": 1.146,
+      "step": 1411
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.551306962966919,
+      "learning_rate": 0.00019137409587045433,
+      "loss": 1.0985,
+      "step": 1412
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5754269957542419,
+      "learning_rate": 0.00019136187953033566,
+      "loss": 1.0251,
+      "step": 1413
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5977479219436646,
+      "learning_rate": 0.00019134965493618848,
+      "loss": 1.1365,
+      "step": 1414
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6409013867378235,
+      "learning_rate": 0.00019133742208911719,
+      "loss": 1.2431,
+      "step": 1415
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5369458198547363,
+      "learning_rate": 0.00019132518099022696,
+      "loss": 1.1019,
+      "step": 1416
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5461875796318054,
+      "learning_rate": 0.00019131293164062368,
+      "loss": 0.9547,
+      "step": 1417
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5358739495277405,
+      "learning_rate": 0.0001913006740414141,
+      "loss": 1.0399,
+      "step": 1418
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6357903480529785,
+      "learning_rate": 0.00019128840819370555,
+      "loss": 1.0574,
+      "step": 1419
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5420630574226379,
+      "learning_rate": 0.00019127613409860615,
+      "loss": 1.08,
+      "step": 1420
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6549668908119202,
+      "learning_rate": 0.00019126385175722487,
+      "loss": 1.1344,
+      "step": 1421
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5863200426101685,
+      "learning_rate": 0.00019125156117067132,
+      "loss": 1.02,
+      "step": 1422
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6721362471580505,
+      "learning_rate": 0.0001912392623400559,
+      "loss": 1.1862,
+      "step": 1423
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5170490145683289,
+      "learning_rate": 0.00019122695526648968,
+      "loss": 0.9705,
+      "step": 1424
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5132006406784058,
+      "learning_rate": 0.00019121463995108463,
+      "loss": 1.1416,
+      "step": 1425
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5969366431236267,
+      "learning_rate": 0.00019120231639495328,
+      "loss": 1.2315,
+      "step": 1426
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5861543416976929,
+      "learning_rate": 0.00019118998459920902,
+      "loss": 1.1351,
+      "step": 1427
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5891702771186829,
+      "learning_rate": 0.000191177644564966,
+      "loss": 1.1291,
+      "step": 1428
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5410522222518921,
+      "learning_rate": 0.00019116529629333902,
+      "loss": 0.9958,
+      "step": 1429
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5884034633636475,
+      "learning_rate": 0.00019115293978544368,
+      "loss": 1.0791,
+      "step": 1430
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6323250532150269,
+      "learning_rate": 0.00019114057504239637,
+      "loss": 1.0047,
+      "step": 1431
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4916994273662567,
+      "learning_rate": 0.00019112820206531412,
+      "loss": 1.0319,
+      "step": 1432
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5568423271179199,
+      "learning_rate": 0.0001911158208553148,
+      "loss": 1.2002,
+      "step": 1433
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5141173601150513,
+      "learning_rate": 0.00019110343141351695,
+      "loss": 1.1509,
+      "step": 1434
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5884588956832886,
+      "learning_rate": 0.00019109103374103986,
+      "loss": 1.2095,
+      "step": 1435
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6008499264717102,
+      "learning_rate": 0.00019107862783900368,
+      "loss": 1.2357,
+      "step": 1436
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5440402626991272,
+      "learning_rate": 0.00019106621370852916,
+      "loss": 1.1466,
+      "step": 1437
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6503211855888367,
+      "learning_rate": 0.00019105379135073784,
+      "loss": 1.1827,
+      "step": 1438
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6284806132316589,
+      "learning_rate": 0.000191041360766752,
+      "loss": 1.203,
+      "step": 1439
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.563576877117157,
+      "learning_rate": 0.00019102892195769472,
+      "loss": 1.043,
+      "step": 1440
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.629399836063385,
+      "learning_rate": 0.0001910164749246897,
+      "loss": 0.9591,
+      "step": 1441
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5576425194740295,
+      "learning_rate": 0.00019100401966886153,
+      "loss": 1.1031,
+      "step": 1442
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5660919547080994,
+      "learning_rate": 0.00019099155619133545,
+      "loss": 1.0291,
+      "step": 1443
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5721517205238342,
+      "learning_rate": 0.00019097908449323744,
+      "loss": 0.9636,
+      "step": 1444
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5158771276473999,
+      "learning_rate": 0.00019096660457569432,
+      "loss": 1.0108,
+      "step": 1445
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5575356483459473,
+      "learning_rate": 0.00019095411643983348,
+      "loss": 0.9741,
+      "step": 1446
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5737133622169495,
+      "learning_rate": 0.0001909416200867832,
+      "loss": 1.0725,
+      "step": 1447
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5935455560684204,
+      "learning_rate": 0.00019092911551767247,
+      "loss": 1.2495,
+      "step": 1448
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.618424654006958,
+      "learning_rate": 0.000190916602733631,
+      "loss": 1.0883,
+      "step": 1449
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5599175095558167,
+      "learning_rate": 0.0001909040817357892,
+      "loss": 0.9976,
+      "step": 1450
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6387348175048828,
+      "learning_rate": 0.0001908915525252784,
+      "loss": 1.1933,
+      "step": 1451
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.507186233997345,
+      "learning_rate": 0.00019087901510323038,
+      "loss": 1.1166,
+      "step": 1452
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.49330127239227295,
+      "learning_rate": 0.00019086646947077794,
+      "loss": 0.8951,
+      "step": 1453
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5779321789741516,
+      "learning_rate": 0.00019085391562905443,
+      "loss": 1.1588,
+      "step": 1454
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6197642087936401,
+      "learning_rate": 0.00019084135357919408,
+      "loss": 1.0809,
+      "step": 1455
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5197485089302063,
+      "learning_rate": 0.00019082878332233176,
+      "loss": 1.0183,
+      "step": 1456
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6561455130577087,
+      "learning_rate": 0.00019081620485960317,
+      "loss": 1.1886,
+      "step": 1457
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5491870641708374,
+      "learning_rate": 0.00019080361819214466,
+      "loss": 1.1449,
+      "step": 1458
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.48696836829185486,
+      "learning_rate": 0.00019079102332109334,
+      "loss": 1.0596,
+      "step": 1459
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.62408447265625,
+      "learning_rate": 0.00019077842024758717,
+      "loss": 1.2558,
+      "step": 1460
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6218500137329102,
+      "learning_rate": 0.0001907658089727647,
+      "loss": 1.1698,
+      "step": 1461
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.558061420917511,
+      "learning_rate": 0.0001907531894977653,
+      "loss": 1.2256,
+      "step": 1462
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5486068725585938,
+      "learning_rate": 0.00019074056182372907,
+      "loss": 1.2703,
+      "step": 1463
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.501006543636322,
+      "learning_rate": 0.00019072792595179684,
+      "loss": 1.1317,
+      "step": 1464
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.562553346157074,
+      "learning_rate": 0.00019071528188311022,
+      "loss": 0.9698,
+      "step": 1465
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6265679001808167,
+      "learning_rate": 0.0001907026296188115,
+      "loss": 1.0556,
+      "step": 1466
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.554051399230957,
+      "learning_rate": 0.00019068996916004377,
+      "loss": 1.1407,
+      "step": 1467
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5753392577171326,
+      "learning_rate": 0.0001906773005079508,
+      "loss": 1.0468,
+      "step": 1468
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5394361019134521,
+      "learning_rate": 0.0001906646236636771,
+      "loss": 1.12,
+      "step": 1469
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5362756252288818,
+      "learning_rate": 0.00019065193862836805,
+      "loss": 1.0795,
+      "step": 1470
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6187636256217957,
+      "learning_rate": 0.00019063924540316956,
+      "loss": 1.0331,
+      "step": 1471
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5482291579246521,
+      "learning_rate": 0.00019062654398922853,
+      "loss": 1.0507,
+      "step": 1472
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5710556507110596,
+      "learning_rate": 0.00019061383438769228,
+      "loss": 1.2762,
+      "step": 1473
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6045145988464355,
+      "learning_rate": 0.0001906011165997092,
+      "loss": 1.1701,
+      "step": 1474
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6226998567581177,
+      "learning_rate": 0.00019058839062642823,
+      "loss": 1.1354,
+      "step": 1475
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5932821035385132,
+      "learning_rate": 0.00019057565646899907,
+      "loss": 1.2439,
+      "step": 1476
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5586472153663635,
+      "learning_rate": 0.00019056291412857218,
+      "loss": 1.0227,
+      "step": 1477
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6210311651229858,
+      "learning_rate": 0.00019055016360629873,
+      "loss": 1.237,
+      "step": 1478
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5757390260696411,
+      "learning_rate": 0.00019053740490333074,
+      "loss": 1.2567,
+      "step": 1479
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5853813290596008,
+      "learning_rate": 0.00019052463802082082,
+      "loss": 1.1551,
+      "step": 1480
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5320205688476562,
+      "learning_rate": 0.0001905118629599224,
+      "loss": 0.8199,
+      "step": 1481
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5871585607528687,
+      "learning_rate": 0.00019049907972178964,
+      "loss": 1.0795,
+      "step": 1482
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6921295523643494,
+      "learning_rate": 0.00019048628830757743,
+      "loss": 1.2403,
+      "step": 1483
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6031093597412109,
+      "learning_rate": 0.00019047348871844145,
+      "loss": 0.9655,
+      "step": 1484
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6309522390365601,
+      "learning_rate": 0.00019046068095553796,
+      "loss": 1.0757,
+      "step": 1485
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5803585648536682,
+      "learning_rate": 0.00019044786502002414,
+      "loss": 1.0815,
+      "step": 1486
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.559934675693512,
+      "learning_rate": 0.00019043504091305786,
+      "loss": 1.1025,
+      "step": 1487
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5299091339111328,
+      "learning_rate": 0.00019042220863579767,
+      "loss": 1.011,
+      "step": 1488
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6164795756340027,
+      "learning_rate": 0.00019040936818940287,
+      "loss": 1.1809,
+      "step": 1489
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6382726430892944,
+      "learning_rate": 0.0001903965195750336,
+      "loss": 1.0985,
+      "step": 1490
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5797393918037415,
+      "learning_rate": 0.00019038366279385056,
+      "loss": 0.984,
+      "step": 1491
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.516485869884491,
+      "learning_rate": 0.00019037079784701538,
+      "loss": 1.1137,
+      "step": 1492
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5419387817382812,
+      "learning_rate": 0.00019035792473569022,
+      "loss": 1.1436,
+      "step": 1493
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5802305936813354,
+      "learning_rate": 0.00019034504346103823,
+      "loss": 1.2381,
+      "step": 1494
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5675755143165588,
+      "learning_rate": 0.0001903321540242231,
+      "loss": 1.029,
+      "step": 1495
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6638576984405518,
+      "learning_rate": 0.00019031925642640926,
+      "loss": 1.0509,
+      "step": 1496
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5375914573669434,
+      "learning_rate": 0.00019030635066876204,
+      "loss": 1.1761,
+      "step": 1497
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5363042950630188,
+      "learning_rate": 0.00019029343675244732,
+      "loss": 0.9477,
+      "step": 1498
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5889085531234741,
+      "learning_rate": 0.00019028051467863185,
+      "loss": 0.9483,
+      "step": 1499
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6252742409706116,
+      "learning_rate": 0.00019026758444848306,
+      "loss": 1.1044,
+      "step": 1500
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5877997279167175,
+      "learning_rate": 0.00019025464606316908,
+      "loss": 0.9745,
+      "step": 1501
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5305163860321045,
+      "learning_rate": 0.00019024169952385885,
+      "loss": 0.9826,
+      "step": 1502
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5609629154205322,
+      "learning_rate": 0.000190228744831722,
+      "loss": 0.953,
+      "step": 1503
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5558766722679138,
+      "learning_rate": 0.00019021578198792898,
+      "loss": 1.1254,
+      "step": 1504
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5775374174118042,
+      "learning_rate": 0.0001902028109936508,
+      "loss": 0.9384,
+      "step": 1505
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5770506858825684,
+      "learning_rate": 0.0001901898318500594,
+      "loss": 1.051,
+      "step": 1506
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5940860509872437,
+      "learning_rate": 0.00019017684455832737,
+      "loss": 1.1082,
+      "step": 1507
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5455056428909302,
+      "learning_rate": 0.000190163849119628,
+      "loss": 1.1003,
+      "step": 1508
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5529404878616333,
+      "learning_rate": 0.00019015084553513533,
+      "loss": 1.1231,
+      "step": 1509
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.489065021276474,
+      "learning_rate": 0.00019013783380602422,
+      "loss": 0.9142,
+      "step": 1510
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5688076615333557,
+      "learning_rate": 0.0001901248139334702,
+      "loss": 0.9035,
+      "step": 1511
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5179052948951721,
+      "learning_rate": 0.00019011178591864952,
+      "loss": 0.978,
+      "step": 1512
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.61092609167099,
+      "learning_rate": 0.0001900987497627392,
+      "loss": 1.1268,
+      "step": 1513
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5672241449356079,
+      "learning_rate": 0.00019008570546691695,
+      "loss": 1.1325,
+      "step": 1514
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5476197600364685,
+      "learning_rate": 0.0001900726530323613,
+      "loss": 0.9854,
+      "step": 1515
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5623148679733276,
+      "learning_rate": 0.00019005959246025142,
+      "loss": 1.0171,
+      "step": 1516
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5329583287239075,
+      "learning_rate": 0.00019004652375176733,
+      "loss": 1.0977,
+      "step": 1517
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.582499086856842,
+      "learning_rate": 0.0001900334469080896,
+      "loss": 1.0895,
+      "step": 1518
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5313141345977783,
+      "learning_rate": 0.00019002036193039974,
+      "loss": 1.0889,
+      "step": 1519
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5586351752281189,
+      "learning_rate": 0.00019000726881987986,
+      "loss": 0.8848,
+      "step": 1520
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6101945638656616,
+      "learning_rate": 0.00018999416757771288,
+      "loss": 1.1436,
+      "step": 1521
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6232748627662659,
+      "learning_rate": 0.00018998105820508236,
+      "loss": 1.1319,
+      "step": 1522
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5600299835205078,
+      "learning_rate": 0.00018996794070317275,
+      "loss": 1.1113,
+      "step": 1523
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5319695472717285,
+      "learning_rate": 0.00018995481507316905,
+      "loss": 0.9104,
+      "step": 1524
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5279593467712402,
+      "learning_rate": 0.00018994168131625719,
+      "loss": 0.9743,
+      "step": 1525
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5348560214042664,
+      "learning_rate": 0.00018992853943362363,
+      "loss": 1.0488,
+      "step": 1526
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6045165061950684,
+      "learning_rate": 0.00018991538942645574,
+      "loss": 1.1206,
+      "step": 1527
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6501975059509277,
+      "learning_rate": 0.00018990223129594148,
+      "loss": 1.111,
+      "step": 1528
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5018801689147949,
+      "learning_rate": 0.00018988906504326965,
+      "loss": 1.0066,
+      "step": 1529
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5682173371315002,
+      "learning_rate": 0.00018987589066962976,
+      "loss": 1.1631,
+      "step": 1530
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6507973670959473,
+      "learning_rate": 0.00018986270817621201,
+      "loss": 0.9727,
+      "step": 1531
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5866438150405884,
+      "learning_rate": 0.00018984951756420738,
+      "loss": 1.0574,
+      "step": 1532
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6363970637321472,
+      "learning_rate": 0.00018983631883480757,
+      "loss": 1.0447,
+      "step": 1533
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5931180715560913,
+      "learning_rate": 0.000189823111989205,
+      "loss": 1.1534,
+      "step": 1534
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.560235857963562,
+      "learning_rate": 0.00018980989702859286,
+      "loss": 0.9453,
+      "step": 1535
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6191147565841675,
+      "learning_rate": 0.00018979667395416501,
+      "loss": 1.0562,
+      "step": 1536
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5215852856636047,
+      "learning_rate": 0.00018978344276711608,
+      "loss": 1.2288,
+      "step": 1537
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5872689485549927,
+      "learning_rate": 0.00018977020346864145,
+      "loss": 1.1983,
+      "step": 1538
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5544451475143433,
+      "learning_rate": 0.00018975695605993724,
+      "loss": 1.1188,
+      "step": 1539
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5902532935142517,
+      "learning_rate": 0.00018974370054220026,
+      "loss": 1.0748,
+      "step": 1540
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6012886762619019,
+      "learning_rate": 0.00018973043691662803,
+      "loss": 1.0795,
+      "step": 1541
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5270823836326599,
+      "learning_rate": 0.00018971716518441888,
+      "loss": 1.1419,
+      "step": 1542
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6644396781921387,
+      "learning_rate": 0.00018970388534677186,
+      "loss": 1.0674,
+      "step": 1543
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6086916923522949,
+      "learning_rate": 0.0001896905974048867,
+      "loss": 1.0743,
+      "step": 1544
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5799842476844788,
+      "learning_rate": 0.00018967730135996383,
+      "loss": 1.0611,
+      "step": 1545
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5766069293022156,
+      "learning_rate": 0.0001896639972132046,
+      "loss": 1.0784,
+      "step": 1546
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6158689856529236,
+      "learning_rate": 0.00018965068496581088,
+      "loss": 1.2381,
+      "step": 1547
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6311269402503967,
+      "learning_rate": 0.00018963736461898538,
+      "loss": 1.1589,
+      "step": 1548
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5501083731651306,
+      "learning_rate": 0.0001896240361739315,
+      "loss": 1.0382,
+      "step": 1549
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6331074237823486,
+      "learning_rate": 0.0001896106996318534,
+      "loss": 1.2187,
+      "step": 1550
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.531475305557251,
+      "learning_rate": 0.00018959735499395598,
+      "loss": 0.9537,
+      "step": 1551
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.69541996717453,
+      "learning_rate": 0.00018958400226144482,
+      "loss": 1.2587,
+      "step": 1552
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5847852230072021,
+      "learning_rate": 0.00018957064143552627,
+      "loss": 1.1522,
+      "step": 1553
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.549366295337677,
+      "learning_rate": 0.00018955727251740743,
+      "loss": 1.1589,
+      "step": 1554
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6188735365867615,
+      "learning_rate": 0.00018954389550829607,
+      "loss": 1.0249,
+      "step": 1555
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5889078974723816,
+      "learning_rate": 0.00018953051040940076,
+      "loss": 1.1075,
+      "step": 1556
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5530256628990173,
+      "learning_rate": 0.00018951711722193078,
+      "loss": 1.0457,
+      "step": 1557
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.586457371711731,
+      "learning_rate": 0.0001895037159470961,
+      "loss": 1.1271,
+      "step": 1558
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5959973931312561,
+      "learning_rate": 0.00018949030658610742,
+      "loss": 1.0652,
+      "step": 1559
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5624261498451233,
+      "learning_rate": 0.00018947688914017623,
+      "loss": 1.1421,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5853245258331299,
+      "learning_rate": 0.00018946346361051474,
+      "loss": 1.1081,
+      "step": 1561
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6063684225082397,
+      "learning_rate": 0.00018945002999833584,
+      "loss": 1.1567,
+      "step": 1562
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5784252882003784,
+      "learning_rate": 0.00018943658830485323,
+      "loss": 1.2009,
+      "step": 1563
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5920445322990417,
+      "learning_rate": 0.00018942313853128123,
+      "loss": 0.939,
+      "step": 1564
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5626775622367859,
+      "learning_rate": 0.00018940968067883497,
+      "loss": 1.1501,
+      "step": 1565
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.49538347125053406,
+      "learning_rate": 0.00018939621474873031,
+      "loss": 0.9118,
+      "step": 1566
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6201627254486084,
+      "learning_rate": 0.0001893827407421838,
+      "loss": 1.2706,
+      "step": 1567
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5373640656471252,
+      "learning_rate": 0.00018936925866041277,
+      "loss": 1.0574,
+      "step": 1568
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5345717072486877,
+      "learning_rate": 0.00018935576850463522,
+      "loss": 1.1574,
+      "step": 1569
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6629395484924316,
+      "learning_rate": 0.00018934227027606987,
+      "loss": 1.2365,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6796295046806335,
+      "learning_rate": 0.0001893287639759363,
+      "loss": 1.1736,
+      "step": 1571
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5906338691711426,
+      "learning_rate": 0.00018931524960545467,
+      "loss": 1.1604,
+      "step": 1572
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.66068035364151,
+      "learning_rate": 0.00018930172716584597,
+      "loss": 1.0991,
+      "step": 1573
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7229920029640198,
+      "learning_rate": 0.00018928819665833178,
+      "loss": 1.366,
+      "step": 1574
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5846989750862122,
+      "learning_rate": 0.00018927465808413465,
+      "loss": 0.9767,
+      "step": 1575
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5244808793067932,
+      "learning_rate": 0.0001892611114444776,
+      "loss": 1.1022,
+      "step": 1576
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.642883837223053,
+      "learning_rate": 0.0001892475567405845,
+      "loss": 1.058,
+      "step": 1577
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6407866477966309,
+      "learning_rate": 0.00018923399397367997,
+      "loss": 1.1835,
+      "step": 1578
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5814583897590637,
+      "learning_rate": 0.00018922042314498935,
+      "loss": 1.1512,
+      "step": 1579
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6063616871833801,
+      "learning_rate": 0.00018920684425573865,
+      "loss": 1.1334,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6133993864059448,
+      "learning_rate": 0.00018919325730715466,
+      "loss": 0.945,
+      "step": 1581
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5505120754241943,
+      "learning_rate": 0.00018917966230046492,
+      "loss": 1.1992,
+      "step": 1582
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5015380382537842,
+      "learning_rate": 0.00018916605923689756,
+      "loss": 1.0274,
+      "step": 1583
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4922814667224884,
+      "learning_rate": 0.00018915244811768168,
+      "loss": 0.9188,
+      "step": 1584
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.562036395072937,
+      "learning_rate": 0.00018913882894404683,
+      "loss": 0.9782,
+      "step": 1585
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.658740758895874,
+      "learning_rate": 0.0001891252017172235,
+      "loss": 0.9769,
+      "step": 1586
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.675102949142456,
+      "learning_rate": 0.00018911156643844283,
+      "loss": 1.2205,
+      "step": 1587
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6200870275497437,
+      "learning_rate": 0.00018909792310893667,
+      "loss": 1.1162,
+      "step": 1588
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5535531044006348,
+      "learning_rate": 0.00018908427172993766,
+      "loss": 1.1918,
+      "step": 1589
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6343311667442322,
+      "learning_rate": 0.00018907061230267905,
+      "loss": 1.0838,
+      "step": 1590
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5962023735046387,
+      "learning_rate": 0.00018905694482839497,
+      "loss": 1.0302,
+      "step": 1591
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6679428815841675,
+      "learning_rate": 0.00018904326930832015,
+      "loss": 1.1094,
+      "step": 1592
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6458826065063477,
+      "learning_rate": 0.0001890295857436901,
+      "loss": 1.2363,
+      "step": 1593
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5808925032615662,
+      "learning_rate": 0.00018901589413574107,
+      "loss": 1.1789,
+      "step": 1594
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5363585352897644,
+      "learning_rate": 0.00018900219448571003,
+      "loss": 1.0798,
+      "step": 1595
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5684288740158081,
+      "learning_rate": 0.0001889884867948346,
+      "loss": 1.047,
+      "step": 1596
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5775579214096069,
+      "learning_rate": 0.00018897477106435327,
+      "loss": 1.1357,
+      "step": 1597
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6103155016899109,
+      "learning_rate": 0.00018896104729550515,
+      "loss": 1.0776,
+      "step": 1598
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6359939575195312,
+      "learning_rate": 0.00018894731548953014,
+      "loss": 1.1528,
+      "step": 1599
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.553428590297699,
+      "learning_rate": 0.00018893357564766874,
+      "loss": 1.0318,
+      "step": 1600
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6345550417900085,
+      "learning_rate": 0.00018891982777116234,
+      "loss": 1.2425,
+      "step": 1601
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5610020756721497,
+      "learning_rate": 0.00018890607186125292,
+      "loss": 1.1793,
+      "step": 1602
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6097350716590881,
+      "learning_rate": 0.00018889230791918336,
+      "loss": 0.9796,
+      "step": 1603
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5626397132873535,
+      "learning_rate": 0.00018887853594619705,
+      "loss": 1.0767,
+      "step": 1604
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5881405472755432,
+      "learning_rate": 0.00018886475594353825,
+      "loss": 1.2583,
+      "step": 1605
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6646904349327087,
+      "learning_rate": 0.0001888509679124519,
+      "loss": 1.1146,
+      "step": 1606
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5926646590232849,
+      "learning_rate": 0.0001888371718541837,
+      "loss": 1.2153,
+      "step": 1607
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6441829204559326,
+      "learning_rate": 0.00018882336776998,
+      "loss": 1.1823,
+      "step": 1608
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5266425013542175,
+      "learning_rate": 0.0001888095556610879,
+      "loss": 0.95,
+      "step": 1609
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5980542302131653,
+      "learning_rate": 0.00018879573552875534,
+      "loss": 1.0822,
+      "step": 1610
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5426793098449707,
+      "learning_rate": 0.0001887819073742308,
+      "loss": 1.0676,
+      "step": 1611
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6400548815727234,
+      "learning_rate": 0.00018876807119876362,
+      "loss": 1.1233,
+      "step": 1612
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5350365042686462,
+      "learning_rate": 0.00018875422700360384,
+      "loss": 0.9547,
+      "step": 1613
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5893256068229675,
+      "learning_rate": 0.00018874037479000216,
+      "loss": 1.266,
+      "step": 1614
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.564774751663208,
+      "learning_rate": 0.00018872651455921006,
+      "loss": 1.0358,
+      "step": 1615
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.588233232498169,
+      "learning_rate": 0.00018871264631247972,
+      "loss": 1.1458,
+      "step": 1616
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.567189633846283,
+      "learning_rate": 0.00018869877005106407,
+      "loss": 0.9776,
+      "step": 1617
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5712881684303284,
+      "learning_rate": 0.00018868488577621678,
+      "loss": 1.1353,
+      "step": 1618
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5613961219787598,
+      "learning_rate": 0.00018867099348919217,
+      "loss": 1.1427,
+      "step": 1619
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6011113524436951,
+      "learning_rate": 0.0001886570931912454,
+      "loss": 1.1082,
+      "step": 1620
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6284022927284241,
+      "learning_rate": 0.00018864318488363217,
+      "loss": 1.148,
+      "step": 1621
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5748017430305481,
+      "learning_rate": 0.00018862926856760912,
+      "loss": 1.116,
+      "step": 1622
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5744466781616211,
+      "learning_rate": 0.0001886153442444335,
+      "loss": 1.0798,
+      "step": 1623
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5814639329910278,
+      "learning_rate": 0.00018860141191536322,
+      "loss": 1.1328,
+      "step": 1624
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5316477417945862,
+      "learning_rate": 0.00018858747158165704,
+      "loss": 0.981,
+      "step": 1625
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6174801588058472,
+      "learning_rate": 0.0001885735232445744,
+      "loss": 1.1805,
+      "step": 1626
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5454065799713135,
+      "learning_rate": 0.0001885595669053754,
+      "loss": 1.1229,
+      "step": 1627
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5038221478462219,
+      "learning_rate": 0.000188545602565321,
+      "loss": 1.0006,
+      "step": 1628
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6330316066741943,
+      "learning_rate": 0.00018853163022567273,
+      "loss": 1.0873,
+      "step": 1629
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5468506813049316,
+      "learning_rate": 0.00018851764988769295,
+      "loss": 1.034,
+      "step": 1630
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5783451795578003,
+      "learning_rate": 0.00018850366155264466,
+      "loss": 1.0622,
+      "step": 1631
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5333491563796997,
+      "learning_rate": 0.00018848966522179168,
+      "loss": 0.9803,
+      "step": 1632
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5402332544326782,
+      "learning_rate": 0.00018847566089639847,
+      "loss": 1.1328,
+      "step": 1633
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5634700059890747,
+      "learning_rate": 0.00018846164857773024,
+      "loss": 0.9754,
+      "step": 1634
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5241762399673462,
+      "learning_rate": 0.00018844762826705292,
+      "loss": 1.1019,
+      "step": 1635
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5816453099250793,
+      "learning_rate": 0.0001884335999656332,
+      "loss": 1.0907,
+      "step": 1636
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5889614224433899,
+      "learning_rate": 0.0001884195636747384,
+      "loss": 1.1334,
+      "step": 1637
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5321390628814697,
+      "learning_rate": 0.00018840551939563666,
+      "loss": 0.9717,
+      "step": 1638
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5475577712059021,
+      "learning_rate": 0.00018839146712959682,
+      "loss": 1.0735,
+      "step": 1639
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5626057386398315,
+      "learning_rate": 0.00018837740687788839,
+      "loss": 1.0199,
+      "step": 1640
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5641455054283142,
+      "learning_rate": 0.00018836333864178162,
+      "loss": 1.1475,
+      "step": 1641
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5993323922157288,
+      "learning_rate": 0.00018834926242254754,
+      "loss": 1.0956,
+      "step": 1642
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5953279733657837,
+      "learning_rate": 0.0001883351782214578,
+      "loss": 1.021,
+      "step": 1643
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5046542286872864,
+      "learning_rate": 0.0001883210860397849,
+      "loss": 0.8364,
+      "step": 1644
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6529065370559692,
+      "learning_rate": 0.00018830698587880194,
+      "loss": 1.0135,
+      "step": 1645
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6167691349983215,
+      "learning_rate": 0.00018829287773978277,
+      "loss": 1.2643,
+      "step": 1646
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5951334238052368,
+      "learning_rate": 0.00018827876162400201,
+      "loss": 1.1444,
+      "step": 1647
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.519690752029419,
+      "learning_rate": 0.000188264637532735,
+      "loss": 1.085,
+      "step": 1648
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5396617650985718,
+      "learning_rate": 0.00018825050546725773,
+      "loss": 1.1176,
+      "step": 1649
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5493987202644348,
+      "learning_rate": 0.00018823636542884695,
+      "loss": 1.0602,
+      "step": 1650
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5464897155761719,
+      "learning_rate": 0.00018822221741878018,
+      "loss": 0.8669,
+      "step": 1651
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5813970565795898,
+      "learning_rate": 0.00018820806143833555,
+      "loss": 1.0391,
+      "step": 1652
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5767297148704529,
+      "learning_rate": 0.00018819389748879202,
+      "loss": 1.0945,
+      "step": 1653
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6370941996574402,
+      "learning_rate": 0.00018817972557142918,
+      "loss": 1.1547,
+      "step": 1654
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6555666923522949,
+      "learning_rate": 0.0001881655456875274,
+      "loss": 1.1207,
+      "step": 1655
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5538688898086548,
+      "learning_rate": 0.0001881513578383678,
+      "loss": 0.9714,
+      "step": 1656
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5678582191467285,
+      "learning_rate": 0.0001881371620252321,
+      "loss": 1.2347,
+      "step": 1657
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5648280382156372,
+      "learning_rate": 0.00018812295824940285,
+      "loss": 1.0975,
+      "step": 1658
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.569893479347229,
+      "learning_rate": 0.0001881087465121633,
+      "loss": 1.0578,
+      "step": 1659
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5792575478553772,
+      "learning_rate": 0.00018809452681479733,
+      "loss": 1.0296,
+      "step": 1660
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6080823540687561,
+      "learning_rate": 0.00018808029915858967,
+      "loss": 1.2352,
+      "step": 1661
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6537753343582153,
+      "learning_rate": 0.0001880660635448257,
+      "loss": 1.1956,
+      "step": 1662
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6456838250160217,
+      "learning_rate": 0.00018805181997479152,
+      "loss": 1.1356,
+      "step": 1663
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5179287791252136,
+      "learning_rate": 0.00018803756844977394,
+      "loss": 1.0559,
+      "step": 1664
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5787820816040039,
+      "learning_rate": 0.00018802330897106051,
+      "loss": 1.1527,
+      "step": 1665
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6485100388526917,
+      "learning_rate": 0.0001880090415399395,
+      "loss": 1.1667,
+      "step": 1666
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.605714738368988,
+      "learning_rate": 0.00018799476615769992,
+      "loss": 1.2807,
+      "step": 1667
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5844491124153137,
+      "learning_rate": 0.00018798048282563144,
+      "loss": 1.1697,
+      "step": 1668
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5353244543075562,
+      "learning_rate": 0.00018796619154502447,
+      "loss": 0.9961,
+      "step": 1669
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5764732360839844,
+      "learning_rate": 0.00018795189231717014,
+      "loss": 1.0053,
+      "step": 1670
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5386641025543213,
+      "learning_rate": 0.00018793758514336033,
+      "loss": 1.1774,
+      "step": 1671
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5288216471672058,
+      "learning_rate": 0.00018792327002488762,
+      "loss": 1.0226,
+      "step": 1672
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5615891218185425,
+      "learning_rate": 0.00018790894696304525,
+      "loss": 1.0576,
+      "step": 1673
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5695824027061462,
+      "learning_rate": 0.0001878946159591273,
+      "loss": 1.1809,
+      "step": 1674
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.598285436630249,
+      "learning_rate": 0.00018788027701442843,
+      "loss": 1.0365,
+      "step": 1675
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6066529154777527,
+      "learning_rate": 0.00018786593013024412,
+      "loss": 1.0722,
+      "step": 1676
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5309323668479919,
+      "learning_rate": 0.00018785157530787052,
+      "loss": 1.0409,
+      "step": 1677
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5708740949630737,
+      "learning_rate": 0.0001878372125486045,
+      "loss": 1.1045,
+      "step": 1678
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5249610543251038,
+      "learning_rate": 0.00018782284185374368,
+      "loss": 1.0355,
+      "step": 1679
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.553139865398407,
+      "learning_rate": 0.00018780846322458632,
+      "loss": 0.9598,
+      "step": 1680
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5853700041770935,
+      "learning_rate": 0.0001877940766624315,
+      "loss": 1.1801,
+      "step": 1681
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.694056510925293,
+      "learning_rate": 0.00018777968216857893,
+      "loss": 1.1596,
+      "step": 1682
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.606585681438446,
+      "learning_rate": 0.00018776527974432908,
+      "loss": 1.2664,
+      "step": 1683
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5530321598052979,
+      "learning_rate": 0.00018775086939098312,
+      "loss": 1.1819,
+      "step": 1684
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6291928291320801,
+      "learning_rate": 0.00018773645110984298,
+      "loss": 1.3132,
+      "step": 1685
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5767596364021301,
+      "learning_rate": 0.00018772202490221124,
+      "loss": 1.0349,
+      "step": 1686
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6143909096717834,
+      "learning_rate": 0.0001877075907693912,
+      "loss": 1.2825,
+      "step": 1687
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5503374338150024,
+      "learning_rate": 0.000187693148712687,
+      "loss": 1.1777,
+      "step": 1688
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5860001444816589,
+      "learning_rate": 0.0001876786987334033,
+      "loss": 0.9836,
+      "step": 1689
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.645011305809021,
+      "learning_rate": 0.00018766424083284557,
+      "loss": 1.1498,
+      "step": 1690
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5632151365280151,
+      "learning_rate": 0.00018764977501232007,
+      "loss": 1.0685,
+      "step": 1691
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5840081572532654,
+      "learning_rate": 0.00018763530127313365,
+      "loss": 1.0215,
+      "step": 1692
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5783519148826599,
+      "learning_rate": 0.00018762081961659396,
+      "loss": 1.0851,
+      "step": 1693
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.554547131061554,
+      "learning_rate": 0.0001876063300440093,
+      "loss": 1.1261,
+      "step": 1694
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5496121644973755,
+      "learning_rate": 0.0001875918325566888,
+      "loss": 0.8681,
+      "step": 1695
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5871601104736328,
+      "learning_rate": 0.00018757732715594214,
+      "loss": 0.9968,
+      "step": 1696
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5784563422203064,
+      "learning_rate": 0.00018756281384307982,
+      "loss": 1.1381,
+      "step": 1697
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5940728187561035,
+      "learning_rate": 0.00018754829261941308,
+      "loss": 1.0509,
+      "step": 1698
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5413823127746582,
+      "learning_rate": 0.00018753376348625377,
+      "loss": 1.091,
+      "step": 1699
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5205764770507812,
+      "learning_rate": 0.00018751922644491456,
+      "loss": 1.077,
+      "step": 1700
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5888503789901733,
+      "learning_rate": 0.0001875046814967088,
+      "loss": 1.3145,
+      "step": 1701
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6186810731887817,
+      "learning_rate": 0.00018749012864295047,
+      "loss": 1.103,
+      "step": 1702
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6271312832832336,
+      "learning_rate": 0.00018747556788495438,
+      "loss": 1.0219,
+      "step": 1703
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.543907642364502,
+      "learning_rate": 0.00018746099922403606,
+      "loss": 0.9024,
+      "step": 1704
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5604395866394043,
+      "learning_rate": 0.00018744642266151163,
+      "loss": 1.0485,
+      "step": 1705
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5746266841888428,
+      "learning_rate": 0.00018743183819869806,
+      "loss": 1.0975,
+      "step": 1706
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5429287552833557,
+      "learning_rate": 0.00018741724583691294,
+      "loss": 1.0327,
+      "step": 1707
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5611682534217834,
+      "learning_rate": 0.00018740264557747457,
+      "loss": 0.9671,
+      "step": 1708
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5795273184776306,
+      "learning_rate": 0.00018738803742170213,
+      "loss": 1.0083,
+      "step": 1709
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.602764368057251,
+      "learning_rate": 0.00018737342137091522,
+      "loss": 1.2469,
+      "step": 1710
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5357661843299866,
+      "learning_rate": 0.00018735879742643442,
+      "loss": 1.1328,
+      "step": 1711
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5171231031417847,
+      "learning_rate": 0.0001873441655895809,
+      "loss": 1.1278,
+      "step": 1712
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6350235342979431,
+      "learning_rate": 0.00018732952586167655,
+      "loss": 1.0691,
+      "step": 1713
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.579694926738739,
+      "learning_rate": 0.00018731487824404403,
+      "loss": 1.1816,
+      "step": 1714
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6309595108032227,
+      "learning_rate": 0.0001873002227380066,
+      "loss": 1.0332,
+      "step": 1715
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5699979662895203,
+      "learning_rate": 0.00018728555934488837,
+      "loss": 1.0802,
+      "step": 1716
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6824408769607544,
+      "learning_rate": 0.00018727088806601405,
+      "loss": 1.0947,
+      "step": 1717
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5273520350456238,
+      "learning_rate": 0.0001872562089027091,
+      "loss": 1.0015,
+      "step": 1718
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5987673997879028,
+      "learning_rate": 0.00018724152185629976,
+      "loss": 1.0188,
+      "step": 1719
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6542358994483948,
+      "learning_rate": 0.00018722682692811282,
+      "loss": 1.2388,
+      "step": 1720
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5636007785797119,
+      "learning_rate": 0.000187212124119476,
+      "loss": 1.1047,
+      "step": 1721
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5961930751800537,
+      "learning_rate": 0.00018719741343171755,
+      "loss": 1.2205,
+      "step": 1722
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5861148834228516,
+      "learning_rate": 0.0001871826948661665,
+      "loss": 1.1424,
+      "step": 1723
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6240936517715454,
+      "learning_rate": 0.00018716796842415258,
+      "loss": 1.2925,
+      "step": 1724
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5183912515640259,
+      "learning_rate": 0.00018715323410700626,
+      "loss": 0.922,
+      "step": 1725
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.611001193523407,
+      "learning_rate": 0.00018713849191605872,
+      "loss": 1.2328,
+      "step": 1726
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6203360557556152,
+      "learning_rate": 0.0001871237418526418,
+      "loss": 1.2026,
+      "step": 1727
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6380226016044617,
+      "learning_rate": 0.00018710898391808808,
+      "loss": 1.193,
+      "step": 1728
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.711965024471283,
+      "learning_rate": 0.00018709421811373091,
+      "loss": 1.1432,
+      "step": 1729
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5226439833641052,
+      "learning_rate": 0.0001870794444409042,
+      "loss": 1.0078,
+      "step": 1730
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.49361833930015564,
+      "learning_rate": 0.0001870646629009428,
+      "loss": 1.0736,
+      "step": 1731
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5229170322418213,
+      "learning_rate": 0.000187049873495182,
+      "loss": 1.1184,
+      "step": 1732
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.518227219581604,
+      "learning_rate": 0.00018703507622495807,
+      "loss": 1.1062,
+      "step": 1733
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5495775938034058,
+      "learning_rate": 0.00018702027109160775,
+      "loss": 0.9128,
+      "step": 1734
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6473329067230225,
+      "learning_rate": 0.00018700545809646863,
+      "loss": 0.9592,
+      "step": 1735
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5847293138504028,
+      "learning_rate": 0.00018699063724087904,
+      "loss": 0.9249,
+      "step": 1736
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6614704728126526,
+      "learning_rate": 0.0001869758085261779,
+      "loss": 1.0977,
+      "step": 1737
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6490361094474792,
+      "learning_rate": 0.0001869609719537049,
+      "loss": 1.1991,
+      "step": 1738
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6661361455917358,
+      "learning_rate": 0.00018694612752480046,
+      "loss": 1.0612,
+      "step": 1739
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5234553813934326,
+      "learning_rate": 0.00018693127524080572,
+      "loss": 0.9233,
+      "step": 1740
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.49033844470977783,
+      "learning_rate": 0.0001869164151030624,
+      "loss": 0.8696,
+      "step": 1741
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5724113583564758,
+      "learning_rate": 0.00018690154711291314,
+      "loss": 1.0518,
+      "step": 1742
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5647688508033752,
+      "learning_rate": 0.00018688667127170113,
+      "loss": 1.1055,
+      "step": 1743
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5435717701911926,
+      "learning_rate": 0.00018687178758077032,
+      "loss": 0.9948,
+      "step": 1744
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6030397415161133,
+      "learning_rate": 0.00018685689604146537,
+      "loss": 0.9858,
+      "step": 1745
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.550806999206543,
+      "learning_rate": 0.00018684199665513164,
+      "loss": 1.0742,
+      "step": 1746
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6276724338531494,
+      "learning_rate": 0.0001868270894231152,
+      "loss": 1.2271,
+      "step": 1747
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7018080949783325,
+      "learning_rate": 0.00018681217434676287,
+      "loss": 1.1183,
+      "step": 1748
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5612937211990356,
+      "learning_rate": 0.00018679725142742212,
+      "loss": 1.1673,
+      "step": 1749
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6091915965080261,
+      "learning_rate": 0.0001867823206664411,
+      "loss": 1.1528,
+      "step": 1750
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5641574263572693,
+      "learning_rate": 0.0001867673820651688,
+      "loss": 1.1589,
+      "step": 1751
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.59687340259552,
+      "learning_rate": 0.0001867524356249548,
+      "loss": 1.1207,
+      "step": 1752
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5552710294723511,
+      "learning_rate": 0.00018673748134714945,
+      "loss": 1.1232,
+      "step": 1753
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.657662570476532,
+      "learning_rate": 0.00018672251923310376,
+      "loss": 1.3406,
+      "step": 1754
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6064003705978394,
+      "learning_rate": 0.00018670754928416946,
+      "loss": 1.1582,
+      "step": 1755
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5749542117118835,
+      "learning_rate": 0.00018669257150169904,
+      "loss": 0.9952,
+      "step": 1756
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5489286780357361,
+      "learning_rate": 0.0001866775858870456,
+      "loss": 1.2523,
+      "step": 1757
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5087161660194397,
+      "learning_rate": 0.0001866625924415631,
+      "loss": 0.91,
+      "step": 1758
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5486624240875244,
+      "learning_rate": 0.00018664759116660603,
+      "loss": 0.8993,
+      "step": 1759
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5084232687950134,
+      "learning_rate": 0.0001866325820635297,
+      "loss": 1.0808,
+      "step": 1760
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6880375742912292,
+      "learning_rate": 0.00018661756513369007,
+      "loss": 1.1413,
+      "step": 1761
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6002033948898315,
+      "learning_rate": 0.00018660254037844388,
+      "loss": 1.0729,
+      "step": 1762
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5921014547348022,
+      "learning_rate": 0.00018658750779914848,
+      "loss": 1.1544,
+      "step": 1763
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6466764807701111,
+      "learning_rate": 0.00018657246739716205,
+      "loss": 1.1871,
+      "step": 1764
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6112008690834045,
+      "learning_rate": 0.00018655741917384332,
+      "loss": 0.9757,
+      "step": 1765
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5578400492668152,
+      "learning_rate": 0.00018654236313055188,
+      "loss": 0.934,
+      "step": 1766
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5956066250801086,
+      "learning_rate": 0.00018652729926864795,
+      "loss": 1.1892,
+      "step": 1767
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.598323404788971,
+      "learning_rate": 0.00018651222758949242,
+      "loss": 0.9415,
+      "step": 1768
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6082397103309631,
+      "learning_rate": 0.00018649714809444696,
+      "loss": 1.148,
+      "step": 1769
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.566173791885376,
+      "learning_rate": 0.0001864820607848739,
+      "loss": 1.1214,
+      "step": 1770
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5706698894500732,
+      "learning_rate": 0.00018646696566213634,
+      "loss": 1.2498,
+      "step": 1771
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5568757653236389,
+      "learning_rate": 0.00018645186272759801,
+      "loss": 0.9619,
+      "step": 1772
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5579240322113037,
+      "learning_rate": 0.00018643675198262333,
+      "loss": 0.9143,
+      "step": 1773
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5389413833618164,
+      "learning_rate": 0.00018642163342857756,
+      "loss": 1.0979,
+      "step": 1774
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6314026713371277,
+      "learning_rate": 0.0001864065070668265,
+      "loss": 1.2771,
+      "step": 1775
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6543720960617065,
+      "learning_rate": 0.00018639137289873676,
+      "loss": 1.2627,
+      "step": 1776
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.49429407715797424,
+      "learning_rate": 0.00018637623092567564,
+      "loss": 1.0445,
+      "step": 1777
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5479093194007874,
+      "learning_rate": 0.0001863610811490111,
+      "loss": 1.0667,
+      "step": 1778
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5306311845779419,
+      "learning_rate": 0.00018634592357011186,
+      "loss": 0.949,
+      "step": 1779
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5764743089675903,
+      "learning_rate": 0.00018633075819034732,
+      "loss": 1.1231,
+      "step": 1780
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5098384618759155,
+      "learning_rate": 0.00018631558501108758,
+      "loss": 1.0341,
+      "step": 1781
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6208810806274414,
+      "learning_rate": 0.00018630040403370344,
+      "loss": 1.1288,
+      "step": 1782
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6083211898803711,
+      "learning_rate": 0.00018628521525956645,
+      "loss": 1.1865,
+      "step": 1783
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5995293259620667,
+      "learning_rate": 0.0001862700186900488,
+      "loss": 1.1144,
+      "step": 1784
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5885659456253052,
+      "learning_rate": 0.00018625481432652338,
+      "loss": 1.0731,
+      "step": 1785
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6302295327186584,
+      "learning_rate": 0.00018623960217036388,
+      "loss": 1.1223,
+      "step": 1786
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5090101957321167,
+      "learning_rate": 0.00018622438222294464,
+      "loss": 0.9719,
+      "step": 1787
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.592470645904541,
+      "learning_rate": 0.00018620915448564068,
+      "loss": 1.1259,
+      "step": 1788
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.602279782295227,
+      "learning_rate": 0.00018619391895982766,
+      "loss": 1.0776,
+      "step": 1789
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6333302855491638,
+      "learning_rate": 0.00018617867564688214,
+      "loss": 1.1915,
+      "step": 1790
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6094706058502197,
+      "learning_rate": 0.0001861634245481812,
+      "loss": 1.101,
+      "step": 1791
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5379626154899597,
+      "learning_rate": 0.00018614816566510272,
+      "loss": 1.0421,
+      "step": 1792
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5724616050720215,
+      "learning_rate": 0.00018613289899902525,
+      "loss": 1.1036,
+      "step": 1793
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.524731457233429,
+      "learning_rate": 0.00018611762455132802,
+      "loss": 0.9453,
+      "step": 1794
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.631876528263092,
+      "learning_rate": 0.00018610234232339102,
+      "loss": 1.2017,
+      "step": 1795
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5800185799598694,
+      "learning_rate": 0.00018608705231659486,
+      "loss": 1.0768,
+      "step": 1796
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7038686275482178,
+      "learning_rate": 0.00018607175453232098,
+      "loss": 1.2161,
+      "step": 1797
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.605658233165741,
+      "learning_rate": 0.00018605644897195145,
+      "loss": 1.1248,
+      "step": 1798
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5098914504051208,
+      "learning_rate": 0.00018604113563686897,
+      "loss": 0.9575,
+      "step": 1799
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.554283082485199,
+      "learning_rate": 0.00018602581452845703,
+      "loss": 1.0522,
+      "step": 1800
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5628049373626709,
+      "learning_rate": 0.00018601048564809985,
+      "loss": 1.1679,
+      "step": 1801
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5380632281303406,
+      "learning_rate": 0.0001859951489971823,
+      "loss": 1.1503,
+      "step": 1802
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.638298749923706,
+      "learning_rate": 0.0001859798045770899,
+      "loss": 1.2309,
+      "step": 1803
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5701687932014465,
+      "learning_rate": 0.00018596445238920902,
+      "loss": 1.2437,
+      "step": 1804
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5610591173171997,
+      "learning_rate": 0.00018594909243492653,
+      "loss": 1.1682,
+      "step": 1805
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5382736921310425,
+      "learning_rate": 0.00018593372471563022,
+      "loss": 0.9795,
+      "step": 1806
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5971048474311829,
+      "learning_rate": 0.00018591834923270843,
+      "loss": 0.9871,
+      "step": 1807
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5558769106864929,
+      "learning_rate": 0.00018590296598755028,
+      "loss": 0.8444,
+      "step": 1808
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5690754055976868,
+      "learning_rate": 0.0001858875749815455,
+      "loss": 1.0904,
+      "step": 1809
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6201112270355225,
+      "learning_rate": 0.00018587217621608462,
+      "loss": 1.1436,
+      "step": 1810
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6081562638282776,
+      "learning_rate": 0.00018585676969255885,
+      "loss": 1.2893,
+      "step": 1811
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6248214244842529,
+      "learning_rate": 0.00018584135541236006,
+      "loss": 1.1327,
+      "step": 1812
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5153765082359314,
+      "learning_rate": 0.0001858259333768808,
+      "loss": 1.0487,
+      "step": 1813
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5564947724342346,
+      "learning_rate": 0.00018581050358751445,
+      "loss": 1.0409,
+      "step": 1814
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5596327781677246,
+      "learning_rate": 0.00018579506604565494,
+      "loss": 1.118,
+      "step": 1815
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6628304719924927,
+      "learning_rate": 0.00018577962075269698,
+      "loss": 1.182,
+      "step": 1816
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5848866701126099,
+      "learning_rate": 0.00018576416771003595,
+      "loss": 1.2644,
+      "step": 1817
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.600448489189148,
+      "learning_rate": 0.00018574870691906797,
+      "loss": 1.1461,
+      "step": 1818
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5690786242485046,
+      "learning_rate": 0.00018573323838118986,
+      "loss": 1.1681,
+      "step": 1819
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5750014781951904,
+      "learning_rate": 0.00018571776209779905,
+      "loss": 0.9931,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5702171921730042,
+      "learning_rate": 0.00018570227807029378,
+      "loss": 1.0003,
+      "step": 1821
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5780484676361084,
+      "learning_rate": 0.00018568678630007294,
+      "loss": 1.1079,
+      "step": 1822
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5208010077476501,
+      "learning_rate": 0.00018567128678853608,
+      "loss": 1.1168,
+      "step": 1823
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5132521986961365,
+      "learning_rate": 0.00018565577953708358,
+      "loss": 0.9701,
+      "step": 1824
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5545271635055542,
+      "learning_rate": 0.00018564026454711633,
+      "loss": 0.9203,
+      "step": 1825
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5755243897438049,
+      "learning_rate": 0.0001856247418200361,
+      "loss": 1.049,
+      "step": 1826
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.601019024848938,
+      "learning_rate": 0.00018560921135724528,
+      "loss": 1.1816,
+      "step": 1827
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6014050841331482,
+      "learning_rate": 0.00018559367316014688,
+      "loss": 1.1182,
+      "step": 1828
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5794886946678162,
+      "learning_rate": 0.00018557812723014476,
+      "loss": 0.9821,
+      "step": 1829
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6025217175483704,
+      "learning_rate": 0.00018556257356864338,
+      "loss": 1.0827,
+      "step": 1830
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.581674337387085,
+      "learning_rate": 0.00018554701217704796,
+      "loss": 0.8873,
+      "step": 1831
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.669243574142456,
+      "learning_rate": 0.00018553144305676433,
+      "loss": 0.9918,
+      "step": 1832
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6118124723434448,
+      "learning_rate": 0.0001855158662091991,
+      "loss": 1.0832,
+      "step": 1833
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5944288969039917,
+      "learning_rate": 0.00018550028163575954,
+      "loss": 1.2669,
+      "step": 1834
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6186957359313965,
+      "learning_rate": 0.00018548468933785365,
+      "loss": 0.9488,
+      "step": 1835
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5669781565666199,
+      "learning_rate": 0.00018546908931689007,
+      "loss": 1.1178,
+      "step": 1836
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6318771839141846,
+      "learning_rate": 0.00018545348157427823,
+      "loss": 1.061,
+      "step": 1837
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6139348745346069,
+      "learning_rate": 0.00018543786611142815,
+      "loss": 1.0234,
+      "step": 1838
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4509977102279663,
+      "learning_rate": 0.00018542224292975062,
+      "loss": 0.7523,
+      "step": 1839
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5004023909568787,
+      "learning_rate": 0.0001854066120306571,
+      "loss": 1.0223,
+      "step": 1840
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6322842836380005,
+      "learning_rate": 0.00018539097341555973,
+      "loss": 1.1863,
+      "step": 1841
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5344803333282471,
+      "learning_rate": 0.00018537532708587143,
+      "loss": 1.0257,
+      "step": 1842
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5191022753715515,
+      "learning_rate": 0.0001853596730430057,
+      "loss": 0.9801,
+      "step": 1843
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5770233869552612,
+      "learning_rate": 0.00018534401128837683,
+      "loss": 1.0975,
+      "step": 1844
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6385951638221741,
+      "learning_rate": 0.0001853283418233997,
+      "loss": 0.9929,
+      "step": 1845
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5498313903808594,
+      "learning_rate": 0.0001853126646494901,
+      "loss": 0.9341,
+      "step": 1846
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6085447669029236,
+      "learning_rate": 0.00018529697976806423,
+      "loss": 1.057,
+      "step": 1847
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5516953468322754,
+      "learning_rate": 0.0001852812871805392,
+      "loss": 1.0169,
+      "step": 1848
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5250498652458191,
+      "learning_rate": 0.00018526558688833274,
+      "loss": 0.9131,
+      "step": 1849
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7380767464637756,
+      "learning_rate": 0.0001852498788928633,
+      "loss": 1.2816,
+      "step": 1850
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5584884285926819,
+      "learning_rate": 0.00018523416319554995,
+      "loss": 0.9567,
+      "step": 1851
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5699655413627625,
+      "learning_rate": 0.00018521843979781258,
+      "loss": 1.1744,
+      "step": 1852
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5850598216056824,
+      "learning_rate": 0.00018520270870107166,
+      "loss": 1.0347,
+      "step": 1853
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5637402534484863,
+      "learning_rate": 0.00018518696990674842,
+      "loss": 1.2212,
+      "step": 1854
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6142070889472961,
+      "learning_rate": 0.00018517122341626477,
+      "loss": 1.07,
+      "step": 1855
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5141515135765076,
+      "learning_rate": 0.00018515546923104337,
+      "loss": 0.9331,
+      "step": 1856
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.473111093044281,
+      "learning_rate": 0.00018513970735250742,
+      "loss": 0.7911,
+      "step": 1857
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6113271713256836,
+      "learning_rate": 0.000185123937782081,
+      "loss": 1.2151,
+      "step": 1858
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6086576581001282,
+      "learning_rate": 0.00018510816052118875,
+      "loss": 1.1222,
+      "step": 1859
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5792139768600464,
+      "learning_rate": 0.0001850923755712561,
+      "loss": 0.9931,
+      "step": 1860
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5663713216781616,
+      "learning_rate": 0.00018507658293370912,
+      "loss": 1.1144,
+      "step": 1861
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6032007336616516,
+      "learning_rate": 0.0001850607826099745,
+      "loss": 1.1773,
+      "step": 1862
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5945179462432861,
+      "learning_rate": 0.00018504497460147988,
+      "loss": 0.9072,
+      "step": 1863
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6139575839042664,
+      "learning_rate": 0.00018502915890965328,
+      "loss": 1.1492,
+      "step": 1864
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5732980370521545,
+      "learning_rate": 0.0001850133355359236,
+      "loss": 0.9399,
+      "step": 1865
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6116182804107666,
+      "learning_rate": 0.00018499750448172045,
+      "loss": 1.1455,
+      "step": 1866
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5685417652130127,
+      "learning_rate": 0.00018498166574847398,
+      "loss": 1.1286,
+      "step": 1867
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6494985222816467,
+      "learning_rate": 0.00018496581933761516,
+      "loss": 1.1148,
+      "step": 1868
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5793482065200806,
+      "learning_rate": 0.00018494996525057566,
+      "loss": 1.1552,
+      "step": 1869
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5607535243034363,
+      "learning_rate": 0.00018493410348878779,
+      "loss": 1.0392,
+      "step": 1870
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5485084652900696,
+      "learning_rate": 0.00018491823405368456,
+      "loss": 0.8355,
+      "step": 1871
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5895860195159912,
+      "learning_rate": 0.0001849023569466997,
+      "loss": 1.3229,
+      "step": 1872
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6231148838996887,
+      "learning_rate": 0.00018488647216926757,
+      "loss": 1.0216,
+      "step": 1873
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5446336269378662,
+      "learning_rate": 0.00018487057972282336,
+      "loss": 1.0869,
+      "step": 1874
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6365904211997986,
+      "learning_rate": 0.00018485467960880275,
+      "loss": 1.0267,
+      "step": 1875
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5633032917976379,
+      "learning_rate": 0.0001848387718286423,
+      "loss": 1.0315,
+      "step": 1876
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5365302562713623,
+      "learning_rate": 0.0001848228563837792,
+      "loss": 1.2495,
+      "step": 1877
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6207290291786194,
+      "learning_rate": 0.0001848069332756513,
+      "loss": 1.1013,
+      "step": 1878
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6174927949905396,
+      "learning_rate": 0.0001847910025056971,
+      "loss": 1.2725,
+      "step": 1879
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.565679132938385,
+      "learning_rate": 0.00018477506407535595,
+      "loss": 0.9717,
+      "step": 1880
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6149706244468689,
+      "learning_rate": 0.00018475911798606772,
+      "loss": 0.9632,
+      "step": 1881
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5572057962417603,
+      "learning_rate": 0.00018474316423927313,
+      "loss": 0.9965,
+      "step": 1882
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5754985213279724,
+      "learning_rate": 0.00018472720283641345,
+      "loss": 0.8882,
+      "step": 1883
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6410358548164368,
+      "learning_rate": 0.0001847112337789307,
+      "loss": 0.9525,
+      "step": 1884
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6138419508934021,
+      "learning_rate": 0.00018469525706826763,
+      "loss": 1.1237,
+      "step": 1885
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5606120824813843,
+      "learning_rate": 0.0001846792727058676,
+      "loss": 0.9746,
+      "step": 1886
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5778054594993591,
+      "learning_rate": 0.0001846632806931748,
+      "loss": 1.0737,
+      "step": 1887
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5834203958511353,
+      "learning_rate": 0.00018464728103163386,
+      "loss": 1.2133,
+      "step": 1888
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6452996134757996,
+      "learning_rate": 0.0001846312737226904,
+      "loss": 1.1829,
+      "step": 1889
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5307098627090454,
+      "learning_rate": 0.00018461525876779057,
+      "loss": 0.9321,
+      "step": 1890
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5604325532913208,
+      "learning_rate": 0.00018459923616838113,
+      "loss": 1.0976,
+      "step": 1891
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5541103482246399,
+      "learning_rate": 0.00018458320592590975,
+      "loss": 1.0378,
+      "step": 1892
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5926581621170044,
+      "learning_rate": 0.0001845671680418246,
+      "loss": 0.9836,
+      "step": 1893
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6027371883392334,
+      "learning_rate": 0.00018455112251757465,
+      "loss": 1.2251,
+      "step": 1894
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6586189270019531,
+      "learning_rate": 0.0001845350693546095,
+      "loss": 1.2198,
+      "step": 1895
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5091075897216797,
+      "learning_rate": 0.0001845190085543795,
+      "loss": 1.041,
+      "step": 1896
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5649153590202332,
+      "learning_rate": 0.00018450294011833557,
+      "loss": 1.1006,
+      "step": 1897
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6161060929298401,
+      "learning_rate": 0.0001844868640479295,
+      "loss": 1.0289,
+      "step": 1898
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5821948647499084,
+      "learning_rate": 0.0001844707803446136,
+      "loss": 0.9917,
+      "step": 1899
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6699268221855164,
+      "learning_rate": 0.00018445468900984098,
+      "loss": 1.2236,
+      "step": 1900
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5464971661567688,
+      "learning_rate": 0.0001844385900450654,
+      "loss": 1.0474,
+      "step": 1901
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6147408485412598,
+      "learning_rate": 0.00018442248345174126,
+      "loss": 1.1151,
+      "step": 1902
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6387994289398193,
+      "learning_rate": 0.00018440636923132376,
+      "loss": 1.2701,
+      "step": 1903
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6076620221138,
+      "learning_rate": 0.00018439024738526873,
+      "loss": 0.957,
+      "step": 1904
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6402626037597656,
+      "learning_rate": 0.00018437411791503262,
+      "loss": 1.032,
+      "step": 1905
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6451716423034668,
+      "learning_rate": 0.00018435798082207274,
+      "loss": 1.0105,
+      "step": 1906
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7566270232200623,
+      "learning_rate": 0.0001843418361078469,
+      "loss": 0.8896,
+      "step": 1907
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.589716911315918,
+      "learning_rate": 0.0001843256837738137,
+      "loss": 1.0328,
+      "step": 1908
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7324854731559753,
+      "learning_rate": 0.00018430952382143242,
+      "loss": 1.0414,
+      "step": 1909
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6748380064964294,
+      "learning_rate": 0.00018429335625216303,
+      "loss": 1.1903,
+      "step": 1910
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5136976838111877,
+      "learning_rate": 0.00018427718106746617,
+      "loss": 0.9545,
+      "step": 1911
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6081967949867249,
+      "learning_rate": 0.00018426099826880318,
+      "loss": 1.0921,
+      "step": 1912
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5389492511749268,
+      "learning_rate": 0.00018424480785763606,
+      "loss": 1.0173,
+      "step": 1913
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6011937260627747,
+      "learning_rate": 0.00018422860983542758,
+      "loss": 0.9444,
+      "step": 1914
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6137906908988953,
+      "learning_rate": 0.00018421240420364108,
+      "loss": 1.0585,
+      "step": 1915
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6015632748603821,
+      "learning_rate": 0.00018419619096374067,
+      "loss": 1.1256,
+      "step": 1916
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6502439975738525,
+      "learning_rate": 0.00018417997011719111,
+      "loss": 1.158,
+      "step": 1917
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5486273765563965,
+      "learning_rate": 0.00018416374166545788,
+      "loss": 0.9079,
+      "step": 1918
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5800416469573975,
+      "learning_rate": 0.0001841475056100072,
+      "loss": 1.1742,
+      "step": 1919
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5277450084686279,
+      "learning_rate": 0.00018413126195230572,
+      "loss": 1.0423,
+      "step": 1920
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5868902802467346,
+      "learning_rate": 0.00018411501069382112,
+      "loss": 1.1206,
+      "step": 1921
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5652511119842529,
+      "learning_rate": 0.00018409875183602158,
+      "loss": 1.0377,
+      "step": 1922
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5721745491027832,
+      "learning_rate": 0.00018408248538037594,
+      "loss": 0.9665,
+      "step": 1923
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6899604201316833,
+      "learning_rate": 0.00018406621132835383,
+      "loss": 1.2337,
+      "step": 1924
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.604383647441864,
+      "learning_rate": 0.00018404992968142552,
+      "loss": 1.0569,
+      "step": 1925
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5258768200874329,
+      "learning_rate": 0.00018403364044106195,
+      "loss": 1.067,
+      "step": 1926
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5347247123718262,
+      "learning_rate": 0.00018401734360873478,
+      "loss": 1.0679,
+      "step": 1927
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5636218786239624,
+      "learning_rate": 0.0001840010391859163,
+      "loss": 1.0746,
+      "step": 1928
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6444405913352966,
+      "learning_rate": 0.00018398472717407954,
+      "loss": 1.1184,
+      "step": 1929
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6043636202812195,
+      "learning_rate": 0.00018396840757469818,
+      "loss": 1.0266,
+      "step": 1930
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5988621115684509,
+      "learning_rate": 0.00018395208038924667,
+      "loss": 0.9645,
+      "step": 1931
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5031013488769531,
+      "learning_rate": 0.0001839357456192,
+      "loss": 0.9798,
+      "step": 1932
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6296197175979614,
+      "learning_rate": 0.00018391940326603395,
+      "loss": 0.9574,
+      "step": 1933
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7154589891433716,
+      "learning_rate": 0.000183903053331225,
+      "loss": 1.233,
+      "step": 1934
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5253052711486816,
+      "learning_rate": 0.00018388669581625023,
+      "loss": 1.051,
+      "step": 1935
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6451107263565063,
+      "learning_rate": 0.00018387033072258743,
+      "loss": 1.1662,
+      "step": 1936
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6134210228919983,
+      "learning_rate": 0.00018385395805171515,
+      "loss": 1.2124,
+      "step": 1937
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.521445631980896,
+      "learning_rate": 0.00018383757780511252,
+      "loss": 1.066,
+      "step": 1938
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6145197749137878,
+      "learning_rate": 0.00018382118998425944,
+      "loss": 1.1536,
+      "step": 1939
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5896132588386536,
+      "learning_rate": 0.00018380479459063643,
+      "loss": 1.1371,
+      "step": 1940
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5753506422042847,
+      "learning_rate": 0.0001837883916257247,
+      "loss": 1.2127,
+      "step": 1941
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6061387062072754,
+      "learning_rate": 0.00018377198109100626,
+      "loss": 0.9953,
+      "step": 1942
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6697282195091248,
+      "learning_rate": 0.0001837555629879636,
+      "loss": 1.2423,
+      "step": 1943
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5806127786636353,
+      "learning_rate": 0.00018373913731808007,
+      "loss": 1.1304,
+      "step": 1944
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.538391649723053,
+      "learning_rate": 0.0001837227040828396,
+      "loss": 1.0623,
+      "step": 1945
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6214932799339294,
+      "learning_rate": 0.00018370626328372683,
+      "loss": 1.147,
+      "step": 1946
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5707278251647949,
+      "learning_rate": 0.00018368981492222717,
+      "loss": 1.0347,
+      "step": 1947
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5317419767379761,
+      "learning_rate": 0.00018367335899982653,
+      "loss": 1.1575,
+      "step": 1948
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6756814122200012,
+      "learning_rate": 0.00018365689551801172,
+      "loss": 1.1641,
+      "step": 1949
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5761415958404541,
+      "learning_rate": 0.00018364042447827,
+      "loss": 1.2391,
+      "step": 1950
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6367659568786621,
+      "learning_rate": 0.00018362394588208953,
+      "loss": 1.0521,
+      "step": 1951
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.612143874168396,
+      "learning_rate": 0.00018360745973095902,
+      "loss": 1.1476,
+      "step": 1952
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6905182600021362,
+      "learning_rate": 0.00018359096602636794,
+      "loss": 1.1497,
+      "step": 1953
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6150467395782471,
+      "learning_rate": 0.00018357446476980634,
+      "loss": 1.0446,
+      "step": 1954
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6133366227149963,
+      "learning_rate": 0.00018355795596276502,
+      "loss": 1.0702,
+      "step": 1955
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8109869360923767,
+      "learning_rate": 0.00018354143960673552,
+      "loss": 1.1479,
+      "step": 1956
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5375014543533325,
+      "learning_rate": 0.00018352491570320997,
+      "loss": 1.0464,
+      "step": 1957
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5784035325050354,
+      "learning_rate": 0.00018350838425368115,
+      "loss": 1.0479,
+      "step": 1958
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6554840207099915,
+      "learning_rate": 0.0001834918452596427,
+      "loss": 1.0229,
+      "step": 1959
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5233038067817688,
+      "learning_rate": 0.00018347529872258874,
+      "loss": 1.1293,
+      "step": 1960
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5993428826332092,
+      "learning_rate": 0.00018345874464401412,
+      "loss": 1.0875,
+      "step": 1961
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5326312184333801,
+      "learning_rate": 0.00018344218302541454,
+      "loss": 1.0039,
+      "step": 1962
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6286762356758118,
+      "learning_rate": 0.00018342561386828615,
+      "loss": 1.1151,
+      "step": 1963
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6470648646354675,
+      "learning_rate": 0.00018340903717412588,
+      "loss": 1.2054,
+      "step": 1964
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5421403050422668,
+      "learning_rate": 0.0001833924529444314,
+      "loss": 1.1086,
+      "step": 1965
+    },
+    {
+      "epoch": 0.75,
+      "eval_loss": 0.9726712107658386,
+      "eval_runtime": 17.8102,
+      "eval_samples_per_second": 153.227,
+      "eval_steps_per_second": 38.349,
+      "step": 1965
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.59051114320755,
+      "learning_rate": 0.00018337586118070091,
+      "loss": 1.1655,
+      "step": 1966
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5358890891075134,
+      "learning_rate": 0.00018335926188443348,
+      "loss": 0.8833,
+      "step": 1967
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.590603232383728,
+      "learning_rate": 0.0001833426550571287,
+      "loss": 1.2004,
+      "step": 1968
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5690146684646606,
+      "learning_rate": 0.00018332604070028693,
+      "loss": 1.0583,
+      "step": 1969
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5618229508399963,
+      "learning_rate": 0.00018330941881540915,
+      "loss": 1.0755,
+      "step": 1970
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6254324316978455,
+      "learning_rate": 0.00018329278940399713,
+      "loss": 1.214,
+      "step": 1971
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6175144910812378,
+      "learning_rate": 0.0001832761524675531,
+      "loss": 1.192,
+      "step": 1972
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5359880924224854,
+      "learning_rate": 0.0001832595080075803,
+      "loss": 0.9377,
+      "step": 1973
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5933471322059631,
+      "learning_rate": 0.0001832428560255823,
+      "loss": 1.1249,
+      "step": 1974
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7249235510826111,
+      "learning_rate": 0.00018322619652306362,
+      "loss": 1.0241,
+      "step": 1975
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6383464932441711,
+      "learning_rate": 0.00018320952950152928,
+      "loss": 1.2613,
+      "step": 1976
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5812982320785522,
+      "learning_rate": 0.00018319285496248508,
+      "loss": 1.0963,
+      "step": 1977
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6639915704727173,
+      "learning_rate": 0.00018317617290743748,
+      "loss": 1.0557,
+      "step": 1978
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6422559022903442,
+      "learning_rate": 0.00018315948333789362,
+      "loss": 1.0014,
+      "step": 1979
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5967164039611816,
+      "learning_rate": 0.00018314278625536126,
+      "loss": 1.0322,
+      "step": 1980
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6338431239128113,
+      "learning_rate": 0.00018312608166134892,
+      "loss": 1.1217,
+      "step": 1981
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5600549578666687,
+      "learning_rate": 0.00018310936955736579,
+      "loss": 1.2031,
+      "step": 1982
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.640558660030365,
+      "learning_rate": 0.00018309264994492163,
+      "loss": 1.1203,
+      "step": 1983
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6528109312057495,
+      "learning_rate": 0.00018307592282552706,
+      "loss": 1.1409,
+      "step": 1984
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5706397891044617,
+      "learning_rate": 0.0001830591882006932,
+      "loss": 1.0011,
+      "step": 1985
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6276172399520874,
+      "learning_rate": 0.00018304244607193198,
+      "loss": 1.2437,
+      "step": 1986
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6464489102363586,
+      "learning_rate": 0.00018302569644075597,
+      "loss": 1.049,
+      "step": 1987
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5966342091560364,
+      "learning_rate": 0.00018300893930867838,
+      "loss": 1.2014,
+      "step": 1988
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5682297945022583,
+      "learning_rate": 0.0001829921746772131,
+      "loss": 0.9777,
+      "step": 1989
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6807078123092651,
+      "learning_rate": 0.00018297540254787472,
+      "loss": 1.0618,
+      "step": 1990
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5794547200202942,
+      "learning_rate": 0.00018295862292217857,
+      "loss": 0.9338,
+      "step": 1991
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5602817535400391,
+      "learning_rate": 0.0001829418358016405,
+      "loss": 1.1029,
+      "step": 1992
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6040206551551819,
+      "learning_rate": 0.0001829250411877772,
+      "loss": 1.0249,
+      "step": 1993
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6918087601661682,
+      "learning_rate": 0.00018290823908210594,
+      "loss": 1.2849,
+      "step": 1994
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6619887351989746,
+      "learning_rate": 0.0001828914294861447,
+      "loss": 1.1101,
+      "step": 1995
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6102719902992249,
+      "learning_rate": 0.00018287461240141216,
+      "loss": 1.0639,
+      "step": 1996
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6364079117774963,
+      "learning_rate": 0.00018285778782942762,
+      "loss": 1.2199,
+      "step": 1997
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6381446123123169,
+      "learning_rate": 0.0001828409557717111,
+      "loss": 1.1562,
+      "step": 1998
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6087682843208313,
+      "learning_rate": 0.00018282411622978323,
+      "loss": 1.0874,
+      "step": 1999
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6255144476890564,
+      "learning_rate": 0.00018280726920516546,
+      "loss": 1.3216,
+      "step": 2000
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5824018716812134,
+      "learning_rate": 0.00018279041469937974,
+      "loss": 1.2186,
+      "step": 2001
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6130787134170532,
+      "learning_rate": 0.0001827735527139488,
+      "loss": 1.164,
+      "step": 2002
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6037015318870544,
+      "learning_rate": 0.00018275668325039605,
+      "loss": 1.2456,
+      "step": 2003
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5351065397262573,
+      "learning_rate": 0.00018273980631024555,
+      "loss": 0.9892,
+      "step": 2004
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6300187706947327,
+      "learning_rate": 0.00018272292189502198,
+      "loss": 1.0067,
+      "step": 2005
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5483942031860352,
+      "learning_rate": 0.00018270603000625083,
+      "loss": 1.0425,
+      "step": 2006
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.607746422290802,
+      "learning_rate": 0.00018268913064545814,
+      "loss": 0.9913,
+      "step": 2007
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6014259457588196,
+      "learning_rate": 0.00018267222381417066,
+      "loss": 1.0893,
+      "step": 2008
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6995277404785156,
+      "learning_rate": 0.0001826553095139159,
+      "loss": 1.2404,
+      "step": 2009
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6149386763572693,
+      "learning_rate": 0.00018263838774622187,
+      "loss": 1.2246,
+      "step": 2010
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5554689764976501,
+      "learning_rate": 0.0001826214585126174,
+      "loss": 1.1945,
+      "step": 2011
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6338213086128235,
+      "learning_rate": 0.00018260452181463198,
+      "loss": 1.1132,
+      "step": 2012
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5740888118743896,
+      "learning_rate": 0.0001825875776537957,
+      "loss": 1.1365,
+      "step": 2013
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6145991683006287,
+      "learning_rate": 0.0001825706260316394,
+      "loss": 1.1372,
+      "step": 2014
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5843204855918884,
+      "learning_rate": 0.00018255366694969457,
+      "loss": 1.0638,
+      "step": 2015
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7211287021636963,
+      "learning_rate": 0.00018253670040949333,
+      "loss": 1.1015,
+      "step": 2016
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6173563599586487,
+      "learning_rate": 0.00018251972641256854,
+      "loss": 1.1816,
+      "step": 2017
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6879176497459412,
+      "learning_rate": 0.00018250274496045367,
+      "loss": 1.2768,
+      "step": 2018
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5964239239692688,
+      "learning_rate": 0.00018248575605468294,
+      "loss": 1.1307,
+      "step": 2019
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5030563473701477,
+      "learning_rate": 0.00018246875969679118,
+      "loss": 0.709,
+      "step": 2020
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6097801327705383,
+      "learning_rate": 0.00018245175588831392,
+      "loss": 1.0092,
+      "step": 2021
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6590239405632019,
+      "learning_rate": 0.00018243474463078735,
+      "loss": 1.1461,
+      "step": 2022
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5847734808921814,
+      "learning_rate": 0.00018241772592574834,
+      "loss": 1.1958,
+      "step": 2023
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5965940356254578,
+      "learning_rate": 0.00018240069977473446,
+      "loss": 1.2313,
+      "step": 2024
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6437633633613586,
+      "learning_rate": 0.0001823836661792839,
+      "loss": 1.0947,
+      "step": 2025
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5720009803771973,
+      "learning_rate": 0.00018236662514093557,
+      "loss": 1.0711,
+      "step": 2026
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6008542776107788,
+      "learning_rate": 0.000182349576661229,
+      "loss": 1.1588,
+      "step": 2027
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.612463116645813,
+      "learning_rate": 0.00018233252074170442,
+      "loss": 0.9719,
+      "step": 2028
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5129092931747437,
+      "learning_rate": 0.0001823154573839028,
+      "loss": 0.932,
+      "step": 2029
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7150174379348755,
+      "learning_rate": 0.00018229838658936564,
+      "loss": 1.2633,
+      "step": 2030
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6581935286521912,
+      "learning_rate": 0.00018228130835963526,
+      "loss": 1.0565,
+      "step": 2031
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5770524144172668,
+      "learning_rate": 0.0001822642226962545,
+      "loss": 1.0407,
+      "step": 2032
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5931928157806396,
+      "learning_rate": 0.000182247129600767,
+      "loss": 1.1073,
+      "step": 2033
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5641647577285767,
+      "learning_rate": 0.00018223002907471708,
+      "loss": 1.0942,
+      "step": 2034
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5922268629074097,
+      "learning_rate": 0.00018221292111964959,
+      "loss": 1.1774,
+      "step": 2035
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.637740969657898,
+      "learning_rate": 0.00018219580573711016,
+      "loss": 1.0944,
+      "step": 2036
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5891668200492859,
+      "learning_rate": 0.00018217868292864505,
+      "loss": 1.2074,
+      "step": 2037
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5810216665267944,
+      "learning_rate": 0.00018216155269580123,
+      "loss": 1.1021,
+      "step": 2038
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5588609576225281,
+      "learning_rate": 0.00018214441504012637,
+      "loss": 1.0642,
+      "step": 2039
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5198193192481995,
+      "learning_rate": 0.00018212726996316868,
+      "loss": 0.8633,
+      "step": 2040
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6176603436470032,
+      "learning_rate": 0.00018211011746647717,
+      "loss": 1.0139,
+      "step": 2041
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5617534518241882,
+      "learning_rate": 0.00018209295755160142,
+      "loss": 1.0715,
+      "step": 2042
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6227233409881592,
+      "learning_rate": 0.0001820757902200918,
+      "loss": 1.0053,
+      "step": 2043
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5912638306617737,
+      "learning_rate": 0.00018205861547349923,
+      "loss": 1.147,
+      "step": 2044
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5654580593109131,
+      "learning_rate": 0.00018204143331337536,
+      "loss": 1.0996,
+      "step": 2045
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7012020945549011,
+      "learning_rate": 0.0001820242437412725,
+      "loss": 1.0514,
+      "step": 2046
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6119007468223572,
+      "learning_rate": 0.00018200704675874367,
+      "loss": 1.0164,
+      "step": 2047
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6899349689483643,
+      "learning_rate": 0.00018198984236734246,
+      "loss": 1.0248,
+      "step": 2048
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6254860758781433,
+      "learning_rate": 0.00018197263056862325,
+      "loss": 1.1599,
+      "step": 2049
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5398896932601929,
+      "learning_rate": 0.00018195541136414097,
+      "loss": 1.0355,
+      "step": 2050
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5548747777938843,
+      "learning_rate": 0.00018193818475545129,
+      "loss": 1.0282,
+      "step": 2051
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.591780424118042,
+      "learning_rate": 0.00018192095074411058,
+      "loss": 1.2166,
+      "step": 2052
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6100587844848633,
+      "learning_rate": 0.00018190370933167578,
+      "loss": 1.1181,
+      "step": 2053
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.699583113193512,
+      "learning_rate": 0.0001818864605197046,
+      "loss": 1.1278,
+      "step": 2054
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6051424741744995,
+      "learning_rate": 0.00018186920430975534,
+      "loss": 1.0937,
+      "step": 2055
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.550672173500061,
+      "learning_rate": 0.00018185194070338703,
+      "loss": 1.0416,
+      "step": 2056
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5645923018455505,
+      "learning_rate": 0.0001818346697021593,
+      "loss": 1.0335,
+      "step": 2057
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5920846462249756,
+      "learning_rate": 0.0001818173913076325,
+      "loss": 1.173,
+      "step": 2058
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7178803086280823,
+      "learning_rate": 0.00018180010552136769,
+      "loss": 0.9348,
+      "step": 2059
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5697028040885925,
+      "learning_rate": 0.00018178281234492648,
+      "loss": 1.1084,
+      "step": 2060
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6761431097984314,
+      "learning_rate": 0.0001817655117798712,
+      "loss": 1.1577,
+      "step": 2061
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.573343813419342,
+      "learning_rate": 0.00018174820382776488,
+      "loss": 1.1266,
+      "step": 2062
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5408897399902344,
+      "learning_rate": 0.00018173088849017123,
+      "loss": 1.054,
+      "step": 2063
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6230862140655518,
+      "learning_rate": 0.00018171356576865457,
+      "loss": 1.0538,
+      "step": 2064
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5958811640739441,
+      "learning_rate": 0.0001816962356647799,
+      "loss": 1.226,
+      "step": 2065
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5987549424171448,
+      "learning_rate": 0.0001816788981801129,
+      "loss": 1.1471,
+      "step": 2066
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6187098622322083,
+      "learning_rate": 0.0001816615533162199,
+      "loss": 1.242,
+      "step": 2067
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5478816032409668,
+      "learning_rate": 0.0001816442010746679,
+      "loss": 1.0554,
+      "step": 2068
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6243496537208557,
+      "learning_rate": 0.00018162684145702466,
+      "loss": 0.9563,
+      "step": 2069
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6071329712867737,
+      "learning_rate": 0.00018160947446485844,
+      "loss": 1.1325,
+      "step": 2070
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6612725853919983,
+      "learning_rate": 0.00018159210009973826,
+      "loss": 1.1448,
+      "step": 2071
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6862730979919434,
+      "learning_rate": 0.00018157471836323383,
+      "loss": 1.2824,
+      "step": 2072
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.604934811592102,
+      "learning_rate": 0.00018155732925691543,
+      "loss": 1.0633,
+      "step": 2073
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6084781885147095,
+      "learning_rate": 0.00018153993278235416,
+      "loss": 1.1974,
+      "step": 2074
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6043122410774231,
+      "learning_rate": 0.00018152252894112163,
+      "loss": 1.1033,
+      "step": 2075
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6114141941070557,
+      "learning_rate": 0.00018150511773479017,
+      "loss": 1.1386,
+      "step": 2076
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6719943284988403,
+      "learning_rate": 0.0001814876991649328,
+      "loss": 1.0866,
+      "step": 2077
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5567785501480103,
+      "learning_rate": 0.00018147027323312323,
+      "loss": 0.9184,
+      "step": 2078
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5580743551254272,
+      "learning_rate": 0.0001814528399409357,
+      "loss": 1.0469,
+      "step": 2079
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5721143484115601,
+      "learning_rate": 0.00018143539928994532,
+      "loss": 1.0805,
+      "step": 2080
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5623195171356201,
+      "learning_rate": 0.00018141795128172767,
+      "loss": 1.1016,
+      "step": 2081
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6376780271530151,
+      "learning_rate": 0.0001814004959178591,
+      "loss": 1.002,
+      "step": 2082
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6629352569580078,
+      "learning_rate": 0.00018138303319991663,
+      "loss": 1.1561,
+      "step": 2083
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5993563532829285,
+      "learning_rate": 0.0001813655631294779,
+      "loss": 1.063,
+      "step": 2084
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5828850865364075,
+      "learning_rate": 0.00018134808570812125,
+      "loss": 1.1868,
+      "step": 2085
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5925980806350708,
+      "learning_rate": 0.0001813306009374256,
+      "loss": 1.0633,
+      "step": 2086
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6339349746704102,
+      "learning_rate": 0.0001813131088189707,
+      "loss": 1.0335,
+      "step": 2087
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5569525361061096,
+      "learning_rate": 0.00018129560935433678,
+      "loss": 0.9869,
+      "step": 2088
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5449732542037964,
+      "learning_rate": 0.00018127810254510486,
+      "loss": 1.2209,
+      "step": 2089
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6194440722465515,
+      "learning_rate": 0.00018126058839285656,
+      "loss": 1.0826,
+      "step": 2090
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5619876384735107,
+      "learning_rate": 0.0001812430668991742,
+      "loss": 1.1765,
+      "step": 2091
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5702633261680603,
+      "learning_rate": 0.0001812255380656408,
+      "loss": 1.1087,
+      "step": 2092
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6091815233230591,
+      "learning_rate": 0.00018120800189383985,
+      "loss": 1.1604,
+      "step": 2093
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6504197120666504,
+      "learning_rate": 0.00018119045838535573,
+      "loss": 0.9817,
+      "step": 2094
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5884442329406738,
+      "learning_rate": 0.00018117290754177345,
+      "loss": 0.9892,
+      "step": 2095
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5776565074920654,
+      "learning_rate": 0.00018115534936467853,
+      "loss": 1.0363,
+      "step": 2096
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6103533506393433,
+      "learning_rate": 0.00018113778385565733,
+      "loss": 1.244,
+      "step": 2097
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6736485958099365,
+      "learning_rate": 0.0001811202110162967,
+      "loss": 1.0085,
+      "step": 2098
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7142513990402222,
+      "learning_rate": 0.00018110263084818436,
+      "loss": 1.1899,
+      "step": 2099
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6481114029884338,
+      "learning_rate": 0.0001810850433529085,
+      "loss": 1.1406,
+      "step": 2100
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5257217288017273,
+      "learning_rate": 0.0001810674485320581,
+      "loss": 1.1025,
+      "step": 2101
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.538659393787384,
+      "learning_rate": 0.0001810498463872227,
+      "loss": 1.0308,
+      "step": 2102
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5943474769592285,
+      "learning_rate": 0.0001810322369199926,
+      "loss": 1.0064,
+      "step": 2103
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6814498901367188,
+      "learning_rate": 0.00018101462013195868,
+      "loss": 1.0761,
+      "step": 2104
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6166239976882935,
+      "learning_rate": 0.00018099699602471255,
+      "loss": 0.9444,
+      "step": 2105
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6369709372520447,
+      "learning_rate": 0.00018097936459984647,
+      "loss": 1.1837,
+      "step": 2106
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5224598050117493,
+      "learning_rate": 0.00018096172585895323,
+      "loss": 1.2496,
+      "step": 2107
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6199701428413391,
+      "learning_rate": 0.0001809440798036265,
+      "loss": 1.0938,
+      "step": 2108
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6109820008277893,
+      "learning_rate": 0.00018092642643546047,
+      "loss": 1.0366,
+      "step": 2109
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6200912594795227,
+      "learning_rate": 0.00018090876575604998,
+      "loss": 1.094,
+      "step": 2110
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5744567513465881,
+      "learning_rate": 0.00018089109776699066,
+      "loss": 1.0253,
+      "step": 2111
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6033176779747009,
+      "learning_rate": 0.00018087342246987862,
+      "loss": 0.9497,
+      "step": 2112
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6371310949325562,
+      "learning_rate": 0.0001808557398663108,
+      "loss": 1.0781,
+      "step": 2113
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6146370768547058,
+      "learning_rate": 0.00018083804995788465,
+      "loss": 1.0779,
+      "step": 2114
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5866037011146545,
+      "learning_rate": 0.00018082035274619843,
+      "loss": 1.1211,
+      "step": 2115
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5422040820121765,
+      "learning_rate": 0.0001808026482328509,
+      "loss": 0.8366,
+      "step": 2116
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5723130106925964,
+      "learning_rate": 0.0001807849364194416,
+      "loss": 1.0537,
+      "step": 2117
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5562235713005066,
+      "learning_rate": 0.00018076721730757073,
+      "loss": 1.0842,
+      "step": 2118
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.608705997467041,
+      "learning_rate": 0.0001807494908988391,
+      "loss": 1.1393,
+      "step": 2119
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6037261486053467,
+      "learning_rate": 0.0001807317571948481,
+      "loss": 1.3242,
+      "step": 2120
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6175596117973328,
+      "learning_rate": 0.0001807140161972,
+      "loss": 1.1733,
+      "step": 2121
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.615269660949707,
+      "learning_rate": 0.00018069626790749753,
+      "loss": 1.0321,
+      "step": 2122
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.603236734867096,
+      "learning_rate": 0.00018067851232734413,
+      "loss": 1.2018,
+      "step": 2123
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5883955955505371,
+      "learning_rate": 0.00018066074945834396,
+      "loss": 1.1032,
+      "step": 2124
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5292845368385315,
+      "learning_rate": 0.00018064297930210174,
+      "loss": 0.9332,
+      "step": 2125
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6402313709259033,
+      "learning_rate": 0.000180625201860223,
+      "loss": 1.0735,
+      "step": 2126
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5985416769981384,
+      "learning_rate": 0.00018060741713431372,
+      "loss": 1.0645,
+      "step": 2127
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5230291485786438,
+      "learning_rate": 0.00018058962512598075,
+      "loss": 0.9831,
+      "step": 2128
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6098759770393372,
+      "learning_rate": 0.0001805718258368314,
+      "loss": 1.1876,
+      "step": 2129
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7077792882919312,
+      "learning_rate": 0.00018055401926847379,
+      "loss": 0.9392,
+      "step": 2130
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6847090721130371,
+      "learning_rate": 0.00018053620542251666,
+      "loss": 1.0481,
+      "step": 2131
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5998674631118774,
+      "learning_rate": 0.00018051838430056935,
+      "loss": 0.916,
+      "step": 2132
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6179681420326233,
+      "learning_rate": 0.0001805005559042419,
+      "loss": 1.0466,
+      "step": 2133
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6567022204399109,
+      "learning_rate": 0.00018048272023514503,
+      "loss": 1.2206,
+      "step": 2134
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6526135206222534,
+      "learning_rate": 0.00018046487729489008,
+      "loss": 1.2091,
+      "step": 2135
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.605446457862854,
+      "learning_rate": 0.00018044702708508908,
+      "loss": 1.2436,
+      "step": 2136
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6243354678153992,
+      "learning_rate": 0.00018042916960735463,
+      "loss": 1.3055,
+      "step": 2137
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5263623595237732,
+      "learning_rate": 0.0001804113048633001,
+      "loss": 0.8699,
+      "step": 2138
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6840293407440186,
+      "learning_rate": 0.0001803934328545395,
+      "loss": 1.0613,
+      "step": 2139
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6112801432609558,
+      "learning_rate": 0.00018037555358268737,
+      "loss": 1.0718,
+      "step": 2140
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5487591028213501,
+      "learning_rate": 0.0001803576670493591,
+      "loss": 0.9538,
+      "step": 2141
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5530500411987305,
+      "learning_rate": 0.00018033977325617058,
+      "loss": 0.9213,
+      "step": 2142
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5962365865707397,
+      "learning_rate": 0.00018032187220473847,
+      "loss": 1.0206,
+      "step": 2143
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6077033877372742,
+      "learning_rate": 0.00018030396389667996,
+      "loss": 1.1739,
+      "step": 2144
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6256138682365417,
+      "learning_rate": 0.00018028604833361298,
+      "loss": 1.2033,
+      "step": 2145
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5926710367202759,
+      "learning_rate": 0.0001802681255171561,
+      "loss": 0.9579,
+      "step": 2146
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.603348433971405,
+      "learning_rate": 0.00018025019544892857,
+      "loss": 1.2326,
+      "step": 2147
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.576933741569519,
+      "learning_rate": 0.00018023225813055028,
+      "loss": 1.1138,
+      "step": 2148
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6577118635177612,
+      "learning_rate": 0.00018021431356364173,
+      "loss": 1.1375,
+      "step": 2149
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6184905767440796,
+      "learning_rate": 0.00018019636174982414,
+      "loss": 1.2606,
+      "step": 2150
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5773909687995911,
+      "learning_rate": 0.00018017840269071928,
+      "loss": 1.1278,
+      "step": 2151
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.593681812286377,
+      "learning_rate": 0.00018016043638794974,
+      "loss": 1.0683,
+      "step": 2152
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5299533009529114,
+      "learning_rate": 0.00018014246284313865,
+      "loss": 1.0488,
+      "step": 2153
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5954784750938416,
+      "learning_rate": 0.00018012448205790977,
+      "loss": 0.9962,
+      "step": 2154
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5385948419570923,
+      "learning_rate": 0.00018010649403388764,
+      "loss": 0.9248,
+      "step": 2155
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5898330807685852,
+      "learning_rate": 0.0001800884987726973,
+      "loss": 1.1551,
+      "step": 2156
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6078624129295349,
+      "learning_rate": 0.0001800704962759646,
+      "loss": 1.0074,
+      "step": 2157
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5408656597137451,
+      "learning_rate": 0.00018005248654531588,
+      "loss": 0.9395,
+      "step": 2158
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5705100893974304,
+      "learning_rate": 0.00018003446958237828,
+      "loss": 1.0721,
+      "step": 2159
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5572664737701416,
+      "learning_rate": 0.0001800164453887795,
+      "loss": 0.8564,
+      "step": 2160
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6223136186599731,
+      "learning_rate": 0.0001799984139661479,
+      "loss": 1.0628,
+      "step": 2161
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5901376008987427,
+      "learning_rate": 0.00017998037531611263,
+      "loss": 1.1279,
+      "step": 2162
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.581200122833252,
+      "learning_rate": 0.00017996232944030323,
+      "loss": 1.2537,
+      "step": 2163
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6000811457633972,
+      "learning_rate": 0.00017994427634035015,
+      "loss": 1.0852,
+      "step": 2164
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.608558177947998,
+      "learning_rate": 0.00017992621601788428,
+      "loss": 1.1352,
+      "step": 2165
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6038708090782166,
+      "learning_rate": 0.00017990814847453738,
+      "loss": 1.3391,
+      "step": 2166
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.588092029094696,
+      "learning_rate": 0.00017989007371194173,
+      "loss": 1.2182,
+      "step": 2167
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5584761500358582,
+      "learning_rate": 0.00017987199173173023,
+      "loss": 1.1986,
+      "step": 2168
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6125190258026123,
+      "learning_rate": 0.0001798539025355365,
+      "loss": 1.1643,
+      "step": 2169
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5272698402404785,
+      "learning_rate": 0.00017983580612499482,
+      "loss": 0.9866,
+      "step": 2170
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6117221117019653,
+      "learning_rate": 0.00017981770250174005,
+      "loss": 0.988,
+      "step": 2171
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5380080342292786,
+      "learning_rate": 0.00017979959166740782,
+      "loss": 1.0272,
+      "step": 2172
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6345922350883484,
+      "learning_rate": 0.00017978147362363432,
+      "loss": 0.9944,
+      "step": 2173
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.685167670249939,
+      "learning_rate": 0.0001797633483720564,
+      "loss": 1.0927,
+      "step": 2174
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.682863712310791,
+      "learning_rate": 0.00017974521591431156,
+      "loss": 1.1548,
+      "step": 2175
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5539425015449524,
+      "learning_rate": 0.00017972707625203796,
+      "loss": 1.1734,
+      "step": 2176
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6049844026565552,
+      "learning_rate": 0.00017970892938687445,
+      "loss": 1.0368,
+      "step": 2177
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5493163466453552,
+      "learning_rate": 0.00017969077532046048,
+      "loss": 1.0534,
+      "step": 2178
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5946351885795593,
+      "learning_rate": 0.00017967261405443617,
+      "loss": 1.0142,
+      "step": 2179
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5894339680671692,
+      "learning_rate": 0.00017965444559044227,
+      "loss": 1.118,
+      "step": 2180
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5812546014785767,
+      "learning_rate": 0.00017963626993012022,
+      "loss": 1.1465,
+      "step": 2181
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7265496850013733,
+      "learning_rate": 0.0001796180870751121,
+      "loss": 0.9756,
+      "step": 2182
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5403939485549927,
+      "learning_rate": 0.00017959989702706058,
+      "loss": 0.9606,
+      "step": 2183
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6526712775230408,
+      "learning_rate": 0.0001795816997876091,
+      "loss": 1.0271,
+      "step": 2184
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5255093574523926,
+      "learning_rate": 0.0001795634953584016,
+      "loss": 0.8756,
+      "step": 2185
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6554507613182068,
+      "learning_rate": 0.0001795452837410828,
+      "loss": 1.2341,
+      "step": 2186
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6162861585617065,
+      "learning_rate": 0.000179527064937298,
+      "loss": 1.2453,
+      "step": 2187
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5804990530014038,
+      "learning_rate": 0.0001795088389486931,
+      "loss": 1.172,
+      "step": 2188
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6742739677429199,
+      "learning_rate": 0.00017949060577691488,
+      "loss": 1.0933,
+      "step": 2189
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5167546272277832,
+      "learning_rate": 0.00017947236542361044,
+      "loss": 1.0269,
+      "step": 2190
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6206005215644836,
+      "learning_rate": 0.00017945411789042776,
+      "loss": 0.9725,
+      "step": 2191
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.58721524477005,
+      "learning_rate": 0.0001794358631790154,
+      "loss": 1.0485,
+      "step": 2192
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5348027348518372,
+      "learning_rate": 0.00017941760129102255,
+      "loss": 0.9812,
+      "step": 2193
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5457707047462463,
+      "learning_rate": 0.00017939933222809915,
+      "loss": 1.0419,
+      "step": 2194
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6082179546356201,
+      "learning_rate": 0.00017938105599189558,
+      "loss": 1.1514,
+      "step": 2195
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5660126209259033,
+      "learning_rate": 0.00017936277258406307,
+      "loss": 1.0089,
+      "step": 2196
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5888327956199646,
+      "learning_rate": 0.00017934448200625342,
+      "loss": 0.9221,
+      "step": 2197
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.638171374797821,
+      "learning_rate": 0.00017932618426011907,
+      "loss": 1.0189,
+      "step": 2198
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7061122059822083,
+      "learning_rate": 0.00017930787934731311,
+      "loss": 1.1216,
+      "step": 2199
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5726503729820251,
+      "learning_rate": 0.00017928956726948926,
+      "loss": 1.075,
+      "step": 2200
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5943982601165771,
+      "learning_rate": 0.000179271248028302,
+      "loss": 0.8889,
+      "step": 2201
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6812533736228943,
+      "learning_rate": 0.00017925292162540628,
+      "loss": 1.2345,
+      "step": 2202
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6768470406532288,
+      "learning_rate": 0.00017923458806245782,
+      "loss": 1.2374,
+      "step": 2203
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5991680026054382,
+      "learning_rate": 0.00017921624734111292,
+      "loss": 1.08,
+      "step": 2204
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6172266602516174,
+      "learning_rate": 0.00017919789946302866,
+      "loss": 1.0463,
+      "step": 2205
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8502815961837769,
+      "learning_rate": 0.00017917954442986256,
+      "loss": 1.089,
+      "step": 2206
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5796491503715515,
+      "learning_rate": 0.00017916118224327296,
+      "loss": 1.1364,
+      "step": 2207
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6637051105499268,
+      "learning_rate": 0.00017914281290491873,
+      "loss": 1.205,
+      "step": 2208
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6383735537528992,
+      "learning_rate": 0.0001791244364164595,
+      "loss": 0.9944,
+      "step": 2209
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5945635437965393,
+      "learning_rate": 0.0001791060527795554,
+      "loss": 1.0487,
+      "step": 2210
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6733584403991699,
+      "learning_rate": 0.00017908766199586733,
+      "loss": 1.1305,
+      "step": 2211
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5791917443275452,
+      "learning_rate": 0.00017906926406705678,
+      "loss": 1.17,
+      "step": 2212
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5425911545753479,
+      "learning_rate": 0.00017905085899478593,
+      "loss": 1.1022,
+      "step": 2213
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6071683168411255,
+      "learning_rate": 0.00017903244678071757,
+      "loss": 1.1017,
+      "step": 2214
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6542885899543762,
+      "learning_rate": 0.00017901402742651512,
+      "loss": 1.1805,
+      "step": 2215
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6654801964759827,
+      "learning_rate": 0.00017899560093384263,
+      "loss": 1.2689,
+      "step": 2216
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5979436039924622,
+      "learning_rate": 0.00017897716730436492,
+      "loss": 0.9942,
+      "step": 2217
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6463541388511658,
+      "learning_rate": 0.00017895872653974729,
+      "loss": 1.0664,
+      "step": 2218
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5794387459754944,
+      "learning_rate": 0.00017894027864165577,
+      "loss": 0.9069,
+      "step": 2219
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6222195029258728,
+      "learning_rate": 0.00017892182361175702,
+      "loss": 1.1433,
+      "step": 2220
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5939921140670776,
+      "learning_rate": 0.0001789033614517184,
+      "loss": 1.1035,
+      "step": 2221
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6801263689994812,
+      "learning_rate": 0.00017888489216320782,
+      "loss": 1.3398,
+      "step": 2222
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6073567867279053,
+      "learning_rate": 0.0001788664157478938,
+      "loss": 1.1463,
+      "step": 2223
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5933486223220825,
+      "learning_rate": 0.00017884793220744575,
+      "loss": 1.0512,
+      "step": 2224
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6283493638038635,
+      "learning_rate": 0.00017882944154353343,
+      "loss": 1.1488,
+      "step": 2225
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5770262479782104,
+      "learning_rate": 0.0001788109437578274,
+      "loss": 1.1998,
+      "step": 2226
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5506201982498169,
+      "learning_rate": 0.00017879243885199884,
+      "loss": 1.1062,
+      "step": 2227
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5642320513725281,
+      "learning_rate": 0.00017877392682771954,
+      "loss": 1.0807,
+      "step": 2228
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5464077591896057,
+      "learning_rate": 0.00017875540768666194,
+      "loss": 0.9634,
+      "step": 2229
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5671146512031555,
+      "learning_rate": 0.0001787368814304992,
+      "loss": 1.189,
+      "step": 2230
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5678768754005432,
+      "learning_rate": 0.00017871834806090501,
+      "loss": 1.0745,
+      "step": 2231
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5408884286880493,
+      "learning_rate": 0.0001786998075795538,
+      "loss": 1.152,
+      "step": 2232
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5448105931282043,
+      "learning_rate": 0.00017868125998812056,
+      "loss": 1.0465,
+      "step": 2233
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6558840870857239,
+      "learning_rate": 0.000178662705288281,
+      "loss": 1.053,
+      "step": 2234
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5681583881378174,
+      "learning_rate": 0.00017864414348171134,
+      "loss": 0.8673,
+      "step": 2235
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.577218234539032,
+      "learning_rate": 0.00017862557457008866,
+      "loss": 1.0605,
+      "step": 2236
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6256998181343079,
+      "learning_rate": 0.00017860699855509044,
+      "loss": 1.087,
+      "step": 2237
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7601686120033264,
+      "learning_rate": 0.000178588415438395,
+      "loss": 1.138,
+      "step": 2238
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6093456745147705,
+      "learning_rate": 0.0001785698252216812,
+      "loss": 1.0941,
+      "step": 2239
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5740756392478943,
+      "learning_rate": 0.00017855122790662854,
+      "loss": 1.2238,
+      "step": 2240
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5732688307762146,
+      "learning_rate": 0.00017853262349491723,
+      "loss": 1.0036,
+      "step": 2241
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5456250905990601,
+      "learning_rate": 0.000178514011988228,
+      "loss": 0.8441,
+      "step": 2242
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5977802276611328,
+      "learning_rate": 0.00017849539338824231,
+      "loss": 1.0905,
+      "step": 2243
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5805135369300842,
+      "learning_rate": 0.00017847676769664232,
+      "loss": 1.0668,
+      "step": 2244
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.544512152671814,
+      "learning_rate": 0.00017845813491511068,
+      "loss": 1.0293,
+      "step": 2245
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6788541078567505,
+      "learning_rate": 0.00017843949504533078,
+      "loss": 1.0246,
+      "step": 2246
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5601325631141663,
+      "learning_rate": 0.00017842084808898661,
+      "loss": 1.0222,
+      "step": 2247
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5927838683128357,
+      "learning_rate": 0.00017840219404776283,
+      "loss": 0.9626,
+      "step": 2248
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.565467894077301,
+      "learning_rate": 0.00017838353292334478,
+      "loss": 0.9815,
+      "step": 2249
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5185115933418274,
+      "learning_rate": 0.0001783648647174183,
+      "loss": 0.9391,
+      "step": 2250
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6661456823348999,
+      "learning_rate": 0.00017834618943167,
+      "loss": 1.1952,
+      "step": 2251
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5766065120697021,
+      "learning_rate": 0.00017832750706778708,
+      "loss": 1.0535,
+      "step": 2252
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5823925733566284,
+      "learning_rate": 0.0001783088176274574,
+      "loss": 1.0761,
+      "step": 2253
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5491415858268738,
+      "learning_rate": 0.0001782901211123694,
+      "loss": 1.0468,
+      "step": 2254
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5908526182174683,
+      "learning_rate": 0.00017827141752421223,
+      "loss": 1.0112,
+      "step": 2255
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5098611116409302,
+      "learning_rate": 0.00017825270686467568,
+      "loss": 0.8527,
+      "step": 2256
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.536832869052887,
+      "learning_rate": 0.00017823398913545014,
+      "loss": 0.9449,
+      "step": 2257
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5211824178695679,
+      "learning_rate": 0.00017821526433822661,
+      "loss": 1.105,
+      "step": 2258
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6235193610191345,
+      "learning_rate": 0.00017819653247469683,
+      "loss": 1.048,
+      "step": 2259
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5279130935668945,
+      "learning_rate": 0.00017817779354655308,
+      "loss": 1.1058,
+      "step": 2260
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5793086886405945,
+      "learning_rate": 0.0001781590475554883,
+      "loss": 1.0004,
+      "step": 2261
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5962411165237427,
+      "learning_rate": 0.00017814029450319616,
+      "loss": 1.0824,
+      "step": 2262
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5811233520507812,
+      "learning_rate": 0.00017812153439137078,
+      "loss": 0.9964,
+      "step": 2263
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5765413641929626,
+      "learning_rate": 0.00017810276722170714,
+      "loss": 1.0507,
+      "step": 2264
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6243969202041626,
+      "learning_rate": 0.00017808399299590067,
+      "loss": 0.9657,
+      "step": 2265
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6792131066322327,
+      "learning_rate": 0.00017806521171564757,
+      "loss": 1.0217,
+      "step": 2266
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6137617230415344,
+      "learning_rate": 0.00017804642338264457,
+      "loss": 1.0808,
+      "step": 2267
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6335213780403137,
+      "learning_rate": 0.00017802762799858912,
+      "loss": 1.0909,
+      "step": 2268
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5913608074188232,
+      "learning_rate": 0.0001780088255651793,
+      "loss": 1.1797,
+      "step": 2269
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5856280326843262,
+      "learning_rate": 0.00017799001608411375,
+      "loss": 1.122,
+      "step": 2270
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5854422450065613,
+      "learning_rate": 0.0001779711995570918,
+      "loss": 1.0876,
+      "step": 2271
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5743233561515808,
+      "learning_rate": 0.0001779523759858135,
+      "loss": 1.0314,
+      "step": 2272
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5637683868408203,
+      "learning_rate": 0.00017793354537197934,
+      "loss": 1.0113,
+      "step": 2273
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5439072847366333,
+      "learning_rate": 0.00017791470771729063,
+      "loss": 1.1764,
+      "step": 2274
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6216204166412354,
+      "learning_rate": 0.00017789586302344924,
+      "loss": 0.9853,
+      "step": 2275
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5557587742805481,
+      "learning_rate": 0.00017787701129215767,
+      "loss": 0.9668,
+      "step": 2276
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6265153884887695,
+      "learning_rate": 0.00017785815252511908,
+      "loss": 1.0337,
+      "step": 2277
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6366235613822937,
+      "learning_rate": 0.00017783928672403723,
+      "loss": 1.3435,
+      "step": 2278
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6350502967834473,
+      "learning_rate": 0.00017782041389061655,
+      "loss": 1.0933,
+      "step": 2279
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6751552820205688,
+      "learning_rate": 0.00017780153402656207,
+      "loss": 1.0766,
+      "step": 2280
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5644265413284302,
+      "learning_rate": 0.00017778264713357952,
+      "loss": 0.9216,
+      "step": 2281
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5799874663352966,
+      "learning_rate": 0.00017776375321337521,
+      "loss": 0.9968,
+      "step": 2282
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6158801913261414,
+      "learning_rate": 0.0001777448522676561,
+      "loss": 1.2637,
+      "step": 2283
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.571167528629303,
+      "learning_rate": 0.00017772594429812976,
+      "loss": 1.0365,
+      "step": 2284
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6398135423660278,
+      "learning_rate": 0.00017770702930650445,
+      "loss": 1.1111,
+      "step": 2285
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5383583307266235,
+      "learning_rate": 0.000177688107294489,
+      "loss": 1.0693,
+      "step": 2286
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.592910647392273,
+      "learning_rate": 0.00017766917826379295,
+      "loss": 1.0026,
+      "step": 2287
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5524596571922302,
+      "learning_rate": 0.0001776502422161264,
+      "loss": 0.9801,
+      "step": 2288
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5932091474533081,
+      "learning_rate": 0.0001776312991532001,
+      "loss": 1.1859,
+      "step": 2289
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6059234738349915,
+      "learning_rate": 0.00017761234907672549,
+      "loss": 1.1686,
+      "step": 2290
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5730354189872742,
+      "learning_rate": 0.00017759339198841455,
+      "loss": 1.1097,
+      "step": 2291
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.545337975025177,
+      "learning_rate": 0.00017757442788998002,
+      "loss": 0.8885,
+      "step": 2292
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7098504304885864,
+      "learning_rate": 0.00017755545678313514,
+      "loss": 1.2253,
+      "step": 2293
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5590181946754456,
+      "learning_rate": 0.00017753647866959385,
+      "loss": 0.8259,
+      "step": 2294
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.804749071598053,
+      "learning_rate": 0.00017751749355107073,
+      "loss": 1.0091,
+      "step": 2295
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.540513277053833,
+      "learning_rate": 0.00017749850142928096,
+      "loss": 0.9203,
+      "step": 2296
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5944012403488159,
+      "learning_rate": 0.00017747950230594039,
+      "loss": 1.0141,
+      "step": 2297
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5729084014892578,
+      "learning_rate": 0.00017746049618276545,
+      "loss": 1.2178,
+      "step": 2298
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6585800051689148,
+      "learning_rate": 0.0001774414830614733,
+      "loss": 0.9808,
+      "step": 2299
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6925854086875916,
+      "learning_rate": 0.00017742246294378159,
+      "loss": 1.0978,
+      "step": 2300
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6280820965766907,
+      "learning_rate": 0.00017740343583140877,
+      "loss": 1.0309,
+      "step": 2301
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6000059247016907,
+      "learning_rate": 0.00017738440172607373,
+      "loss": 1.1107,
+      "step": 2302
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5967001914978027,
+      "learning_rate": 0.00017736536062949615,
+      "loss": 1.0392,
+      "step": 2303
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4993484318256378,
+      "learning_rate": 0.0001773463125433963,
+      "loss": 0.8711,
+      "step": 2304
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5703268647193909,
+      "learning_rate": 0.00017732725746949504,
+      "loss": 1.0827,
+      "step": 2305
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5676262378692627,
+      "learning_rate": 0.0001773081954095139,
+      "loss": 0.9512,
+      "step": 2306
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5920840501785278,
+      "learning_rate": 0.00017728912636517499,
+      "loss": 1.0837,
+      "step": 2307
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5826390385627747,
+      "learning_rate": 0.00017727005033820115,
+      "loss": 0.9103,
+      "step": 2308
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.718740701675415,
+      "learning_rate": 0.00017725096733031576,
+      "loss": 1.1066,
+      "step": 2309
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.593582034111023,
+      "learning_rate": 0.00017723187734324287,
+      "loss": 0.9903,
+      "step": 2310
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5643889904022217,
+      "learning_rate": 0.00017721278037870715,
+      "loss": 1.0689,
+      "step": 2311
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5696119666099548,
+      "learning_rate": 0.00017719367643843388,
+      "loss": 1.01,
+      "step": 2312
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5642480254173279,
+      "learning_rate": 0.00017717456552414902,
+      "loss": 1.1625,
+      "step": 2313
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.506824791431427,
+      "learning_rate": 0.00017715544763757915,
+      "loss": 0.9964,
+      "step": 2314
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7369400262832642,
+      "learning_rate": 0.0001771363227804514,
+      "loss": 1.2391,
+      "step": 2315
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5189867615699768,
+      "learning_rate": 0.00017711719095449366,
+      "loss": 1.0312,
+      "step": 2316
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5520070791244507,
+      "learning_rate": 0.00017709805216143432,
+      "loss": 1.1712,
+      "step": 2317
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5789135098457336,
+      "learning_rate": 0.00017707890640300254,
+      "loss": 1.1366,
+      "step": 2318
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6521427631378174,
+      "learning_rate": 0.00017705975368092794,
+      "loss": 1.0508,
+      "step": 2319
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5626633167266846,
+      "learning_rate": 0.00017704059399694096,
+      "loss": 1.1635,
+      "step": 2320
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.685055136680603,
+      "learning_rate": 0.00017702142735277247,
+      "loss": 1.1603,
+      "step": 2321
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5628002285957336,
+      "learning_rate": 0.0001770022537501541,
+      "loss": 0.9771,
+      "step": 2322
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.631969153881073,
+      "learning_rate": 0.00017698307319081815,
+      "loss": 1.1721,
+      "step": 2323
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.691218912601471,
+      "learning_rate": 0.00017696388567649737,
+      "loss": 1.2345,
+      "step": 2324
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7056717276573181,
+      "learning_rate": 0.0001769446912089253,
+      "loss": 1.0213,
+      "step": 2325
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5915741324424744,
+      "learning_rate": 0.00017692548978983605,
+      "loss": 1.2047,
+      "step": 2326
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5657656192779541,
+      "learning_rate": 0.00017690628142096433,
+      "loss": 1.0482,
+      "step": 2327
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6127615571022034,
+      "learning_rate": 0.00017688706610404555,
+      "loss": 1.1017,
+      "step": 2328
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5989276170730591,
+      "learning_rate": 0.00017686784384081565,
+      "loss": 1.2391,
+      "step": 2329
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.49377620220184326,
+      "learning_rate": 0.0001768486146330113,
+      "loss": 0.9627,
+      "step": 2330
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6099086999893188,
+      "learning_rate": 0.00017682937848236972,
+      "loss": 1.0978,
+      "step": 2331
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6590273976325989,
+      "learning_rate": 0.00017681013539062884,
+      "loss": 1.0869,
+      "step": 2332
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5015077590942383,
+      "learning_rate": 0.00017679088535952709,
+      "loss": 1.082,
+      "step": 2333
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5334069132804871,
+      "learning_rate": 0.00017677162839080363,
+      "loss": 1.0413,
+      "step": 2334
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6499801278114319,
+      "learning_rate": 0.00017675236448619824,
+      "loss": 1.1829,
+      "step": 2335
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.744327962398529,
+      "learning_rate": 0.0001767330936474513,
+      "loss": 1.2113,
+      "step": 2336
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6160250902175903,
+      "learning_rate": 0.0001767138158763038,
+      "loss": 1.1968,
+      "step": 2337
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6244109869003296,
+      "learning_rate": 0.0001766945311744974,
+      "loss": 1.1436,
+      "step": 2338
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5933175683021545,
+      "learning_rate": 0.00017667523954377433,
+      "loss": 0.966,
+      "step": 2339
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.621781051158905,
+      "learning_rate": 0.0001766559409858775,
+      "loss": 1.0309,
+      "step": 2340
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5982803702354431,
+      "learning_rate": 0.00017663663550255044,
+      "loss": 1.2157,
+      "step": 2341
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6001816391944885,
+      "learning_rate": 0.00017661732309553728,
+      "loss": 1.0603,
+      "step": 2342
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5300409197807312,
+      "learning_rate": 0.00017659800376658278,
+      "loss": 1.0337,
+      "step": 2343
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5697085857391357,
+      "learning_rate": 0.00017657867751743234,
+      "loss": 0.9433,
+      "step": 2344
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5476548671722412,
+      "learning_rate": 0.00017655934434983198,
+      "loss": 1.0204,
+      "step": 2345
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5725254416465759,
+      "learning_rate": 0.0001765400042655283,
+      "loss": 1.1516,
+      "step": 2346
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.666840672492981,
+      "learning_rate": 0.00017652065726626862,
+      "loss": 1.0513,
+      "step": 2347
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6322093605995178,
+      "learning_rate": 0.00017650130335380084,
+      "loss": 1.1666,
+      "step": 2348
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6339440941810608,
+      "learning_rate": 0.00017648194252987344,
+      "loss": 1.02,
+      "step": 2349
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5936554074287415,
+      "learning_rate": 0.00017646257479623557,
+      "loss": 1.0905,
+      "step": 2350
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6231417655944824,
+      "learning_rate": 0.000176443200154637,
+      "loss": 1.1442,
+      "step": 2351
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5736232399940491,
+      "learning_rate": 0.00017642381860682812,
+      "loss": 1.0681,
+      "step": 2352
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6252473592758179,
+      "learning_rate": 0.0001764044301545599,
+      "loss": 1.1322,
+      "step": 2353
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.56855708360672,
+      "learning_rate": 0.00017638503479958404,
+      "loss": 1.1346,
+      "step": 2354
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6396982073783875,
+      "learning_rate": 0.0001763656325436528,
+      "loss": 1.1003,
+      "step": 2355
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7210072875022888,
+      "learning_rate": 0.000176346223388519,
+      "loss": 1.1606,
+      "step": 2356
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6793826222419739,
+      "learning_rate": 0.00017632680733593623,
+      "loss": 1.0003,
+      "step": 2357
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6272571682929993,
+      "learning_rate": 0.00017630738438765853,
+      "loss": 1.0742,
+      "step": 2358
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5288051962852478,
+      "learning_rate": 0.00017628795454544075,
+      "loss": 1.093,
+      "step": 2359
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6225223541259766,
+      "learning_rate": 0.0001762685178110382,
+      "loss": 1.0729,
+      "step": 2360
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5822621583938599,
+      "learning_rate": 0.00017624907418620691,
+      "loss": 0.9997,
+      "step": 2361
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5485069751739502,
+      "learning_rate": 0.00017622962367270346,
+      "loss": 1.0984,
+      "step": 2362
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6365689039230347,
+      "learning_rate": 0.00017621016627228516,
+      "loss": 1.087,
+      "step": 2363
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6249110698699951,
+      "learning_rate": 0.00017619070198670981,
+      "loss": 1.0656,
+      "step": 2364
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5620887279510498,
+      "learning_rate": 0.00017617123081773591,
+      "loss": 1.1772,
+      "step": 2365
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5600476264953613,
+      "learning_rate": 0.00017615175276712263,
+      "loss": 1.0348,
+      "step": 2366
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7794972062110901,
+      "learning_rate": 0.00017613226783662965,
+      "loss": 1.0991,
+      "step": 2367
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5799472332000732,
+      "learning_rate": 0.0001761127760280173,
+      "loss": 0.9012,
+      "step": 2368
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6447935104370117,
+      "learning_rate": 0.00017609327734304658,
+      "loss": 1.0947,
+      "step": 2369
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6327322721481323,
+      "learning_rate": 0.0001760737717834791,
+      "loss": 1.0207,
+      "step": 2370
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6323292851448059,
+      "learning_rate": 0.00017605425935107708,
+      "loss": 1.1466,
+      "step": 2371
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6471980214118958,
+      "learning_rate": 0.00017603474004760337,
+      "loss": 1.0487,
+      "step": 2372
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6297856569290161,
+      "learning_rate": 0.00017601521387482135,
+      "loss": 1.2354,
+      "step": 2373
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6297726631164551,
+      "learning_rate": 0.00017599568083449522,
+      "loss": 0.926,
+      "step": 2374
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.716359555721283,
+      "learning_rate": 0.00017597614092838956,
+      "loss": 1.0752,
+      "step": 2375
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6178591251373291,
+      "learning_rate": 0.00017595659415826977,
+      "loss": 1.1487,
+      "step": 2376
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5869417786598206,
+      "learning_rate": 0.00017593704052590173,
+      "loss": 1.0316,
+      "step": 2377
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5557582974433899,
+      "learning_rate": 0.0001759174800330521,
+      "loss": 1.0543,
+      "step": 2378
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5786042213439941,
+      "learning_rate": 0.00017589791268148792,
+      "loss": 1.0099,
+      "step": 2379
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6963510513305664,
+      "learning_rate": 0.0001758783384729771,
+      "loss": 1.1509,
+      "step": 2380
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5620383024215698,
+      "learning_rate": 0.000175858757409288,
+      "loss": 0.8293,
+      "step": 2381
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5947554111480713,
+      "learning_rate": 0.0001758391694921897,
+      "loss": 1.1298,
+      "step": 2382
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5493107438087463,
+      "learning_rate": 0.00017581957472345182,
+      "loss": 1.153,
+      "step": 2383
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.664533793926239,
+      "learning_rate": 0.00017579997310484467,
+      "loss": 1.0192,
+      "step": 2384
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.591524064540863,
+      "learning_rate": 0.0001757803646381391,
+      "loss": 1.1513,
+      "step": 2385
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5664580464363098,
+      "learning_rate": 0.00017576074932510667,
+      "loss": 1.0468,
+      "step": 2386
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.631363034248352,
+      "learning_rate": 0.0001757411271675195,
+      "loss": 1.1021,
+      "step": 2387
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5752427577972412,
+      "learning_rate": 0.00017572149816715032,
+      "loss": 1.0026,
+      "step": 2388
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.561696469783783,
+      "learning_rate": 0.00017570186232577256,
+      "loss": 1.0329,
+      "step": 2389
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.588409423828125,
+      "learning_rate": 0.0001756822196451601,
+      "loss": 0.9594,
+      "step": 2390
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5990773439407349,
+      "learning_rate": 0.00017566257012708762,
+      "loss": 1.0183,
+      "step": 2391
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5430911183357239,
+      "learning_rate": 0.00017564291377333035,
+      "loss": 1.0256,
+      "step": 2392
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6061146259307861,
+      "learning_rate": 0.00017562325058566408,
+      "loss": 1.0068,
+      "step": 2393
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.718323290348053,
+      "learning_rate": 0.00017560358056586535,
+      "loss": 1.1601,
+      "step": 2394
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5919448137283325,
+      "learning_rate": 0.00017558390371571117,
+      "loss": 1.1336,
+      "step": 2395
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6243352293968201,
+      "learning_rate": 0.00017556422003697919,
+      "loss": 1.084,
+      "step": 2396
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5345236659049988,
+      "learning_rate": 0.00017554452953144782,
+      "loss": 1.029,
+      "step": 2397
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6053178906440735,
+      "learning_rate": 0.00017552483220089593,
+      "loss": 1.0874,
+      "step": 2398
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6026574373245239,
+      "learning_rate": 0.0001755051280471031,
+      "loss": 1.1356,
+      "step": 2399
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5435675382614136,
+      "learning_rate": 0.0001754854170718494,
+      "loss": 1.022,
+      "step": 2400
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5904632806777954,
+      "learning_rate": 0.0001754656992769157,
+      "loss": 1.0769,
+      "step": 2401
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6012750864028931,
+      "learning_rate": 0.00017544597466408334,
+      "loss": 1.0033,
+      "step": 2402
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5695236325263977,
+      "learning_rate": 0.00017542624323513436,
+      "loss": 0.9283,
+      "step": 2403
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7243800759315491,
+      "learning_rate": 0.00017540650499185136,
+      "loss": 1.2303,
+      "step": 2404
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6951823830604553,
+      "learning_rate": 0.00017538675993601757,
+      "loss": 1.144,
+      "step": 2405
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6282033920288086,
+      "learning_rate": 0.00017536700806941688,
+      "loss": 0.8576,
+      "step": 2406
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6453385949134827,
+      "learning_rate": 0.00017534724939383372,
+      "loss": 1.1128,
+      "step": 2407
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6555474400520325,
+      "learning_rate": 0.0001753274839110532,
+      "loss": 1.0861,
+      "step": 2408
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5819057822227478,
+      "learning_rate": 0.00017530771162286101,
+      "loss": 0.9768,
+      "step": 2409
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5652881264686584,
+      "learning_rate": 0.00017528793253104348,
+      "loss": 1.1722,
+      "step": 2410
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5035662055015564,
+      "learning_rate": 0.0001752681466373875,
+      "loss": 0.8257,
+      "step": 2411
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6228415369987488,
+      "learning_rate": 0.00017524835394368065,
+      "loss": 0.9524,
+      "step": 2412
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5981736779212952,
+      "learning_rate": 0.0001752285544517111,
+      "loss": 1.1211,
+      "step": 2413
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.623490035533905,
+      "learning_rate": 0.00017520874816326755,
+      "loss": 1.1335,
+      "step": 2414
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6124420762062073,
+      "learning_rate": 0.0001751889350801395,
+      "loss": 1.1728,
+      "step": 2415
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6511784791946411,
+      "learning_rate": 0.00017516911520411684,
+      "loss": 1.0441,
+      "step": 2416
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6015797853469849,
+      "learning_rate": 0.0001751492885369902,
+      "loss": 0.908,
+      "step": 2417
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6543349027633667,
+      "learning_rate": 0.0001751294550805509,
+      "loss": 1.1191,
+      "step": 2418
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6257116198539734,
+      "learning_rate": 0.00017510961483659066,
+      "loss": 1.0538,
+      "step": 2419
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5716103315353394,
+      "learning_rate": 0.00017508976780690203,
+      "loss": 1.0378,
+      "step": 2420
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5667004585266113,
+      "learning_rate": 0.000175069913993278,
+      "loss": 1.0132,
+      "step": 2421
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5886109471321106,
+      "learning_rate": 0.0001750500533975123,
+      "loss": 0.9482,
+      "step": 2422
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6722474694252014,
+      "learning_rate": 0.00017503018602139925,
+      "loss": 1.3397,
+      "step": 2423
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5232262015342712,
+      "learning_rate": 0.00017501031186673366,
+      "loss": 1.035,
+      "step": 2424
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5664156079292297,
+      "learning_rate": 0.00017499043093531112,
+      "loss": 1.0928,
+      "step": 2425
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5467848181724548,
+      "learning_rate": 0.00017497054322892773,
+      "loss": 1.1207,
+      "step": 2426
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.561660647392273,
+      "learning_rate": 0.00017495064874938025,
+      "loss": 0.9222,
+      "step": 2427
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.659675657749176,
+      "learning_rate": 0.00017493074749846602,
+      "loss": 0.7637,
+      "step": 2428
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5654435753822327,
+      "learning_rate": 0.000174910839477983,
+      "loss": 1.1098,
+      "step": 2429
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6611302495002747,
+      "learning_rate": 0.0001748909246897298,
+      "loss": 1.1681,
+      "step": 2430
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5984782576560974,
+      "learning_rate": 0.00017487100313550555,
+      "loss": 1.1746,
+      "step": 2431
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7086513638496399,
+      "learning_rate": 0.00017485107481711012,
+      "loss": 1.1975,
+      "step": 2432
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.639403760433197,
+      "learning_rate": 0.00017483113973634388,
+      "loss": 1.2715,
+      "step": 2433
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6293968558311462,
+      "learning_rate": 0.00017481119789500787,
+      "loss": 1.0939,
+      "step": 2434
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6304031610488892,
+      "learning_rate": 0.00017479124929490366,
+      "loss": 1.0429,
+      "step": 2435
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6312044858932495,
+      "learning_rate": 0.0001747712939378336,
+      "loss": 1.2176,
+      "step": 2436
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7063558101654053,
+      "learning_rate": 0.00017475133182560045,
+      "loss": 1.1881,
+      "step": 2437
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6040260791778564,
+      "learning_rate": 0.00017473136296000772,
+      "loss": 1.1133,
+      "step": 2438
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6029963493347168,
+      "learning_rate": 0.00017471138734285948,
+      "loss": 1.071,
+      "step": 2439
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5627874135971069,
+      "learning_rate": 0.00017469140497596038,
+      "loss": 0.9752,
+      "step": 2440
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5305935740470886,
+      "learning_rate": 0.0001746714158611158,
+      "loss": 1.0148,
+      "step": 2441
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5892313122749329,
+      "learning_rate": 0.00017465142000013155,
+      "loss": 1.0881,
+      "step": 2442
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6180660724639893,
+      "learning_rate": 0.00017463141739481418,
+      "loss": 1.071,
+      "step": 2443
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.559648334980011,
+      "learning_rate": 0.0001746114080469708,
+      "loss": 0.9962,
+      "step": 2444
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6373436450958252,
+      "learning_rate": 0.00017459139195840918,
+      "loss": 1.1598,
+      "step": 2445
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6689919829368591,
+      "learning_rate": 0.00017457136913093761,
+      "loss": 1.1186,
+      "step": 2446
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.49747997522354126,
+      "learning_rate": 0.00017455133956636507,
+      "loss": 0.9106,
+      "step": 2447
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6205312609672546,
+      "learning_rate": 0.0001745313032665011,
+      "loss": 0.9376,
+      "step": 2448
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5468583703041077,
+      "learning_rate": 0.00017451126023315588,
+      "loss": 0.8509,
+      "step": 2449
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.720237672328949,
+      "learning_rate": 0.00017449121046814016,
+      "loss": 1.0091,
+      "step": 2450
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6043667793273926,
+      "learning_rate": 0.0001744711539732654,
+      "loss": 1.0248,
+      "step": 2451
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5988819599151611,
+      "learning_rate": 0.00017445109075034347,
+      "loss": 1.1108,
+      "step": 2452
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.62172931432724,
+      "learning_rate": 0.00017443102080118704,
+      "loss": 1.127,
+      "step": 2453
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5728936791419983,
+      "learning_rate": 0.00017441094412760935,
+      "loss": 1.0426,
+      "step": 2454
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5967807173728943,
+      "learning_rate": 0.00017439086073142414,
+      "loss": 1.2082,
+      "step": 2455
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5857275724411011,
+      "learning_rate": 0.00017437077061444587,
+      "loss": 1.1496,
+      "step": 2456
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.569225549697876,
+      "learning_rate": 0.00017435067377848955,
+      "loss": 0.8923,
+      "step": 2457
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5777881741523743,
+      "learning_rate": 0.00017433057022537085,
+      "loss": 1.0297,
+      "step": 2458
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6602001786231995,
+      "learning_rate": 0.00017431045995690595,
+      "loss": 0.9484,
+      "step": 2459
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6171995997428894,
+      "learning_rate": 0.00017429034297491178,
+      "loss": 1.053,
+      "step": 2460
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5582581758499146,
+      "learning_rate": 0.00017427021928120574,
+      "loss": 0.9963,
+      "step": 2461
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6040249466896057,
+      "learning_rate": 0.0001742500888776059,
+      "loss": 1.0635,
+      "step": 2462
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6399948596954346,
+      "learning_rate": 0.00017422995176593093,
+      "loss": 1.1903,
+      "step": 2463
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6034595966339111,
+      "learning_rate": 0.00017420980794800013,
+      "loss": 1.1586,
+      "step": 2464
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6262980103492737,
+      "learning_rate": 0.00017418965742563332,
+      "loss": 1.0906,
+      "step": 2465
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6708945035934448,
+      "learning_rate": 0.00017416950020065103,
+      "loss": 0.9319,
+      "step": 2466
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7182316780090332,
+      "learning_rate": 0.00017414933627487436,
+      "loss": 1.1271,
+      "step": 2467
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5667241811752319,
+      "learning_rate": 0.000174129165650125,
+      "loss": 1.1542,
+      "step": 2468
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6271225214004517,
+      "learning_rate": 0.0001741089883282252,
+      "loss": 1.0171,
+      "step": 2469
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7024561166763306,
+      "learning_rate": 0.00017408880431099794,
+      "loss": 0.9372,
+      "step": 2470
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5504549145698547,
+      "learning_rate": 0.0001740686136002667,
+      "loss": 1.0443,
+      "step": 2471
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5320385098457336,
+      "learning_rate": 0.00017404841619785558,
+      "loss": 1.0304,
+      "step": 2472
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6917148232460022,
+      "learning_rate": 0.00017402821210558934,
+      "loss": 0.9552,
+      "step": 2473
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5703467726707458,
+      "learning_rate": 0.00017400800132529327,
+      "loss": 1.1731,
+      "step": 2474
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5451155304908752,
+      "learning_rate": 0.0001739877838587933,
+      "loss": 1.0043,
+      "step": 2475
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.653205156326294,
+      "learning_rate": 0.000173967559707916,
+      "loss": 1.0316,
+      "step": 2476
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6300532817840576,
+      "learning_rate": 0.00017394732887448847,
+      "loss": 1.0411,
+      "step": 2477
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5900035500526428,
+      "learning_rate": 0.00017392709136033847,
+      "loss": 1.117,
+      "step": 2478
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6111758351325989,
+      "learning_rate": 0.00017390684716729434,
+      "loss": 1.1896,
+      "step": 2479
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5686002969741821,
+      "learning_rate": 0.000173886596297185,
+      "loss": 1.1497,
+      "step": 2480
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6367338299751282,
+      "learning_rate": 0.00017386633875184006,
+      "loss": 1.2072,
+      "step": 2481
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6269874572753906,
+      "learning_rate": 0.00017384607453308965,
+      "loss": 1.0348,
+      "step": 2482
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6493733525276184,
+      "learning_rate": 0.00017382580364276451,
+      "loss": 1.0231,
+      "step": 2483
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7141128778457642,
+      "learning_rate": 0.00017380552608269604,
+      "loss": 1.1997,
+      "step": 2484
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5094428658485413,
+      "learning_rate": 0.0001737852418547161,
+      "loss": 0.8767,
+      "step": 2485
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.635170578956604,
+      "learning_rate": 0.00017376495096065739,
+      "loss": 1.0596,
+      "step": 2486
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5259461998939514,
+      "learning_rate": 0.00017374465340235298,
+      "loss": 0.9905,
+      "step": 2487
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.47776302695274353,
+      "learning_rate": 0.00017372434918163672,
+      "loss": 0.864,
+      "step": 2488
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5800095200538635,
+      "learning_rate": 0.0001737040383003429,
+      "loss": 0.9926,
+      "step": 2489
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6474664211273193,
+      "learning_rate": 0.00017368372076030655,
+      "loss": 1.183,
+      "step": 2490
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6319321393966675,
+      "learning_rate": 0.0001736633965633632,
+      "loss": 1.1478,
+      "step": 2491
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5537928342819214,
+      "learning_rate": 0.00017364306571134909,
+      "loss": 1.0313,
+      "step": 2492
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6153177618980408,
+      "learning_rate": 0.00017362272820610095,
+      "loss": 0.9352,
+      "step": 2493
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6004303097724915,
+      "learning_rate": 0.00017360238404945616,
+      "loss": 0.8545,
+      "step": 2494
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6483585834503174,
+      "learning_rate": 0.0001735820332432527,
+      "loss": 1.0338,
+      "step": 2495
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5554136633872986,
+      "learning_rate": 0.00017356167578932915,
+      "loss": 1.1218,
+      "step": 2496
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.582006573677063,
+      "learning_rate": 0.00017354131168952473,
+      "loss": 1.0416,
+      "step": 2497
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.582472562789917,
+      "learning_rate": 0.00017352094094567915,
+      "loss": 0.9425,
+      "step": 2498
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5998402237892151,
+      "learning_rate": 0.00017350056355963287,
+      "loss": 0.9139,
+      "step": 2499
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5740723609924316,
+      "learning_rate": 0.00017348017953322683,
+      "loss": 1.2384,
+      "step": 2500
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5838708281517029,
+      "learning_rate": 0.00017345978886830257,
+      "loss": 1.1484,
+      "step": 2501
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6191444396972656,
+      "learning_rate": 0.00017343939156670235,
+      "loss": 0.9293,
+      "step": 2502
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.569363534450531,
+      "learning_rate": 0.00017341898763026888,
+      "loss": 0.8821,
+      "step": 2503
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6086107492446899,
+      "learning_rate": 0.0001733985770608456,
+      "loss": 1.1533,
+      "step": 2504
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6043040156364441,
+      "learning_rate": 0.00017337815986027648,
+      "loss": 1.116,
+      "step": 2505
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5391929149627686,
+      "learning_rate": 0.00017335773603040604,
+      "loss": 1.0267,
+      "step": 2506
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5986438989639282,
+      "learning_rate": 0.00017333730557307953,
+      "loss": 1.194,
+      "step": 2507
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5731180310249329,
+      "learning_rate": 0.00017331686849014267,
+      "loss": 1.1851,
+      "step": 2508
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6088848114013672,
+      "learning_rate": 0.0001732964247834419,
+      "loss": 1.0551,
+      "step": 2509
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5705273747444153,
+      "learning_rate": 0.00017327597445482415,
+      "loss": 0.9899,
+      "step": 2510
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7256391644477844,
+      "learning_rate": 0.00017325551750613696,
+      "loss": 1.2971,
+      "step": 2511
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.659708559513092,
+      "learning_rate": 0.00017323505393922856,
+      "loss": 1.1237,
+      "step": 2512
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6605255603790283,
+      "learning_rate": 0.0001732145837559477,
+      "loss": 1.0982,
+      "step": 2513
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5549720525741577,
+      "learning_rate": 0.0001731941069581437,
+      "loss": 1.0215,
+      "step": 2514
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5785793662071228,
+      "learning_rate": 0.0001731736235476666,
+      "loss": 0.9916,
+      "step": 2515
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5541704297065735,
+      "learning_rate": 0.0001731531335263669,
+      "loss": 1.0952,
+      "step": 2516
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5742295980453491,
+      "learning_rate": 0.0001731326368960958,
+      "loss": 1.049,
+      "step": 2517
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5734418630599976,
+      "learning_rate": 0.000173112133658705,
+      "loss": 1.2567,
+      "step": 2518
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5206436514854431,
+      "learning_rate": 0.00017309162381604691,
+      "loss": 0.9286,
+      "step": 2519
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.624740481376648,
+      "learning_rate": 0.00017307110736997444,
+      "loss": 1.0974,
+      "step": 2520
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8554314970970154,
+      "learning_rate": 0.00017305058432234117,
+      "loss": 1.1071,
+      "step": 2521
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5889794230461121,
+      "learning_rate": 0.00017303005467500117,
+      "loss": 1.1397,
+      "step": 2522
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5917510986328125,
+      "learning_rate": 0.00017300951842980926,
+      "loss": 1.0686,
+      "step": 2523
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6076456308364868,
+      "learning_rate": 0.00017298897558862072,
+      "loss": 1.1758,
+      "step": 2524
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5394027829170227,
+      "learning_rate": 0.00017296842615329152,
+      "loss": 1.0842,
+      "step": 2525
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5625893473625183,
+      "learning_rate": 0.00017294787012567814,
+      "loss": 0.9352,
+      "step": 2526
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6789762377738953,
+      "learning_rate": 0.00017292730750763773,
+      "loss": 1.2374,
+      "step": 2527
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6913893818855286,
+      "learning_rate": 0.00017290673830102802,
+      "loss": 1.1266,
+      "step": 2528
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5793453454971313,
+      "learning_rate": 0.0001728861625077073,
+      "loss": 0.9519,
+      "step": 2529
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5948161482810974,
+      "learning_rate": 0.00017286558012953444,
+      "loss": 1.0246,
+      "step": 2530
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5573294758796692,
+      "learning_rate": 0.00017284499116836902,
+      "loss": 1.0322,
+      "step": 2531
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6499856114387512,
+      "learning_rate": 0.00017282439562607107,
+      "loss": 1.0345,
+      "step": 2532
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5683536529541016,
+      "learning_rate": 0.0001728037935045013,
+      "loss": 0.9092,
+      "step": 2533
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6661322712898254,
+      "learning_rate": 0.000172783184805521,
+      "loss": 1.0459,
+      "step": 2534
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5732278823852539,
+      "learning_rate": 0.0001727625695309921,
+      "loss": 0.981,
+      "step": 2535
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6493289470672607,
+      "learning_rate": 0.00017274194768277697,
+      "loss": 1.1291,
+      "step": 2536
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6122599840164185,
+      "learning_rate": 0.00017272131926273875,
+      "loss": 1.0597,
+      "step": 2537
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6348492503166199,
+      "learning_rate": 0.00017270068427274109,
+      "loss": 1.1,
+      "step": 2538
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5080891847610474,
+      "learning_rate": 0.00017268004271464823,
+      "loss": 0.9227,
+      "step": 2539
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5522621273994446,
+      "learning_rate": 0.00017265939459032504,
+      "loss": 0.9829,
+      "step": 2540
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6143836975097656,
+      "learning_rate": 0.00017263873990163692,
+      "loss": 0.9779,
+      "step": 2541
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5581562519073486,
+      "learning_rate": 0.00017261807865044993,
+      "loss": 0.9664,
+      "step": 2542
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6005426049232483,
+      "learning_rate": 0.0001725974108386307,
+      "loss": 1.0958,
+      "step": 2543
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6451091766357422,
+      "learning_rate": 0.00017257673646804643,
+      "loss": 0.9855,
+      "step": 2544
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6124886274337769,
+      "learning_rate": 0.00017255605554056496,
+      "loss": 1.02,
+      "step": 2545
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6112679243087769,
+      "learning_rate": 0.00017253536805805474,
+      "loss": 0.9499,
+      "step": 2546
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5976582765579224,
+      "learning_rate": 0.00017251467402238464,
+      "loss": 0.8698,
+      "step": 2547
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6313769817352295,
+      "learning_rate": 0.00017249397343542433,
+      "loss": 1.163,
+      "step": 2548
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6301968097686768,
+      "learning_rate": 0.000172473266299044,
+      "loss": 1.0018,
+      "step": 2549
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5670313835144043,
+      "learning_rate": 0.0001724525526151144,
+      "loss": 1.1514,
+      "step": 2550
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5379245281219482,
+      "learning_rate": 0.0001724318323855069,
+      "loss": 0.9412,
+      "step": 2551
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5700402855873108,
+      "learning_rate": 0.00017241110561209342,
+      "loss": 1.2688,
+      "step": 2552
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6314685940742493,
+      "learning_rate": 0.00017239037229674654,
+      "loss": 1.1394,
+      "step": 2553
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6362370252609253,
+      "learning_rate": 0.00017236963244133944,
+      "loss": 1.147,
+      "step": 2554
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6027795672416687,
+      "learning_rate": 0.00017234888604774574,
+      "loss": 1.0329,
+      "step": 2555
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6116631031036377,
+      "learning_rate": 0.00017232813311783988,
+      "loss": 0.9994,
+      "step": 2556
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6491808891296387,
+      "learning_rate": 0.00017230737365349672,
+      "loss": 0.9094,
+      "step": 2557
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5615575909614563,
+      "learning_rate": 0.0001722866076565917,
+      "loss": 1.1796,
+      "step": 2558
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5690613389015198,
+      "learning_rate": 0.000172265835129001,
+      "loss": 1.0572,
+      "step": 2559
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5822938680648804,
+      "learning_rate": 0.00017224505607260128,
+      "loss": 1.1249,
+      "step": 2560
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5287598967552185,
+      "learning_rate": 0.00017222427048926977,
+      "loss": 0.9504,
+      "step": 2561
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5965141654014587,
+      "learning_rate": 0.00017220347838088436,
+      "loss": 1.0806,
+      "step": 2562
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6563447713851929,
+      "learning_rate": 0.00017218267974932348,
+      "loss": 1.0252,
+      "step": 2563
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5706745386123657,
+      "learning_rate": 0.0001721618745964662,
+      "loss": 1.1904,
+      "step": 2564
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6056011915206909,
+      "learning_rate": 0.00017214106292419212,
+      "loss": 1.2825,
+      "step": 2565
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6403053402900696,
+      "learning_rate": 0.00017212024473438147,
+      "loss": 0.9715,
+      "step": 2566
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6166976690292358,
+      "learning_rate": 0.00017209942002891504,
+      "loss": 1.0931,
+      "step": 2567
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6298193335533142,
+      "learning_rate": 0.00017207858880967426,
+      "loss": 1.1611,
+      "step": 2568
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5657613277435303,
+      "learning_rate": 0.0001720577510785411,
+      "loss": 1.0368,
+      "step": 2569
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8520029187202454,
+      "learning_rate": 0.00017203690683739807,
+      "loss": 1.2765,
+      "step": 2570
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5342356562614441,
+      "learning_rate": 0.0001720160560881284,
+      "loss": 1.1253,
+      "step": 2571
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6659888029098511,
+      "learning_rate": 0.00017199519883261583,
+      "loss": 1.0596,
+      "step": 2572
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6437025666236877,
+      "learning_rate": 0.00017197433507274465,
+      "loss": 1.2124,
+      "step": 2573
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6187822818756104,
+      "learning_rate": 0.00017195346481039986,
+      "loss": 1.0649,
+      "step": 2574
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5546347498893738,
+      "learning_rate": 0.0001719325880474669,
+      "loss": 1.1158,
+      "step": 2575
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6503075361251831,
+      "learning_rate": 0.00017191170478583187,
+      "loss": 1.1343,
+      "step": 2576
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6349958777427673,
+      "learning_rate": 0.0001718908150273815,
+      "loss": 0.935,
+      "step": 2577
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6106913089752197,
+      "learning_rate": 0.00017186991877400304,
+      "loss": 1.1899,
+      "step": 2578
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6525137424468994,
+      "learning_rate": 0.00017184901602758433,
+      "loss": 1.0135,
+      "step": 2579
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5244576334953308,
+      "learning_rate": 0.00017182810679001385,
+      "loss": 1.0043,
+      "step": 2580
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5864934921264648,
+      "learning_rate": 0.0001718071910631806,
+      "loss": 1.0766,
+      "step": 2581
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6121466159820557,
+      "learning_rate": 0.0001717862688489742,
+      "loss": 0.9471,
+      "step": 2582
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5585313439369202,
+      "learning_rate": 0.00017176534014928487,
+      "loss": 1.0705,
+      "step": 2583
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5822098851203918,
+      "learning_rate": 0.0001717444049660034,
+      "loss": 1.0575,
+      "step": 2584
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5953527092933655,
+      "learning_rate": 0.00017172346330102115,
+      "loss": 0.9999,
+      "step": 2585
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5756900906562805,
+      "learning_rate": 0.00017170251515623007,
+      "loss": 1.0358,
+      "step": 2586
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.714945912361145,
+      "learning_rate": 0.00017168156053352273,
+      "loss": 1.2321,
+      "step": 2587
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6626083850860596,
+      "learning_rate": 0.0001716605994347923,
+      "loss": 1.1833,
+      "step": 2588
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5595118999481201,
+      "learning_rate": 0.00017163963186193242,
+      "loss": 0.9832,
+      "step": 2589
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5993494391441345,
+      "learning_rate": 0.00017161865781683741,
+      "loss": 1.1693,
+      "step": 2590
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5663042068481445,
+      "learning_rate": 0.0001715976773014022,
+      "loss": 1.1872,
+      "step": 2591
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5583269000053406,
+      "learning_rate": 0.00017157669031752222,
+      "loss": 0.9559,
+      "step": 2592
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6719366312026978,
+      "learning_rate": 0.0001715556968670935,
+      "loss": 0.9072,
+      "step": 2593
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5847353935241699,
+      "learning_rate": 0.00017153469695201277,
+      "loss": 0.9896,
+      "step": 2594
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5907499194145203,
+      "learning_rate": 0.00017151369057417717,
+      "loss": 1.1507,
+      "step": 2595
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5951755046844482,
+      "learning_rate": 0.00017149267773548455,
+      "loss": 1.1303,
+      "step": 2596
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5762290954589844,
+      "learning_rate": 0.00017147165843783325,
+      "loss": 1.0785,
+      "step": 2597
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5002837181091309,
+      "learning_rate": 0.0001714506326831223,
+      "loss": 0.8424,
+      "step": 2598
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5358802676200867,
+      "learning_rate": 0.00017142960047325127,
+      "loss": 0.9938,
+      "step": 2599
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6255794167518616,
+      "learning_rate": 0.00017140856181012027,
+      "loss": 1.0858,
+      "step": 2600
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7021377086639404,
+      "learning_rate": 0.00017138751669562997,
+      "loss": 1.1827,
+      "step": 2601
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6269333958625793,
+      "learning_rate": 0.00017136646513168179,
+      "loss": 1.1467,
+      "step": 2602
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5892650485038757,
+      "learning_rate": 0.00017134540712017752,
+      "loss": 0.9954,
+      "step": 2603
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5547927021980286,
+      "learning_rate": 0.0001713243426630197,
+      "loss": 1.1428,
+      "step": 2604
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5758942365646362,
+      "learning_rate": 0.00017130327176211133,
+      "loss": 1.0432,
+      "step": 2605
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6165540218353271,
+      "learning_rate": 0.0001712821944193561,
+      "loss": 1.1172,
+      "step": 2606
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6256709694862366,
+      "learning_rate": 0.0001712611106366582,
+      "loss": 0.9987,
+      "step": 2607
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6734545230865479,
+      "learning_rate": 0.00017124002041592245,
+      "loss": 0.9803,
+      "step": 2608
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5715333819389343,
+      "learning_rate": 0.0001712189237590542,
+      "loss": 1.1031,
+      "step": 2609
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6795665621757507,
+      "learning_rate": 0.0001711978206679594,
+      "loss": 1.2371,
+      "step": 2610
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6391113996505737,
+      "learning_rate": 0.00017117671114454463,
+      "loss": 1.2347,
+      "step": 2611
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5957304835319519,
+      "learning_rate": 0.00017115559519071704,
+      "loss": 1.138,
+      "step": 2612
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6582696437835693,
+      "learning_rate": 0.0001711344728083843,
+      "loss": 1.2427,
+      "step": 2613
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.610765814781189,
+      "learning_rate": 0.00017111334399945468,
+      "loss": 1.0626,
+      "step": 2614
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6205570101737976,
+      "learning_rate": 0.0001710922087658371,
+      "loss": 0.9739,
+      "step": 2615
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5730193853378296,
+      "learning_rate": 0.00017107106710944096,
+      "loss": 1.0047,
+      "step": 2616
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5567137598991394,
+      "learning_rate": 0.00017104991903217634,
+      "loss": 1.2085,
+      "step": 2617
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5671441555023193,
+      "learning_rate": 0.0001710287645359538,
+      "loss": 1.1132,
+      "step": 2618
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 10472,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 2618,
+  "total_flos": 2.0326700047296102e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}