diff --git "a/checkpoint-5236/trainer_state.json" "b/checkpoint-5236/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-5236/trainer_state.json"
@@ -0,0 +1,36737 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9809014514896868,
+  "eval_steps": 655,
+  "global_step": 5236,
+  "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
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5712023973464966,
+      "learning_rate": 0.00017100760362268454,
+      "loss": 1.052,
+      "step": 2619
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6445497274398804,
+      "learning_rate": 0.00017098643629428036,
+      "loss": 0.9957,
+      "step": 2620
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.959863007068634,
+      "eval_runtime": 17.8087,
+      "eval_samples_per_second": 153.24,
+      "eval_steps_per_second": 38.352,
+      "step": 2620
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5775024890899658,
+      "learning_rate": 0.00017096526255265356,
+      "loss": 1.0435,
+      "step": 2621
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.603359580039978,
+      "learning_rate": 0.0001709440823997171,
+      "loss": 1.1612,
+      "step": 2622
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5576832890510559,
+      "learning_rate": 0.00017092289583738447,
+      "loss": 0.8875,
+      "step": 2623
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5404002070426941,
+      "learning_rate": 0.00017090170286756978,
+      "loss": 1.0764,
+      "step": 2624
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6048089861869812,
+      "learning_rate": 0.00017088050349218768,
+      "loss": 0.9467,
+      "step": 2625
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6076110005378723,
+      "learning_rate": 0.00017085929771315338,
+      "loss": 1.0436,
+      "step": 2626
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6579375863075256,
+      "learning_rate": 0.00017083808553238278,
+      "loss": 1.0591,
+      "step": 2627
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.515089213848114,
+      "learning_rate": 0.00017081686695179221,
+      "loss": 0.9998,
+      "step": 2628
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6199119091033936,
+      "learning_rate": 0.0001707956419732987,
+      "loss": 0.9776,
+      "step": 2629
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6219343543052673,
+      "learning_rate": 0.00017077441059881978,
+      "loss": 1.0895,
+      "step": 2630
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7458560466766357,
+      "learning_rate": 0.00017075317283027355,
+      "loss": 1.0071,
+      "step": 2631
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6140450239181519,
+      "learning_rate": 0.00017073192866957882,
+      "loss": 1.0347,
+      "step": 2632
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6091210246086121,
+      "learning_rate": 0.00017071067811865476,
+      "loss": 0.9943,
+      "step": 2633
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6610286235809326,
+      "learning_rate": 0.0001706894211794213,
+      "loss": 1.2903,
+      "step": 2634
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5978135466575623,
+      "learning_rate": 0.00017066815785379892,
+      "loss": 1.0517,
+      "step": 2635
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6416430473327637,
+      "learning_rate": 0.00017064688814370858,
+      "loss": 1.2506,
+      "step": 2636
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6144441962242126,
+      "learning_rate": 0.0001706256120510719,
+      "loss": 1.0151,
+      "step": 2637
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6584609150886536,
+      "learning_rate": 0.00017060432957781104,
+      "loss": 1.0537,
+      "step": 2638
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6167966723442078,
+      "learning_rate": 0.00017058304072584878,
+      "loss": 1.0422,
+      "step": 2639
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5490273237228394,
+      "learning_rate": 0.00017056174549710842,
+      "loss": 1.1807,
+      "step": 2640
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5920153856277466,
+      "learning_rate": 0.00017054044389351383,
+      "loss": 1.0671,
+      "step": 2641
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6045279502868652,
+      "learning_rate": 0.00017051913591698958,
+      "loss": 1.0548,
+      "step": 2642
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6131229400634766,
+      "learning_rate": 0.0001704978215694607,
+      "loss": 1.0219,
+      "step": 2643
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.578783392906189,
+      "learning_rate": 0.00017047650085285276,
+      "loss": 1.1181,
+      "step": 2644
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5949702858924866,
+      "learning_rate": 0.00017045517376909205,
+      "loss": 1.0815,
+      "step": 2645
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6813076734542847,
+      "learning_rate": 0.00017043384032010522,
+      "loss": 1.2671,
+      "step": 2646
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6607581377029419,
+      "learning_rate": 0.00017041250050781978,
+      "loss": 0.9366,
+      "step": 2647
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6341320276260376,
+      "learning_rate": 0.0001703911543341636,
+      "loss": 1.2367,
+      "step": 2648
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5848129987716675,
+      "learning_rate": 0.00017036980180106515,
+      "loss": 1.0295,
+      "step": 2649
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5940001010894775,
+      "learning_rate": 0.00017034844291045356,
+      "loss": 1.0005,
+      "step": 2650
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5980600118637085,
+      "learning_rate": 0.00017032707766425844,
+      "loss": 1.0023,
+      "step": 2651
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6173316240310669,
+      "learning_rate": 0.00017030570606441005,
+      "loss": 1.1167,
+      "step": 2652
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5713648796081543,
+      "learning_rate": 0.00017028432811283918,
+      "loss": 0.9816,
+      "step": 2653
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.690399169921875,
+      "learning_rate": 0.00017026294381147726,
+      "loss": 1.2074,
+      "step": 2654
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6921668648719788,
+      "learning_rate": 0.00017024155316225614,
+      "loss": 1.1248,
+      "step": 2655
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5611989498138428,
+      "learning_rate": 0.0001702201561671084,
+      "loss": 1.0523,
+      "step": 2656
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6552536487579346,
+      "learning_rate": 0.0001701987528279672,
+      "loss": 1.0417,
+      "step": 2657
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5799632668495178,
+      "learning_rate": 0.0001701773431467661,
+      "loss": 1.0133,
+      "step": 2658
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5619133710861206,
+      "learning_rate": 0.0001701559271254394,
+      "loss": 1.0702,
+      "step": 2659
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6272364854812622,
+      "learning_rate": 0.00017013450476592189,
+      "loss": 1.029,
+      "step": 2660
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6038535237312317,
+      "learning_rate": 0.000170113076070149,
+      "loss": 1.0668,
+      "step": 2661
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5458052158355713,
+      "learning_rate": 0.00017009164104005663,
+      "loss": 1.0212,
+      "step": 2662
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5674341320991516,
+      "learning_rate": 0.00017007019967758138,
+      "loss": 1.0305,
+      "step": 2663
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6523269414901733,
+      "learning_rate": 0.00017004875198466031,
+      "loss": 0.983,
+      "step": 2664
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6157127022743225,
+      "learning_rate": 0.00017002729796323113,
+      "loss": 0.8491,
+      "step": 2665
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.581650972366333,
+      "learning_rate": 0.00017000583761523204,
+      "loss": 1.0964,
+      "step": 2666
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5540935397148132,
+      "learning_rate": 0.00016998437094260194,
+      "loss": 1.0273,
+      "step": 2667
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.565251350402832,
+      "learning_rate": 0.00016996289794728013,
+      "loss": 1.0797,
+      "step": 2668
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5763837099075317,
+      "learning_rate": 0.00016994141863120666,
+      "loss": 1.0146,
+      "step": 2669
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5568943023681641,
+      "learning_rate": 0.000169919932996322,
+      "loss": 0.8508,
+      "step": 2670
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.639254629611969,
+      "learning_rate": 0.00016989844104456727,
+      "loss": 1.1043,
+      "step": 2671
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5605832934379578,
+      "learning_rate": 0.00016987694277788417,
+      "loss": 0.9942,
+      "step": 2672
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6005716323852539,
+      "learning_rate": 0.00016985543819821492,
+      "loss": 0.9229,
+      "step": 2673
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5722012519836426,
+      "learning_rate": 0.0001698339273075024,
+      "loss": 1.0464,
+      "step": 2674
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5585935711860657,
+      "learning_rate": 0.0001698124101076899,
+      "loss": 0.754,
+      "step": 2675
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6257965564727783,
+      "learning_rate": 0.0001697908866007214,
+      "loss": 1.1917,
+      "step": 2676
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5758666396141052,
+      "learning_rate": 0.00016976935678854148,
+      "loss": 1.0265,
+      "step": 2677
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6103315949440002,
+      "learning_rate": 0.00016974782067309522,
+      "loss": 1.1078,
+      "step": 2678
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6202539205551147,
+      "learning_rate": 0.00016972627825632828,
+      "loss": 1.119,
+      "step": 2679
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6085442900657654,
+      "learning_rate": 0.00016970472954018687,
+      "loss": 0.987,
+      "step": 2680
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6156657338142395,
+      "learning_rate": 0.0001696831745266178,
+      "loss": 1.0947,
+      "step": 2681
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5341770648956299,
+      "learning_rate": 0.00016966161321756848,
+      "loss": 0.7939,
+      "step": 2682
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5123429894447327,
+      "learning_rate": 0.0001696400456149868,
+      "loss": 0.9622,
+      "step": 2683
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6400977969169617,
+      "learning_rate": 0.00016961847172082135,
+      "loss": 0.6955,
+      "step": 2684
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.9171468615531921,
+      "learning_rate": 0.0001695968915370211,
+      "loss": 0.8986,
+      "step": 2685
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6452581286430359,
+      "learning_rate": 0.00016957530506553578,
+      "loss": 0.8457,
+      "step": 2686
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6027610301971436,
+      "learning_rate": 0.0001695537123083156,
+      "loss": 0.9135,
+      "step": 2687
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5201352834701538,
+      "learning_rate": 0.00016953211326731132,
+      "loss": 0.8204,
+      "step": 2688
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.611875593662262,
+      "learning_rate": 0.00016951050794447426,
+      "loss": 0.9314,
+      "step": 2689
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7601920962333679,
+      "learning_rate": 0.0001694888963417564,
+      "loss": 0.8574,
+      "step": 2690
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6485564112663269,
+      "learning_rate": 0.0001694672784611102,
+      "loss": 1.0105,
+      "step": 2691
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7118054032325745,
+      "learning_rate": 0.00016944565430448874,
+      "loss": 1.0206,
+      "step": 2692
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.624322772026062,
+      "learning_rate": 0.00016942402387384558,
+      "loss": 0.8937,
+      "step": 2693
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6083493232727051,
+      "learning_rate": 0.00016940238717113494,
+      "loss": 0.9125,
+      "step": 2694
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6026355624198914,
+      "learning_rate": 0.0001693807441983116,
+      "loss": 0.905,
+      "step": 2695
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7009247541427612,
+      "learning_rate": 0.00016935909495733082,
+      "loss": 1.2486,
+      "step": 2696
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5891932845115662,
+      "learning_rate": 0.00016933743945014856,
+      "loss": 0.8315,
+      "step": 2697
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.656340479850769,
+      "learning_rate": 0.0001693157776787212,
+      "loss": 1.048,
+      "step": 2698
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6057400107383728,
+      "learning_rate": 0.0001692941096450058,
+      "loss": 0.973,
+      "step": 2699
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.680121123790741,
+      "learning_rate": 0.00016927243535095997,
+      "loss": 1.0588,
+      "step": 2700
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6800230741500854,
+      "learning_rate": 0.00016925075479854175,
+      "loss": 0.9017,
+      "step": 2701
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.589103102684021,
+      "learning_rate": 0.00016922906798971,
+      "loss": 0.8216,
+      "step": 2702
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6343634724617004,
+      "learning_rate": 0.0001692073749264239,
+      "loss": 0.883,
+      "step": 2703
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6688684821128845,
+      "learning_rate": 0.00016918567561064336,
+      "loss": 0.8563,
+      "step": 2704
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6180203557014465,
+      "learning_rate": 0.00016916397004432872,
+      "loss": 1.0055,
+      "step": 2705
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6281496286392212,
+      "learning_rate": 0.00016914225822944098,
+      "loss": 0.9944,
+      "step": 2706
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7170965671539307,
+      "learning_rate": 0.00016912054016794174,
+      "loss": 1.145,
+      "step": 2707
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6044749021530151,
+      "learning_rate": 0.00016909881586179304,
+      "loss": 0.9927,
+      "step": 2708
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6250945329666138,
+      "learning_rate": 0.00016907708531295755,
+      "loss": 0.8074,
+      "step": 2709
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7008484601974487,
+      "learning_rate": 0.0001690553485233985,
+      "loss": 0.9173,
+      "step": 2710
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7483898997306824,
+      "learning_rate": 0.0001690336054950797,
+      "loss": 0.8804,
+      "step": 2711
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6421265006065369,
+      "learning_rate": 0.00016901185622996554,
+      "loss": 1.0949,
+      "step": 2712
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6270898580551147,
+      "learning_rate": 0.00016899010073002092,
+      "loss": 0.8987,
+      "step": 2713
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6082978844642639,
+      "learning_rate": 0.0001689683389972113,
+      "loss": 0.7974,
+      "step": 2714
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5921205878257751,
+      "learning_rate": 0.00016894657103350274,
+      "loss": 1.0373,
+      "step": 2715
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6575934886932373,
+      "learning_rate": 0.00016892479684086186,
+      "loss": 0.944,
+      "step": 2716
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6679452657699585,
+      "learning_rate": 0.00016890301642125588,
+      "loss": 0.9262,
+      "step": 2717
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7515625357627869,
+      "learning_rate": 0.0001688812297766525,
+      "loss": 0.944,
+      "step": 2718
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6797657608985901,
+      "learning_rate": 0.00016885943690901994,
+      "loss": 0.8558,
+      "step": 2719
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6033437848091125,
+      "learning_rate": 0.00016883763782032718,
+      "loss": 0.9379,
+      "step": 2720
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6078616976737976,
+      "learning_rate": 0.0001688158325125436,
+      "loss": 0.7629,
+      "step": 2721
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6192131638526917,
+      "learning_rate": 0.00016879402098763917,
+      "loss": 1.0707,
+      "step": 2722
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6294496059417725,
+      "learning_rate": 0.00016877220324758448,
+      "loss": 0.8759,
+      "step": 2723
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6831909418106079,
+      "learning_rate": 0.0001687503792943506,
+      "loss": 0.9751,
+      "step": 2724
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6219105124473572,
+      "learning_rate": 0.00016872854912990923,
+      "loss": 0.8283,
+      "step": 2725
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6364472508430481,
+      "learning_rate": 0.00016870671275623256,
+      "loss": 1.0008,
+      "step": 2726
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6604315042495728,
+      "learning_rate": 0.00016868487017529342,
+      "loss": 1.0695,
+      "step": 2727
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6434652805328369,
+      "learning_rate": 0.00016866302138906516,
+      "loss": 0.8471,
+      "step": 2728
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6357761025428772,
+      "learning_rate": 0.00016864116639952168,
+      "loss": 0.9723,
+      "step": 2729
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7642642259597778,
+      "learning_rate": 0.00016861930520863748,
+      "loss": 1.154,
+      "step": 2730
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6624659895896912,
+      "learning_rate": 0.0001685974378183875,
+      "loss": 0.8315,
+      "step": 2731
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6619630455970764,
+      "learning_rate": 0.00016857556423074748,
+      "loss": 0.8667,
+      "step": 2732
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6592692136764526,
+      "learning_rate": 0.0001685536844476935,
+      "loss": 1.0389,
+      "step": 2733
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6404480338096619,
+      "learning_rate": 0.0001685317984712023,
+      "loss": 1.0442,
+      "step": 2734
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5911348462104797,
+      "learning_rate": 0.0001685099063032511,
+      "loss": 0.7707,
+      "step": 2735
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6581184267997742,
+      "learning_rate": 0.00016848800794581775,
+      "loss": 0.9005,
+      "step": 2736
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6644151210784912,
+      "learning_rate": 0.0001684661034008807,
+      "loss": 0.9491,
+      "step": 2737
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6669042706489563,
+      "learning_rate": 0.00016844419267041886,
+      "loss": 0.9936,
+      "step": 2738
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6686713099479675,
+      "learning_rate": 0.00016842227575641173,
+      "loss": 1.0133,
+      "step": 2739
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.639211118221283,
+      "learning_rate": 0.0001684003526608394,
+      "loss": 0.8211,
+      "step": 2740
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6182137727737427,
+      "learning_rate": 0.00016837842338568245,
+      "loss": 0.9588,
+      "step": 2741
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7168389558792114,
+      "learning_rate": 0.00016835648793292216,
+      "loss": 0.7844,
+      "step": 2742
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5751428008079529,
+      "learning_rate": 0.00016833454630454017,
+      "loss": 0.8838,
+      "step": 2743
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6621695756912231,
+      "learning_rate": 0.00016831259850251886,
+      "loss": 1.0437,
+      "step": 2744
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.646742045879364,
+      "learning_rate": 0.00016829064452884108,
+      "loss": 0.9284,
+      "step": 2745
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6294827461242676,
+      "learning_rate": 0.00016826868438549018,
+      "loss": 0.9306,
+      "step": 2746
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6392683386802673,
+      "learning_rate": 0.0001682467180744502,
+      "loss": 0.9498,
+      "step": 2747
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6557241082191467,
+      "learning_rate": 0.00016822474559770564,
+      "loss": 1.1063,
+      "step": 2748
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6353508830070496,
+      "learning_rate": 0.00016820276695724164,
+      "loss": 0.8402,
+      "step": 2749
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6505456566810608,
+      "learning_rate": 0.0001681807821550438,
+      "loss": 0.9349,
+      "step": 2750
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7468292713165283,
+      "learning_rate": 0.0001681587911930983,
+      "loss": 1.1376,
+      "step": 2751
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.608337938785553,
+      "learning_rate": 0.00016813679407339197,
+      "loss": 1.0272,
+      "step": 2752
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6507999897003174,
+      "learning_rate": 0.00016811479079791204,
+      "loss": 1.1247,
+      "step": 2753
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6512243151664734,
+      "learning_rate": 0.00016809278136864646,
+      "loss": 0.901,
+      "step": 2754
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5723342299461365,
+      "learning_rate": 0.0001680707657875836,
+      "loss": 0.8635,
+      "step": 2755
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6601933240890503,
+      "learning_rate": 0.0001680487440567125,
+      "loss": 0.876,
+      "step": 2756
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6232926845550537,
+      "learning_rate": 0.00016802671617802264,
+      "loss": 0.9157,
+      "step": 2757
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6270859241485596,
+      "learning_rate": 0.00016800468215350412,
+      "loss": 0.9454,
+      "step": 2758
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6160213351249695,
+      "learning_rate": 0.00016798264198514767,
+      "loss": 0.8243,
+      "step": 2759
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6409580111503601,
+      "learning_rate": 0.00016796059567494439,
+      "loss": 1.014,
+      "step": 2760
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7464109063148499,
+      "learning_rate": 0.00016793854322488607,
+      "loss": 1.0413,
+      "step": 2761
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.58450847864151,
+      "learning_rate": 0.00016791648463696505,
+      "loss": 0.7682,
+      "step": 2762
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6073700189590454,
+      "learning_rate": 0.0001678944199131742,
+      "loss": 0.8193,
+      "step": 2763
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6749382615089417,
+      "learning_rate": 0.00016787234905550692,
+      "loss": 0.9554,
+      "step": 2764
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6255000829696655,
+      "learning_rate": 0.0001678502720659572,
+      "loss": 0.8979,
+      "step": 2765
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.625436544418335,
+      "learning_rate": 0.00016782818894651956,
+      "loss": 0.9665,
+      "step": 2766
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6645047664642334,
+      "learning_rate": 0.0001678060996991891,
+      "loss": 0.993,
+      "step": 2767
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7271724939346313,
+      "learning_rate": 0.00016778400432596144,
+      "loss": 0.9922,
+      "step": 2768
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7113057374954224,
+      "learning_rate": 0.00016776190282883277,
+      "loss": 1.029,
+      "step": 2769
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.650078535079956,
+      "learning_rate": 0.00016773979520979987,
+      "loss": 0.9139,
+      "step": 2770
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6887482404708862,
+      "learning_rate": 0.00016771768147086,
+      "loss": 1.0467,
+      "step": 2771
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.675455629825592,
+      "learning_rate": 0.00016769556161401103,
+      "loss": 0.9832,
+      "step": 2772
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.690849244594574,
+      "learning_rate": 0.00016767343564125136,
+      "loss": 0.9428,
+      "step": 2773
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7624968886375427,
+      "learning_rate": 0.00016765130355457996,
+      "loss": 1.0352,
+      "step": 2774
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6946026682853699,
+      "learning_rate": 0.00016762916535599631,
+      "loss": 0.9284,
+      "step": 2775
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6219006776809692,
+      "learning_rate": 0.00016760702104750045,
+      "loss": 0.8471,
+      "step": 2776
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6577418446540833,
+      "learning_rate": 0.0001675848706310931,
+      "loss": 1.0124,
+      "step": 2777
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6933895945549011,
+      "learning_rate": 0.00016756271410877532,
+      "loss": 0.8792,
+      "step": 2778
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6292385458946228,
+      "learning_rate": 0.00016754055148254885,
+      "loss": 0.936,
+      "step": 2779
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6794418692588806,
+      "learning_rate": 0.00016751838275441596,
+      "loss": 0.9986,
+      "step": 2780
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6378180980682373,
+      "learning_rate": 0.00016749620792637948,
+      "loss": 0.942,
+      "step": 2781
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6403058767318726,
+      "learning_rate": 0.0001674740270004428,
+      "loss": 0.844,
+      "step": 2782
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7571424841880798,
+      "learning_rate": 0.0001674518399786098,
+      "loss": 0.9293,
+      "step": 2783
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6841951012611389,
+      "learning_rate": 0.00016742964686288495,
+      "loss": 0.7867,
+      "step": 2784
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7075207233428955,
+      "learning_rate": 0.00016740744765527334,
+      "loss": 0.9051,
+      "step": 2785
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6528226733207703,
+      "learning_rate": 0.00016738524235778045,
+      "loss": 0.8037,
+      "step": 2786
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6237984895706177,
+      "learning_rate": 0.00016736303097241245,
+      "loss": 0.9536,
+      "step": 2787
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6472983360290527,
+      "learning_rate": 0.00016734081350117604,
+      "loss": 0.8897,
+      "step": 2788
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6402902007102966,
+      "learning_rate": 0.00016731858994607838,
+      "loss": 0.8731,
+      "step": 2789
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7781495451927185,
+      "learning_rate": 0.0001672963603091273,
+      "loss": 1.0459,
+      "step": 2790
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7584108710289001,
+      "learning_rate": 0.0001672741245923311,
+      "loss": 1.1999,
+      "step": 2791
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6754958629608154,
+      "learning_rate": 0.00016725188279769865,
+      "loss": 0.8714,
+      "step": 2792
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6159543395042419,
+      "learning_rate": 0.00016722963492723936,
+      "loss": 0.9094,
+      "step": 2793
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7796074748039246,
+      "learning_rate": 0.0001672073809829632,
+      "loss": 1.037,
+      "step": 2794
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6288584470748901,
+      "learning_rate": 0.0001671851209668807,
+      "loss": 0.8631,
+      "step": 2795
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6400723457336426,
+      "learning_rate": 0.00016716285488100292,
+      "loss": 0.8373,
+      "step": 2796
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6898773312568665,
+      "learning_rate": 0.00016714058272734148,
+      "loss": 0.9886,
+      "step": 2797
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6610246300697327,
+      "learning_rate": 0.0001671183045079085,
+      "loss": 0.9466,
+      "step": 2798
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7253816723823547,
+      "learning_rate": 0.0001670960202247168,
+      "loss": 1.0565,
+      "step": 2799
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6904633641242981,
+      "learning_rate": 0.00016707372987977953,
+      "loss": 0.9772,
+      "step": 2800
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7199797630310059,
+      "learning_rate": 0.00016705143347511055,
+      "loss": 1.0893,
+      "step": 2801
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.650538980960846,
+      "learning_rate": 0.0001670291310127242,
+      "loss": 0.9298,
+      "step": 2802
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6637884378433228,
+      "learning_rate": 0.00016700682249463534,
+      "loss": 1.1245,
+      "step": 2803
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7162839770317078,
+      "learning_rate": 0.00016698450792285952,
+      "loss": 0.9432,
+      "step": 2804
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6651907563209534,
+      "learning_rate": 0.00016696218729941265,
+      "loss": 0.7817,
+      "step": 2805
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6402869820594788,
+      "learning_rate": 0.00016693986062631127,
+      "loss": 0.9302,
+      "step": 2806
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6910731196403503,
+      "learning_rate": 0.00016691752790557254,
+      "loss": 0.9083,
+      "step": 2807
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6888274550437927,
+      "learning_rate": 0.000166895189139214,
+      "loss": 1.0216,
+      "step": 2808
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7038168907165527,
+      "learning_rate": 0.00016687284432925393,
+      "loss": 0.9906,
+      "step": 2809
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7101284265518188,
+      "learning_rate": 0.000166850493477711,
+      "loss": 0.8463,
+      "step": 2810
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6386854648590088,
+      "learning_rate": 0.00016682813658660446,
+      "loss": 1.073,
+      "step": 2811
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6103797554969788,
+      "learning_rate": 0.00016680577365795415,
+      "loss": 0.9996,
+      "step": 2812
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6694577932357788,
+      "learning_rate": 0.00016678340469378049,
+      "loss": 0.9692,
+      "step": 2813
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6260998845100403,
+      "learning_rate": 0.00016676102969610432,
+      "loss": 0.9769,
+      "step": 2814
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6177822351455688,
+      "learning_rate": 0.00016673864866694716,
+      "loss": 0.8814,
+      "step": 2815
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7221593260765076,
+      "learning_rate": 0.00016671626160833089,
+      "loss": 1.1002,
+      "step": 2816
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7593296766281128,
+      "learning_rate": 0.00016669386852227817,
+      "loss": 0.7499,
+      "step": 2817
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.676494836807251,
+      "learning_rate": 0.00016667146941081205,
+      "loss": 0.942,
+      "step": 2818
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.693713903427124,
+      "learning_rate": 0.00016664906427595616,
+      "loss": 0.992,
+      "step": 2819
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7619845867156982,
+      "learning_rate": 0.00016662665311973466,
+      "loss": 1.044,
+      "step": 2820
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6537740230560303,
+      "learning_rate": 0.00016660423594417231,
+      "loss": 0.988,
+      "step": 2821
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7443823218345642,
+      "learning_rate": 0.00016658181275129436,
+      "loss": 1.0774,
+      "step": 2822
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7338739633560181,
+      "learning_rate": 0.00016655938354312658,
+      "loss": 1.006,
+      "step": 2823
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7927771806716919,
+      "learning_rate": 0.0001665369483216954,
+      "loss": 1.2174,
+      "step": 2824
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6222347021102905,
+      "learning_rate": 0.00016651450708902765,
+      "loss": 0.8883,
+      "step": 2825
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6589560508728027,
+      "learning_rate": 0.00016649205984715077,
+      "loss": 0.9508,
+      "step": 2826
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6752550601959229,
+      "learning_rate": 0.00016646960659809276,
+      "loss": 0.9706,
+      "step": 2827
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6681761741638184,
+      "learning_rate": 0.00016644714734388217,
+      "loss": 0.9949,
+      "step": 2828
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6704165935516357,
+      "learning_rate": 0.00016642468208654805,
+      "loss": 1.1179,
+      "step": 2829
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7063694596290588,
+      "learning_rate": 0.00016640221082811997,
+      "loss": 0.9225,
+      "step": 2830
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7499638199806213,
+      "learning_rate": 0.00016637973357062812,
+      "loss": 0.8692,
+      "step": 2831
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6111122965812683,
+      "learning_rate": 0.0001663572503161032,
+      "loss": 0.9102,
+      "step": 2832
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6734527945518494,
+      "learning_rate": 0.0001663347610665764,
+      "loss": 0.9193,
+      "step": 2833
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7178148031234741,
+      "learning_rate": 0.00016631226582407952,
+      "loss": 1.1397,
+      "step": 2834
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7453985810279846,
+      "learning_rate": 0.0001662897645906449,
+      "loss": 0.7565,
+      "step": 2835
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7964417338371277,
+      "learning_rate": 0.00016626725736830538,
+      "loss": 0.9191,
+      "step": 2836
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6423993706703186,
+      "learning_rate": 0.00016624474415909436,
+      "loss": 0.8625,
+      "step": 2837
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6401289105415344,
+      "learning_rate": 0.00016622222496504577,
+      "loss": 0.7904,
+      "step": 2838
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6781225204467773,
+      "learning_rate": 0.00016619969978819415,
+      "loss": 0.8491,
+      "step": 2839
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.672544538974762,
+      "learning_rate": 0.00016617716863057442,
+      "loss": 1.0306,
+      "step": 2840
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6570728421211243,
+      "learning_rate": 0.0001661546314942222,
+      "loss": 0.8669,
+      "step": 2841
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7074602842330933,
+      "learning_rate": 0.00016613208838117362,
+      "loss": 1.0322,
+      "step": 2842
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7880011200904846,
+      "learning_rate": 0.00016610953929346528,
+      "loss": 1.106,
+      "step": 2843
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7000258564949036,
+      "learning_rate": 0.0001660869842331344,
+      "loss": 0.9041,
+      "step": 2844
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7121238708496094,
+      "learning_rate": 0.00016606442320221863,
+      "loss": 1.0979,
+      "step": 2845
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6337499022483826,
+      "learning_rate": 0.0001660418562027563,
+      "loss": 1.0048,
+      "step": 2846
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6515573263168335,
+      "learning_rate": 0.0001660192832367862,
+      "loss": 0.9387,
+      "step": 2847
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5596105456352234,
+      "learning_rate": 0.0001659967043063477,
+      "loss": 0.8096,
+      "step": 2848
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5970057249069214,
+      "learning_rate": 0.00016597411941348056,
+      "loss": 0.7847,
+      "step": 2849
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6861267685890198,
+      "learning_rate": 0.00016595152856022533,
+      "loss": 0.9259,
+      "step": 2850
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6038246750831604,
+      "learning_rate": 0.00016592893174862286,
+      "loss": 0.7822,
+      "step": 2851
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7510673403739929,
+      "learning_rate": 0.00016590632898071475,
+      "loss": 1.031,
+      "step": 2852
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7487673759460449,
+      "learning_rate": 0.00016588372025854296,
+      "loss": 0.9588,
+      "step": 2853
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6498749852180481,
+      "learning_rate": 0.00016586110558415004,
+      "loss": 0.8258,
+      "step": 2854
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6465758085250854,
+      "learning_rate": 0.0001658384849595792,
+      "loss": 1.0278,
+      "step": 2855
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6948072910308838,
+      "learning_rate": 0.00016581585838687396,
+      "loss": 1.0128,
+      "step": 2856
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7159262895584106,
+      "learning_rate": 0.0001657932258680786,
+      "loss": 0.8581,
+      "step": 2857
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.672397255897522,
+      "learning_rate": 0.0001657705874052378,
+      "loss": 0.8933,
+      "step": 2858
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6281071305274963,
+      "learning_rate": 0.0001657479430003968,
+      "loss": 0.8891,
+      "step": 2859
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7124696969985962,
+      "learning_rate": 0.00016572529265560144,
+      "loss": 0.7881,
+      "step": 2860
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8034883737564087,
+      "learning_rate": 0.000165702636372898,
+      "loss": 0.9732,
+      "step": 2861
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7420786619186401,
+      "learning_rate": 0.00016567997415433338,
+      "loss": 0.9587,
+      "step": 2862
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6224200129508972,
+      "learning_rate": 0.00016565730600195497,
+      "loss": 0.9015,
+      "step": 2863
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7891188859939575,
+      "learning_rate": 0.00016563463191781073,
+      "loss": 1.0276,
+      "step": 2864
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.659827709197998,
+      "learning_rate": 0.00016561195190394908,
+      "loss": 0.9243,
+      "step": 2865
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6081632375717163,
+      "learning_rate": 0.0001655892659624191,
+      "loss": 0.8685,
+      "step": 2866
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7055192589759827,
+      "learning_rate": 0.0001655665740952703,
+      "loss": 0.92,
+      "step": 2867
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6848809719085693,
+      "learning_rate": 0.0001655438763045528,
+      "loss": 0.8801,
+      "step": 2868
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7062857747077942,
+      "learning_rate": 0.00016552117259231714,
+      "loss": 0.8699,
+      "step": 2869
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7173714637756348,
+      "learning_rate": 0.00016549846296061454,
+      "loss": 1.0067,
+      "step": 2870
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6474378108978271,
+      "learning_rate": 0.00016547574741149665,
+      "loss": 0.9772,
+      "step": 2871
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7030521631240845,
+      "learning_rate": 0.0001654530259470157,
+      "loss": 0.906,
+      "step": 2872
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7073749303817749,
+      "learning_rate": 0.00016543029856922449,
+      "loss": 1.0654,
+      "step": 2873
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6579055786132812,
+      "learning_rate": 0.00016540756528017622,
+      "loss": 1.1778,
+      "step": 2874
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7000943422317505,
+      "learning_rate": 0.00016538482608192476,
+      "loss": 1.0288,
+      "step": 2875
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6581107378005981,
+      "learning_rate": 0.00016536208097652448,
+      "loss": 0.7962,
+      "step": 2876
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7475805282592773,
+      "learning_rate": 0.00016533932996603027,
+      "loss": 1.0036,
+      "step": 2877
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5845509171485901,
+      "learning_rate": 0.00016531657305249756,
+      "loss": 0.927,
+      "step": 2878
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7028359174728394,
+      "learning_rate": 0.00016529381023798232,
+      "loss": 0.9387,
+      "step": 2879
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.763886570930481,
+      "learning_rate": 0.00016527104152454095,
+      "loss": 1.0064,
+      "step": 2880
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6850041747093201,
+      "learning_rate": 0.0001652482669142306,
+      "loss": 0.8708,
+      "step": 2881
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8486610054969788,
+      "learning_rate": 0.00016522548640910873,
+      "loss": 1.0059,
+      "step": 2882
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6982050538063049,
+      "learning_rate": 0.0001652027000112335,
+      "loss": 0.8896,
+      "step": 2883
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6296117305755615,
+      "learning_rate": 0.0001651799077226635,
+      "loss": 0.8382,
+      "step": 2884
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6951981782913208,
+      "learning_rate": 0.00016515710954545786,
+      "loss": 0.9466,
+      "step": 2885
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6647318005561829,
+      "learning_rate": 0.0001651343054816763,
+      "loss": 0.9548,
+      "step": 2886
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6394795775413513,
+      "learning_rate": 0.00016511149553337904,
+      "loss": 0.8903,
+      "step": 2887
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6727846264839172,
+      "learning_rate": 0.0001650886797026268,
+      "loss": 0.9943,
+      "step": 2888
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6580429673194885,
+      "learning_rate": 0.00016506585799148092,
+      "loss": 0.9331,
+      "step": 2889
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6348177194595337,
+      "learning_rate": 0.00016504303040200312,
+      "loss": 1.1005,
+      "step": 2890
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.709359884262085,
+      "learning_rate": 0.00016502019693625584,
+      "loss": 0.9764,
+      "step": 2891
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7029663920402527,
+      "learning_rate": 0.0001649973575963019,
+      "loss": 1.0282,
+      "step": 2892
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7602015137672424,
+      "learning_rate": 0.00016497451238420467,
+      "loss": 1.0261,
+      "step": 2893
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6693970561027527,
+      "learning_rate": 0.00016495166130202818,
+      "loss": 1.0764,
+      "step": 2894
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6728341579437256,
+      "learning_rate": 0.00016492880435183685,
+      "loss": 0.8672,
+      "step": 2895
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6348689794540405,
+      "learning_rate": 0.00016490594153569562,
+      "loss": 0.9328,
+      "step": 2896
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7363536357879639,
+      "learning_rate": 0.00016488307285567012,
+      "loss": 0.8982,
+      "step": 2897
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6520256400108337,
+      "learning_rate": 0.00016486019831382634,
+      "loss": 1.1188,
+      "step": 2898
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7071019411087036,
+      "learning_rate": 0.00016483731791223084,
+      "loss": 0.9359,
+      "step": 2899
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.9263062477111816,
+      "learning_rate": 0.00016481443165295085,
+      "loss": 1.1566,
+      "step": 2900
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7108592987060547,
+      "learning_rate": 0.0001647915395380539,
+      "loss": 0.8543,
+      "step": 2901
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7122064232826233,
+      "learning_rate": 0.00016476864156960818,
+      "loss": 0.9805,
+      "step": 2902
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6613624691963196,
+      "learning_rate": 0.00016474573774968244,
+      "loss": 0.9594,
+      "step": 2903
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6157373785972595,
+      "learning_rate": 0.0001647228280803459,
+      "loss": 0.9154,
+      "step": 2904
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6679324507713318,
+      "learning_rate": 0.00016469991256366824,
+      "loss": 1.0046,
+      "step": 2905
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6375901699066162,
+      "learning_rate": 0.00016467699120171987,
+      "loss": 0.9018,
+      "step": 2906
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6826538443565369,
+      "learning_rate": 0.00016465406399657155,
+      "loss": 0.9845,
+      "step": 2907
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7681721448898315,
+      "learning_rate": 0.0001646311309502946,
+      "loss": 0.8873,
+      "step": 2908
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7108680605888367,
+      "learning_rate": 0.00016460819206496087,
+      "loss": 1.0239,
+      "step": 2909
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6449684500694275,
+      "learning_rate": 0.00016458524734264284,
+      "loss": 1.0171,
+      "step": 2910
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6971015334129333,
+      "learning_rate": 0.00016456229678541337,
+      "loss": 0.8124,
+      "step": 2911
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6763261556625366,
+      "learning_rate": 0.00016453934039534598,
+      "loss": 0.9602,
+      "step": 2912
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7010252475738525,
+      "learning_rate": 0.00016451637817451457,
+      "loss": 0.8596,
+      "step": 2913
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6786449551582336,
+      "learning_rate": 0.00016449341012499368,
+      "loss": 1.0699,
+      "step": 2914
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.718379557132721,
+      "learning_rate": 0.00016447043624885836,
+      "loss": 0.852,
+      "step": 2915
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6729423999786377,
+      "learning_rate": 0.00016444745654818415,
+      "loss": 1.0049,
+      "step": 2916
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6310647130012512,
+      "learning_rate": 0.00016442447102504716,
+      "loss": 0.852,
+      "step": 2917
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5880166888237,
+      "learning_rate": 0.00016440147968152394,
+      "loss": 0.9871,
+      "step": 2918
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6796461343765259,
+      "learning_rate": 0.00016437848251969173,
+      "loss": 1.1046,
+      "step": 2919
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.750541627407074,
+      "learning_rate": 0.0001643554795416281,
+      "loss": 0.9814,
+      "step": 2920
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7341058850288391,
+      "learning_rate": 0.00016433247074941124,
+      "loss": 1.0621,
+      "step": 2921
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6043815612792969,
+      "learning_rate": 0.00016430945614511996,
+      "loss": 0.9073,
+      "step": 2922
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6640098690986633,
+      "learning_rate": 0.0001642864357308334,
+      "loss": 1.0701,
+      "step": 2923
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6690117716789246,
+      "learning_rate": 0.00016426340950863137,
+      "loss": 0.9208,
+      "step": 2924
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6856228709220886,
+      "learning_rate": 0.00016424037748059416,
+      "loss": 0.998,
+      "step": 2925
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7497897148132324,
+      "learning_rate": 0.00016421733964880252,
+      "loss": 0.8903,
+      "step": 2926
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.8069843053817749,
+      "learning_rate": 0.0001641942960153379,
+      "loss": 0.9125,
+      "step": 2927
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7221735119819641,
+      "learning_rate": 0.00016417124658228208,
+      "loss": 1.0176,
+      "step": 2928
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6690991520881653,
+      "learning_rate": 0.00016414819135171749,
+      "loss": 0.8777,
+      "step": 2929
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7555813789367676,
+      "learning_rate": 0.000164125130325727,
+      "loss": 1.0156,
+      "step": 2930
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.615086555480957,
+      "learning_rate": 0.00016410206350639403,
+      "loss": 0.8144,
+      "step": 2931
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6863670349121094,
+      "learning_rate": 0.00016407899089580262,
+      "loss": 1.0584,
+      "step": 2932
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7557387351989746,
+      "learning_rate": 0.00016405591249603715,
+      "loss": 1.0769,
+      "step": 2933
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6460722088813782,
+      "learning_rate": 0.0001640328283091827,
+      "loss": 0.8902,
+      "step": 2934
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6990112662315369,
+      "learning_rate": 0.00016400973833732477,
+      "loss": 0.9806,
+      "step": 2935
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7056894302368164,
+      "learning_rate": 0.0001639866425825494,
+      "loss": 0.8175,
+      "step": 2936
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6659440994262695,
+      "learning_rate": 0.00016396354104694316,
+      "loss": 0.9374,
+      "step": 2937
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7583017945289612,
+      "learning_rate": 0.00016394043373259311,
+      "loss": 1.0661,
+      "step": 2938
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7467166185379028,
+      "learning_rate": 0.00016391732064158695,
+      "loss": 0.9484,
+      "step": 2939
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6958436369895935,
+      "learning_rate": 0.00016389420177601277,
+      "loss": 0.978,
+      "step": 2940
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7176093459129333,
+      "learning_rate": 0.0001638710771379592,
+      "loss": 0.9416,
+      "step": 2941
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6962888240814209,
+      "learning_rate": 0.00016384794672951545,
+      "loss": 0.9828,
+      "step": 2942
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6958292126655579,
+      "learning_rate": 0.00016382481055277125,
+      "loss": 0.9887,
+      "step": 2943
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7200213074684143,
+      "learning_rate": 0.00016380166860981675,
+      "loss": 1.0077,
+      "step": 2944
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.668902575969696,
+      "learning_rate": 0.00016377852090274276,
+      "loss": 0.9923,
+      "step": 2945
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6741634011268616,
+      "learning_rate": 0.0001637553674336405,
+      "loss": 0.871,
+      "step": 2946
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6794645190238953,
+      "learning_rate": 0.0001637322082046018,
+      "loss": 1.0129,
+      "step": 2947
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.68716961145401,
+      "learning_rate": 0.00016370904321771893,
+      "loss": 0.8878,
+      "step": 2948
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7383739948272705,
+      "learning_rate": 0.00016368587247508472,
+      "loss": 1.0159,
+      "step": 2949
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7284265756607056,
+      "learning_rate": 0.00016366269597879255,
+      "loss": 1.0451,
+      "step": 2950
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6858283281326294,
+      "learning_rate": 0.0001636395137309362,
+      "loss": 0.9618,
+      "step": 2951
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7832133769989014,
+      "learning_rate": 0.00016361632573361014,
+      "loss": 1.0356,
+      "step": 2952
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7388131618499756,
+      "learning_rate": 0.00016359313198890924,
+      "loss": 0.991,
+      "step": 2953
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6978746056556702,
+      "learning_rate": 0.00016356993249892895,
+      "loss": 0.8741,
+      "step": 2954
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7001532316207886,
+      "learning_rate": 0.0001635467272657652,
+      "loss": 0.817,
+      "step": 2955
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6634668111801147,
+      "learning_rate": 0.0001635235162915144,
+      "loss": 0.8336,
+      "step": 2956
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.669437050819397,
+      "learning_rate": 0.00016350029957827355,
+      "loss": 0.9604,
+      "step": 2957
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7847330570220947,
+      "learning_rate": 0.00016347707712814023,
+      "loss": 1.0175,
+      "step": 2958
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6637473702430725,
+      "learning_rate": 0.0001634538489432124,
+      "loss": 0.9076,
+      "step": 2959
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.766559362411499,
+      "learning_rate": 0.00016343061502558856,
+      "loss": 1.0943,
+      "step": 2960
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7539775371551514,
+      "learning_rate": 0.00016340737537736777,
+      "loss": 1.1895,
+      "step": 2961
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6763971447944641,
+      "learning_rate": 0.00016338413000064968,
+      "loss": 1.0199,
+      "step": 2962
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6886430978775024,
+      "learning_rate": 0.00016336087889753432,
+      "loss": 0.9311,
+      "step": 2963
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6830310821533203,
+      "learning_rate": 0.0001633376220701223,
+      "loss": 0.8723,
+      "step": 2964
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6681257486343384,
+      "learning_rate": 0.0001633143595205147,
+      "loss": 0.871,
+      "step": 2965
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7215926051139832,
+      "learning_rate": 0.0001632910912508132,
+      "loss": 0.9047,
+      "step": 2966
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6446928381919861,
+      "learning_rate": 0.00016326781726312,
+      "loss": 0.8068,
+      "step": 2967
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.9273528456687927,
+      "learning_rate": 0.00016324453755953773,
+      "loss": 0.9627,
+      "step": 2968
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.70436692237854,
+      "learning_rate": 0.00016322125214216957,
+      "loss": 1.0719,
+      "step": 2969
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6768920421600342,
+      "learning_rate": 0.00016319796101311925,
+      "loss": 0.8396,
+      "step": 2970
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6657084822654724,
+      "learning_rate": 0.00016317466417449094,
+      "loss": 0.7522,
+      "step": 2971
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7143867015838623,
+      "learning_rate": 0.00016315136162838948,
+      "loss": 1.0002,
+      "step": 2972
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.808310866355896,
+      "learning_rate": 0.00016312805337691998,
+      "loss": 1.0182,
+      "step": 2973
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6517670750617981,
+      "learning_rate": 0.00016310473942218837,
+      "loss": 0.9692,
+      "step": 2974
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6355635523796082,
+      "learning_rate": 0.0001630814197663008,
+      "loss": 0.9723,
+      "step": 2975
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6290289163589478,
+      "learning_rate": 0.00016305809441136414,
+      "loss": 1.1122,
+      "step": 2976
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7016021609306335,
+      "learning_rate": 0.00016303476335948567,
+      "loss": 0.7976,
+      "step": 2977
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7270112037658691,
+      "learning_rate": 0.00016301142661277323,
+      "loss": 0.9878,
+      "step": 2978
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7442315220832825,
+      "learning_rate": 0.00016298808417333516,
+      "loss": 0.9822,
+      "step": 2979
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6447054147720337,
+      "learning_rate": 0.00016296473604328034,
+      "loss": 0.6995,
+      "step": 2980
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.76390141248703,
+      "learning_rate": 0.0001629413822247181,
+      "loss": 1.0782,
+      "step": 2981
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7055835723876953,
+      "learning_rate": 0.00016291802271975838,
+      "loss": 1.006,
+      "step": 2982
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.648766040802002,
+      "learning_rate": 0.00016289465753051153,
+      "loss": 0.8927,
+      "step": 2983
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6227762699127197,
+      "learning_rate": 0.0001628712866590885,
+      "loss": 0.9941,
+      "step": 2984
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7030214667320251,
+      "learning_rate": 0.00016284791010760067,
+      "loss": 0.6903,
+      "step": 2985
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6903812289237976,
+      "learning_rate": 0.00016282452787816002,
+      "loss": 0.8768,
+      "step": 2986
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6601381301879883,
+      "learning_rate": 0.000162801139972879,
+      "loss": 0.95,
+      "step": 2987
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6409837603569031,
+      "learning_rate": 0.00016277774639387053,
+      "loss": 0.8799,
+      "step": 2988
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7373121380805969,
+      "learning_rate": 0.00016275434714324816,
+      "loss": 0.8155,
+      "step": 2989
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6737827062606812,
+      "learning_rate": 0.0001627309422231258,
+      "loss": 0.9852,
+      "step": 2990
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.680059552192688,
+      "learning_rate": 0.00016270753163561801,
+      "loss": 0.9731,
+      "step": 2991
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.722334623336792,
+      "learning_rate": 0.00016268411538283977,
+      "loss": 1.1091,
+      "step": 2992
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6662548184394836,
+      "learning_rate": 0.00016266069346690666,
+      "loss": 0.7737,
+      "step": 2993
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6991024017333984,
+      "learning_rate": 0.00016263726588993467,
+      "loss": 0.8982,
+      "step": 2994
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7102436423301697,
+      "learning_rate": 0.00016261383265404034,
+      "loss": 0.9056,
+      "step": 2995
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6874223947525024,
+      "learning_rate": 0.00016259039376134075,
+      "loss": 0.9987,
+      "step": 2996
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7075768709182739,
+      "learning_rate": 0.00016256694921395348,
+      "loss": 1.0127,
+      "step": 2997
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6447400450706482,
+      "learning_rate": 0.0001625434990139966,
+      "loss": 0.8433,
+      "step": 2998
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7359217405319214,
+      "learning_rate": 0.0001625200431635887,
+      "loss": 1.0163,
+      "step": 2999
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7554381489753723,
+      "learning_rate": 0.0001624965816648489,
+      "loss": 0.9526,
+      "step": 3000
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6736909747123718,
+      "learning_rate": 0.00016247311451989677,
+      "loss": 0.9332,
+      "step": 3001
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6281025409698486,
+      "learning_rate": 0.0001624496417308525,
+      "loss": 0.8652,
+      "step": 3002
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8367699384689331,
+      "learning_rate": 0.00016242616329983665,
+      "loss": 1.0465,
+      "step": 3003
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6902986168861389,
+      "learning_rate": 0.00016240267922897043,
+      "loss": 0.7807,
+      "step": 3004
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6087586879730225,
+      "learning_rate": 0.00016237918952037545,
+      "loss": 0.8017,
+      "step": 3005
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6701231002807617,
+      "learning_rate": 0.0001623556941761739,
+      "loss": 0.9221,
+      "step": 3006
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6698204278945923,
+      "learning_rate": 0.00016233219319848843,
+      "loss": 0.9335,
+      "step": 3007
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7633811831474304,
+      "learning_rate": 0.00016230868658944223,
+      "loss": 0.9768,
+      "step": 3008
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7034717202186584,
+      "learning_rate": 0.000162285174351159,
+      "loss": 0.9893,
+      "step": 3009
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7095836997032166,
+      "learning_rate": 0.0001622616564857629,
+      "loss": 0.8335,
+      "step": 3010
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7027610540390015,
+      "learning_rate": 0.0001622381329953787,
+      "loss": 0.9,
+      "step": 3011
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6865144371986389,
+      "learning_rate": 0.00016221460388213154,
+      "loss": 0.7792,
+      "step": 3012
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7391175627708435,
+      "learning_rate": 0.0001621910691481472,
+      "loss": 0.9275,
+      "step": 3013
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8024181723594666,
+      "learning_rate": 0.00016216752879555187,
+      "loss": 0.956,
+      "step": 3014
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6928214430809021,
+      "learning_rate": 0.00016214398282647232,
+      "loss": 0.9826,
+      "step": 3015
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.669105052947998,
+      "learning_rate": 0.00016212043124303572,
+      "loss": 0.9564,
+      "step": 3016
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7236560583114624,
+      "learning_rate": 0.00016209687404736996,
+      "loss": 0.9277,
+      "step": 3017
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8411710262298584,
+      "learning_rate": 0.0001620733112416032,
+      "loss": 1.0461,
+      "step": 3018
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6422951221466064,
+      "learning_rate": 0.00016204974282786424,
+      "loss": 0.8462,
+      "step": 3019
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7195187211036682,
+      "learning_rate": 0.0001620261688082823,
+      "loss": 0.94,
+      "step": 3020
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7143303155899048,
+      "learning_rate": 0.00016200258918498724,
+      "loss": 0.9178,
+      "step": 3021
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7510295510292053,
+      "learning_rate": 0.0001619790039601093,
+      "loss": 1.0951,
+      "step": 3022
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7327381372451782,
+      "learning_rate": 0.00016195541313577923,
+      "loss": 1.0614,
+      "step": 3023
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7491642236709595,
+      "learning_rate": 0.0001619318167141284,
+      "loss": 1.097,
+      "step": 3024
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7633213996887207,
+      "learning_rate": 0.00016190821469728862,
+      "loss": 0.8888,
+      "step": 3025
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7773520946502686,
+      "learning_rate": 0.0001618846070873921,
+      "loss": 1.1425,
+      "step": 3026
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7367396950721741,
+      "learning_rate": 0.00016186099388657176,
+      "loss": 0.8996,
+      "step": 3027
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7460110783576965,
+      "learning_rate": 0.00016183737509696089,
+      "loss": 1.006,
+      "step": 3028
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7337568402290344,
+      "learning_rate": 0.00016181375072069328,
+      "loss": 0.9945,
+      "step": 3029
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6886031031608582,
+      "learning_rate": 0.0001617901207599033,
+      "loss": 0.9766,
+      "step": 3030
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7199036478996277,
+      "learning_rate": 0.00016176648521672575,
+      "loss": 0.9924,
+      "step": 3031
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7087488770484924,
+      "learning_rate": 0.00016174284409329598,
+      "loss": 0.9453,
+      "step": 3032
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7542042136192322,
+      "learning_rate": 0.00016171919739174984,
+      "loss": 0.8999,
+      "step": 3033
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7130687832832336,
+      "learning_rate": 0.0001616955451142237,
+      "loss": 1.0728,
+      "step": 3034
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7343533039093018,
+      "learning_rate": 0.00016167188726285434,
+      "loss": 0.9288,
+      "step": 3035
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6714083552360535,
+      "learning_rate": 0.00016164822383977914,
+      "loss": 1.029,
+      "step": 3036
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7564050555229187,
+      "learning_rate": 0.00016162455484713598,
+      "loss": 0.873,
+      "step": 3037
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6941550374031067,
+      "learning_rate": 0.0001616008802870632,
+      "loss": 0.862,
+      "step": 3038
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7139540314674377,
+      "learning_rate": 0.0001615772001616997,
+      "loss": 1.0259,
+      "step": 3039
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7256305813789368,
+      "learning_rate": 0.00016155351447318477,
+      "loss": 0.9499,
+      "step": 3040
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7770005464553833,
+      "learning_rate": 0.00016152982322365834,
+      "loss": 0.9313,
+      "step": 3041
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7816600203514099,
+      "learning_rate": 0.00016150612641526078,
+      "loss": 1.0,
+      "step": 3042
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.9455228447914124,
+      "learning_rate": 0.00016148242405013294,
+      "loss": 0.9221,
+      "step": 3043
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7099979519844055,
+      "learning_rate": 0.00016145871613041618,
+      "loss": 0.964,
+      "step": 3044
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.8116803765296936,
+      "learning_rate": 0.00016143500265825242,
+      "loss": 0.9761,
+      "step": 3045
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7324463129043579,
+      "learning_rate": 0.000161411283635784,
+      "loss": 1.0304,
+      "step": 3046
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7201613783836365,
+      "learning_rate": 0.00016138755906515382,
+      "loss": 1.1438,
+      "step": 3047
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6950251460075378,
+      "learning_rate": 0.00016136382894850525,
+      "loss": 0.9848,
+      "step": 3048
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6754145622253418,
+      "learning_rate": 0.00016134009328798218,
+      "loss": 1.0516,
+      "step": 3049
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7823755145072937,
+      "learning_rate": 0.000161316352085729,
+      "loss": 1.0282,
+      "step": 3050
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7333949208259583,
+      "learning_rate": 0.00016129260534389057,
+      "loss": 0.8644,
+      "step": 3051
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7001619338989258,
+      "learning_rate": 0.00016126885306461225,
+      "loss": 0.9158,
+      "step": 3052
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7424895763397217,
+      "learning_rate": 0.00016124509525004002,
+      "loss": 1.1199,
+      "step": 3053
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7295093536376953,
+      "learning_rate": 0.00016122133190232018,
+      "loss": 0.931,
+      "step": 3054
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7016628384590149,
+      "learning_rate": 0.00016119756302359963,
+      "loss": 1.1039,
+      "step": 3055
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7510272264480591,
+      "learning_rate": 0.00016117378861602575,
+      "loss": 0.9644,
+      "step": 3056
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6671194434165955,
+      "learning_rate": 0.00016115000868174644,
+      "loss": 0.7718,
+      "step": 3057
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7244356870651245,
+      "learning_rate": 0.0001611262232229101,
+      "loss": 0.9918,
+      "step": 3058
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6976906657218933,
+      "learning_rate": 0.00016110243224166557,
+      "loss": 0.8962,
+      "step": 3059
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6452173590660095,
+      "learning_rate": 0.00016107863574016222,
+      "loss": 0.8798,
+      "step": 3060
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6883850693702698,
+      "learning_rate": 0.00016105483372055,
+      "loss": 1.0826,
+      "step": 3061
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6843141913414001,
+      "learning_rate": 0.00016103102618497922,
+      "loss": 1.0096,
+      "step": 3062
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7020462155342102,
+      "learning_rate": 0.0001610072131356008,
+      "loss": 0.8949,
+      "step": 3063
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7438017129898071,
+      "learning_rate": 0.00016098339457456607,
+      "loss": 0.9674,
+      "step": 3064
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6534295678138733,
+      "learning_rate": 0.00016095957050402696,
+      "loss": 0.8764,
+      "step": 3065
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6544593572616577,
+      "learning_rate": 0.00016093574092613578,
+      "loss": 0.8218,
+      "step": 3066
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6656880974769592,
+      "learning_rate": 0.0001609119058430454,
+      "loss": 0.999,
+      "step": 3067
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7666613459587097,
+      "learning_rate": 0.00016088806525690926,
+      "loss": 1.0574,
+      "step": 3068
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7083278894424438,
+      "learning_rate": 0.00016086421916988117,
+      "loss": 1.0048,
+      "step": 3069
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6306576728820801,
+      "learning_rate": 0.00016084036758411547,
+      "loss": 0.9398,
+      "step": 3070
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7330958247184753,
+      "learning_rate": 0.00016081651050176704,
+      "loss": 0.9751,
+      "step": 3071
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.901148796081543,
+      "learning_rate": 0.00016079264792499124,
+      "loss": 0.8909,
+      "step": 3072
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7629517316818237,
+      "learning_rate": 0.0001607687798559439,
+      "loss": 1.0978,
+      "step": 3073
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.8078415393829346,
+      "learning_rate": 0.00016074490629678137,
+      "loss": 0.9588,
+      "step": 3074
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.8237608671188354,
+      "learning_rate": 0.00016072102724966046,
+      "loss": 1.1306,
+      "step": 3075
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6664127111434937,
+      "learning_rate": 0.00016069714271673856,
+      "loss": 0.954,
+      "step": 3076
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6480268239974976,
+      "learning_rate": 0.00016067325270017352,
+      "loss": 0.8522,
+      "step": 3077
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7148239612579346,
+      "learning_rate": 0.0001606493572021236,
+      "loss": 0.9399,
+      "step": 3078
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6466790437698364,
+      "learning_rate": 0.00016062545622474764,
+      "loss": 0.7129,
+      "step": 3079
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6735865473747253,
+      "learning_rate": 0.00016060154977020502,
+      "loss": 0.8239,
+      "step": 3080
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7058208584785461,
+      "learning_rate": 0.00016057763784065542,
+      "loss": 0.9673,
+      "step": 3081
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6586633324623108,
+      "learning_rate": 0.0001605537204382593,
+      "loss": 1.0528,
+      "step": 3082
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6823360323905945,
+      "learning_rate": 0.0001605297975651774,
+      "loss": 1.0073,
+      "step": 3083
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7752652168273926,
+      "learning_rate": 0.00016050586922357094,
+      "loss": 1.0964,
+      "step": 3084
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7399365305900574,
+      "learning_rate": 0.00016048193541560183,
+      "loss": 1.0044,
+      "step": 3085
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.689049482345581,
+      "learning_rate": 0.00016045799614343228,
+      "loss": 0.8952,
+      "step": 3086
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7327082753181458,
+      "learning_rate": 0.00016043405140922512,
+      "loss": 1.0547,
+      "step": 3087
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7117489576339722,
+      "learning_rate": 0.00016041010121514357,
+      "loss": 1.0172,
+      "step": 3088
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7927287817001343,
+      "learning_rate": 0.00016038614556335143,
+      "loss": 0.9599,
+      "step": 3089
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6660590767860413,
+      "learning_rate": 0.00016036218445601293,
+      "loss": 0.9384,
+      "step": 3090
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6778466105461121,
+      "learning_rate": 0.00016033821789529284,
+      "loss": 0.8565,
+      "step": 3091
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.694474995136261,
+      "learning_rate": 0.0001603142458833564,
+      "loss": 0.9739,
+      "step": 3092
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.8222754001617432,
+      "learning_rate": 0.00016029026842236935,
+      "loss": 1.0106,
+      "step": 3093
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7755143642425537,
+      "learning_rate": 0.0001602662855144979,
+      "loss": 0.9953,
+      "step": 3094
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6970244646072388,
+      "learning_rate": 0.00016024229716190877,
+      "loss": 0.8954,
+      "step": 3095
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.9485830068588257,
+      "learning_rate": 0.0001602183033667692,
+      "loss": 1.0471,
+      "step": 3096
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7483344078063965,
+      "learning_rate": 0.00016019430413124686,
+      "loss": 1.123,
+      "step": 3097
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7733362913131714,
+      "learning_rate": 0.00016017029945750996,
+      "loss": 0.9967,
+      "step": 3098
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.857623815536499,
+      "learning_rate": 0.0001601462893477272,
+      "loss": 1.0583,
+      "step": 3099
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7281124591827393,
+      "learning_rate": 0.00016012227380406775,
+      "loss": 1.0528,
+      "step": 3100
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.673368513584137,
+      "learning_rate": 0.00016009825282870126,
+      "loss": 1.1355,
+      "step": 3101
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7686245441436768,
+      "learning_rate": 0.0001600742264237979,
+      "loss": 1.1185,
+      "step": 3102
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7917776703834534,
+      "learning_rate": 0.00016005019459152834,
+      "loss": 1.0218,
+      "step": 3103
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.626288652420044,
+      "learning_rate": 0.0001600261573340637,
+      "loss": 0.8897,
+      "step": 3104
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7067705392837524,
+      "learning_rate": 0.0001600021146535756,
+      "loss": 1.0538,
+      "step": 3105
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6985976099967957,
+      "learning_rate": 0.00015997806655223618,
+      "loss": 0.9316,
+      "step": 3106
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6789809465408325,
+      "learning_rate": 0.000159954013032218,
+      "loss": 0.8942,
+      "step": 3107
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7024329900741577,
+      "learning_rate": 0.00015992995409569428,
+      "loss": 0.8517,
+      "step": 3108
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7696155309677124,
+      "learning_rate": 0.0001599058897448385,
+      "loss": 1.0287,
+      "step": 3109
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.697361946105957,
+      "learning_rate": 0.00015988181998182475,
+      "loss": 1.0239,
+      "step": 3110
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6460540294647217,
+      "learning_rate": 0.00015985774480882758,
+      "loss": 0.8986,
+      "step": 3111
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7324708104133606,
+      "learning_rate": 0.00015983366422802213,
+      "loss": 1.0674,
+      "step": 3112
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.8241015076637268,
+      "learning_rate": 0.0001598095782415839,
+      "loss": 0.9858,
+      "step": 3113
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6638197302818298,
+      "learning_rate": 0.00015978548685168892,
+      "loss": 0.8558,
+      "step": 3114
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7642309069633484,
+      "learning_rate": 0.00015976139006051366,
+      "loss": 0.9775,
+      "step": 3115
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7588913440704346,
+      "learning_rate": 0.0001597372878702352,
+      "loss": 0.9801,
+      "step": 3116
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6915333271026611,
+      "learning_rate": 0.00015971318028303104,
+      "loss": 0.9893,
+      "step": 3117
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7214621305465698,
+      "learning_rate": 0.0001596890673010791,
+      "loss": 0.8739,
+      "step": 3118
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6884421706199646,
+      "learning_rate": 0.0001596649489265579,
+      "loss": 0.9756,
+      "step": 3119
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.8056734800338745,
+      "learning_rate": 0.0001596408251616464,
+      "loss": 1.0987,
+      "step": 3120
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7120291590690613,
+      "learning_rate": 0.000159616696008524,
+      "loss": 1.0054,
+      "step": 3121
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7776278853416443,
+      "learning_rate": 0.00015959256146937068,
+      "loss": 1.0415,
+      "step": 3122
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7185929417610168,
+      "learning_rate": 0.00015956842154636687,
+      "loss": 0.9976,
+      "step": 3123
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7339257001876831,
+      "learning_rate": 0.00015954427624169344,
+      "loss": 0.9121,
+      "step": 3124
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.8103228807449341,
+      "learning_rate": 0.00015952012555753178,
+      "loss": 0.9984,
+      "step": 3125
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7015619277954102,
+      "learning_rate": 0.00015949596949606376,
+      "loss": 0.9733,
+      "step": 3126
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.9180814027786255,
+      "learning_rate": 0.0001594718080594718,
+      "loss": 1.0364,
+      "step": 3127
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7444248199462891,
+      "learning_rate": 0.00015944764124993868,
+      "loss": 0.9475,
+      "step": 3128
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6662657856941223,
+      "learning_rate": 0.00015942346906964777,
+      "loss": 0.7493,
+      "step": 3129
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6935024261474609,
+      "learning_rate": 0.00015939929152078287,
+      "loss": 1.0452,
+      "step": 3130
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7142334580421448,
+      "learning_rate": 0.0001593751086055283,
+      "loss": 0.7778,
+      "step": 3131
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7099422812461853,
+      "learning_rate": 0.00015935092032606888,
+      "loss": 0.9093,
+      "step": 3132
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7817436456680298,
+      "learning_rate": 0.00015932672668458986,
+      "loss": 1.115,
+      "step": 3133
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7280943393707275,
+      "learning_rate": 0.00015930252768327696,
+      "loss": 0.998,
+      "step": 3134
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7231597900390625,
+      "learning_rate": 0.00015927832332431646,
+      "loss": 0.92,
+      "step": 3135
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7060204744338989,
+      "learning_rate": 0.0001592541136098951,
+      "loss": 1.0271,
+      "step": 3136
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.717904806137085,
+      "learning_rate": 0.00015922989854220005,
+      "loss": 0.981,
+      "step": 3137
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7248113751411438,
+      "learning_rate": 0.00015920567812341904,
+      "loss": 0.9549,
+      "step": 3138
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6890550851821899,
+      "learning_rate": 0.0001591814523557402,
+      "loss": 1.0144,
+      "step": 3139
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8595572710037231,
+      "learning_rate": 0.00015915722124135227,
+      "loss": 1.1475,
+      "step": 3140
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7589665651321411,
+      "learning_rate": 0.0001591329847824443,
+      "loss": 0.9801,
+      "step": 3141
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6806226372718811,
+      "learning_rate": 0.000159108742981206,
+      "loss": 0.7727,
+      "step": 3142
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.9598560333251953,
+      "learning_rate": 0.00015908449583982743,
+      "loss": 1.101,
+      "step": 3143
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.734533429145813,
+      "learning_rate": 0.00015906024336049922,
+      "loss": 0.9844,
+      "step": 3144
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7020285725593567,
+      "learning_rate": 0.0001590359855454124,
+      "loss": 1.0375,
+      "step": 3145
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8799060583114624,
+      "learning_rate": 0.00015901172239675851,
+      "loss": 0.9193,
+      "step": 3146
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7654073238372803,
+      "learning_rate": 0.00015898745391672967,
+      "loss": 0.9879,
+      "step": 3147
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6300762295722961,
+      "learning_rate": 0.00015896318010751837,
+      "loss": 1.0095,
+      "step": 3148
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7860686779022217,
+      "learning_rate": 0.0001589389009713176,
+      "loss": 1.0894,
+      "step": 3149
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7179495692253113,
+      "learning_rate": 0.0001589146165103208,
+      "loss": 0.9753,
+      "step": 3150
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7529764175415039,
+      "learning_rate": 0.00015889032672672198,
+      "loss": 0.9269,
+      "step": 3151
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6052923798561096,
+      "learning_rate": 0.00015886603162271556,
+      "loss": 0.7845,
+      "step": 3152
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7489050626754761,
+      "learning_rate": 0.00015884173120049653,
+      "loss": 0.9919,
+      "step": 3153
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7350334525108337,
+      "learning_rate": 0.00015881742546226018,
+      "loss": 1.0606,
+      "step": 3154
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7081111669540405,
+      "learning_rate": 0.0001587931144102025,
+      "loss": 0.8885,
+      "step": 3155
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.775686502456665,
+      "learning_rate": 0.00015876879804651984,
+      "loss": 0.9711,
+      "step": 3156
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7543262839317322,
+      "learning_rate": 0.00015874447637340898,
+      "loss": 1.1627,
+      "step": 3157
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7249549627304077,
+      "learning_rate": 0.00015872014939306733,
+      "loss": 0.8961,
+      "step": 3158
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7103540897369385,
+      "learning_rate": 0.00015869581710769266,
+      "loss": 0.9714,
+      "step": 3159
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7673888802528381,
+      "learning_rate": 0.00015867147951948324,
+      "loss": 1.08,
+      "step": 3160
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7711588144302368,
+      "learning_rate": 0.0001586471366306378,
+      "loss": 1.1145,
+      "step": 3161
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.616401195526123,
+      "learning_rate": 0.00015862278844335566,
+      "loss": 0.935,
+      "step": 3162
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6484968066215515,
+      "learning_rate": 0.00015859843495983651,
+      "loss": 0.9124,
+      "step": 3163
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6611524224281311,
+      "learning_rate": 0.00015857407618228054,
+      "loss": 0.7792,
+      "step": 3164
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6965488195419312,
+      "learning_rate": 0.00015854971211288845,
+      "loss": 0.7451,
+      "step": 3165
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.638028085231781,
+      "learning_rate": 0.00015852534275386134,
+      "loss": 0.9095,
+      "step": 3166
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.740543007850647,
+      "learning_rate": 0.0001585009681074009,
+      "loss": 1.0657,
+      "step": 3167
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6465707421302795,
+      "learning_rate": 0.00015847658817570923,
+      "loss": 0.9527,
+      "step": 3168
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.692480742931366,
+      "learning_rate": 0.0001584522029609889,
+      "loss": 0.961,
+      "step": 3169
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7205979824066162,
+      "learning_rate": 0.00015842781246544298,
+      "loss": 1.0782,
+      "step": 3170
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6649343967437744,
+      "learning_rate": 0.00015840341669127504,
+      "loss": 0.8621,
+      "step": 3171
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6777862906455994,
+      "learning_rate": 0.00015837901564068907,
+      "loss": 0.8381,
+      "step": 3172
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7257649898529053,
+      "learning_rate": 0.00015835460931588958,
+      "loss": 1.0025,
+      "step": 3173
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.764773964881897,
+      "learning_rate": 0.00015833019771908154,
+      "loss": 0.9403,
+      "step": 3174
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7970273494720459,
+      "learning_rate": 0.00015830578085247043,
+      "loss": 1.1609,
+      "step": 3175
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7514870762825012,
+      "learning_rate": 0.0001582813587182621,
+      "loss": 0.8338,
+      "step": 3176
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6733028888702393,
+      "learning_rate": 0.00015825693131866304,
+      "loss": 0.8313,
+      "step": 3177
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7544176578521729,
+      "learning_rate": 0.00015823249865588003,
+      "loss": 0.9311,
+      "step": 3178
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7014653086662292,
+      "learning_rate": 0.00015820806073212055,
+      "loss": 0.9614,
+      "step": 3179
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7118635177612305,
+      "learning_rate": 0.00015818361754959233,
+      "loss": 1.0199,
+      "step": 3180
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7127907872200012,
+      "learning_rate": 0.00015815916911050365,
+      "loss": 1.1185,
+      "step": 3181
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6535012722015381,
+      "learning_rate": 0.00015813471541706336,
+      "loss": 0.8695,
+      "step": 3182
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.8021343946456909,
+      "learning_rate": 0.00015811025647148073,
+      "loss": 0.8705,
+      "step": 3183
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7726711630821228,
+      "learning_rate": 0.00015808579227596543,
+      "loss": 0.8476,
+      "step": 3184
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7215268015861511,
+      "learning_rate": 0.00015806132283272768,
+      "loss": 0.9388,
+      "step": 3185
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7374808192253113,
+      "learning_rate": 0.00015803684814397813,
+      "loss": 1.0008,
+      "step": 3186
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7936142683029175,
+      "learning_rate": 0.00015801236821192798,
+      "loss": 0.9965,
+      "step": 3187
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6195435523986816,
+      "learning_rate": 0.00015798788303878882,
+      "loss": 0.9651,
+      "step": 3188
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7308238744735718,
+      "learning_rate": 0.00015796339262677273,
+      "loss": 0.8503,
+      "step": 3189
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.672542154788971,
+      "learning_rate": 0.00015793889697809235,
+      "loss": 0.9087,
+      "step": 3190
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6192045211791992,
+      "learning_rate": 0.00015791439609496064,
+      "loss": 0.9253,
+      "step": 3191
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7407450079917908,
+      "learning_rate": 0.00015788988997959114,
+      "loss": 1.1328,
+      "step": 3192
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6472103595733643,
+      "learning_rate": 0.00015786537863419787,
+      "loss": 0.8499,
+      "step": 3193
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.765039324760437,
+      "learning_rate": 0.0001578408620609953,
+      "loss": 0.9891,
+      "step": 3194
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6693131923675537,
+      "learning_rate": 0.00015781634026219835,
+      "loss": 0.8916,
+      "step": 3195
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6833876371383667,
+      "learning_rate": 0.00015779181324002234,
+      "loss": 0.9281,
+      "step": 3196
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7620227336883545,
+      "learning_rate": 0.00015776728099668326,
+      "loss": 0.9676,
+      "step": 3197
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6861891150474548,
+      "learning_rate": 0.00015774274353439743,
+      "loss": 1.0581,
+      "step": 3198
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6912551522254944,
+      "learning_rate": 0.00015771820085538164,
+      "loss": 0.889,
+      "step": 3199
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7697704434394836,
+      "learning_rate": 0.00015769365296185324,
+      "loss": 1.0187,
+      "step": 3200
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7657372355461121,
+      "learning_rate": 0.00015766909985602992,
+      "loss": 1.1127,
+      "step": 3201
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7877035140991211,
+      "learning_rate": 0.00015764454154012993,
+      "loss": 0.9326,
+      "step": 3202
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7275406122207642,
+      "learning_rate": 0.00015761997801637203,
+      "loss": 0.9775,
+      "step": 3203
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6876859068870544,
+      "learning_rate": 0.00015759540928697536,
+      "loss": 0.9551,
+      "step": 3204
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6654580235481262,
+      "learning_rate": 0.00015757083535415957,
+      "loss": 0.8806,
+      "step": 3205
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7412167191505432,
+      "learning_rate": 0.00015754625622014474,
+      "loss": 0.9021,
+      "step": 3206
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7922194004058838,
+      "learning_rate": 0.0001575216718871515,
+      "loss": 0.9801,
+      "step": 3207
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6838287115097046,
+      "learning_rate": 0.00015749708235740091,
+      "loss": 1.0322,
+      "step": 3208
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8312490582466125,
+      "learning_rate": 0.00015747248763311444,
+      "loss": 0.928,
+      "step": 3209
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7289003729820251,
+      "learning_rate": 0.00015744788771651416,
+      "loss": 0.9353,
+      "step": 3210
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7136335968971252,
+      "learning_rate": 0.00015742328260982247,
+      "loss": 1.0114,
+      "step": 3211
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6795276999473572,
+      "learning_rate": 0.00015739867231526233,
+      "loss": 0.9974,
+      "step": 3212
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6614225506782532,
+      "learning_rate": 0.00015737405683505714,
+      "loss": 0.8994,
+      "step": 3213
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.748741865158081,
+      "learning_rate": 0.00015734943617143077,
+      "loss": 0.9645,
+      "step": 3214
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.657394289970398,
+      "learning_rate": 0.00015732481032660755,
+      "loss": 0.9442,
+      "step": 3215
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7445186376571655,
+      "learning_rate": 0.0001573001793028123,
+      "loss": 0.9267,
+      "step": 3216
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7444228529930115,
+      "learning_rate": 0.00015727554310227023,
+      "loss": 0.7558,
+      "step": 3217
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7758005261421204,
+      "learning_rate": 0.0001572509017272072,
+      "loss": 0.8827,
+      "step": 3218
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7533059120178223,
+      "learning_rate": 0.00015722625517984934,
+      "loss": 0.8436,
+      "step": 3219
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7460200190544128,
+      "learning_rate": 0.00015720160346242333,
+      "loss": 0.9243,
+      "step": 3220
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6741672158241272,
+      "learning_rate": 0.00015717694657715629,
+      "loss": 0.9178,
+      "step": 3221
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.744194746017456,
+      "learning_rate": 0.0001571522845262759,
+      "loss": 0.935,
+      "step": 3222
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6983630657196045,
+      "learning_rate": 0.0001571276173120102,
+      "loss": 0.8223,
+      "step": 3223
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7060294151306152,
+      "learning_rate": 0.00015710294493658772,
+      "loss": 0.972,
+      "step": 3224
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7305655479431152,
+      "learning_rate": 0.00015707826740223748,
+      "loss": 0.7972,
+      "step": 3225
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7339190244674683,
+      "learning_rate": 0.00015705358471118896,
+      "loss": 1.0512,
+      "step": 3226
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8077118992805481,
+      "learning_rate": 0.0001570288968656721,
+      "loss": 0.8467,
+      "step": 3227
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8146405816078186,
+      "learning_rate": 0.0001570042038679173,
+      "loss": 0.9306,
+      "step": 3228
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.628921627998352,
+      "learning_rate": 0.00015697950572015544,
+      "loss": 0.8804,
+      "step": 3229
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6738828420639038,
+      "learning_rate": 0.00015695480242461783,
+      "loss": 0.6829,
+      "step": 3230
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.720912754535675,
+      "learning_rate": 0.00015693009398353634,
+      "loss": 1.0782,
+      "step": 3231
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7239159941673279,
+      "learning_rate": 0.00015690538039914316,
+      "loss": 0.9751,
+      "step": 3232
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6557945609092712,
+      "learning_rate": 0.00015688066167367107,
+      "loss": 0.9095,
+      "step": 3233
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7212446331977844,
+      "learning_rate": 0.00015685593780935325,
+      "loss": 1.041,
+      "step": 3234
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7091822624206543,
+      "learning_rate": 0.00015683120880842342,
+      "loss": 1.034,
+      "step": 3235
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.755889356136322,
+      "learning_rate": 0.00015680647467311557,
+      "loss": 1.0469,
+      "step": 3236
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7285449504852295,
+      "learning_rate": 0.00015678173540566444,
+      "loss": 0.9419,
+      "step": 3237
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7696725726127625,
+      "learning_rate": 0.00015675699100830496,
+      "loss": 1.1016,
+      "step": 3238
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6976828575134277,
+      "learning_rate": 0.00015673224148327275,
+      "loss": 0.9678,
+      "step": 3239
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6844419836997986,
+      "learning_rate": 0.0001567074868328037,
+      "loss": 0.893,
+      "step": 3240
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7798510193824768,
+      "learning_rate": 0.00015668272705913428,
+      "loss": 0.953,
+      "step": 3241
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6004636287689209,
+      "learning_rate": 0.00015665796216450146,
+      "loss": 0.7275,
+      "step": 3242
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.8678481578826904,
+      "learning_rate": 0.0001566331921511425,
+      "loss": 1.0891,
+      "step": 3243
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7131955027580261,
+      "learning_rate": 0.00015660841702129532,
+      "loss": 1.0129,
+      "step": 3244
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7723953723907471,
+      "learning_rate": 0.00015658363677719818,
+      "loss": 1.0193,
+      "step": 3245
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7669752836227417,
+      "learning_rate": 0.0001565588514210898,
+      "loss": 0.8838,
+      "step": 3246
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7441872954368591,
+      "learning_rate": 0.00015653406095520947,
+      "loss": 1.0323,
+      "step": 3247
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7313008308410645,
+      "learning_rate": 0.00015650926538179682,
+      "loss": 1.075,
+      "step": 3248
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.8604608178138733,
+      "learning_rate": 0.00015648446470309198,
+      "loss": 0.8467,
+      "step": 3249
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7757964134216309,
+      "learning_rate": 0.00015645965892133559,
+      "loss": 0.984,
+      "step": 3250
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6894923448562622,
+      "learning_rate": 0.00015643484803876866,
+      "loss": 0.7665,
+      "step": 3251
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7314807176589966,
+      "learning_rate": 0.00015641003205763275,
+      "loss": 1.0821,
+      "step": 3252
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6835300922393799,
+      "learning_rate": 0.00015638521098016982,
+      "loss": 0.7995,
+      "step": 3253
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7496686577796936,
+      "learning_rate": 0.00015636038480862239,
+      "loss": 0.9026,
+      "step": 3254
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7328741550445557,
+      "learning_rate": 0.0001563355535452333,
+      "loss": 0.8307,
+      "step": 3255
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.8298733234405518,
+      "learning_rate": 0.00015631071719224583,
+      "loss": 0.9103,
+      "step": 3256
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.8060025572776794,
+      "learning_rate": 0.00015628587575190395,
+      "loss": 1.0579,
+      "step": 3257
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7077762484550476,
+      "learning_rate": 0.00015626102922645192,
+      "loss": 0.8934,
+      "step": 3258
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7120048403739929,
+      "learning_rate": 0.0001562361776181344,
+      "loss": 0.9242,
+      "step": 3259
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.8317088484764099,
+      "learning_rate": 0.00015621132092919667,
+      "loss": 0.94,
+      "step": 3260
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7645663022994995,
+      "learning_rate": 0.00015618645916188432,
+      "loss": 0.9364,
+      "step": 3261
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7070802450180054,
+      "learning_rate": 0.00015616159231844354,
+      "loss": 0.84,
+      "step": 3262
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7429453134536743,
+      "learning_rate": 0.00015613672040112086,
+      "loss": 0.8429,
+      "step": 3263
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7282542586326599,
+      "learning_rate": 0.0001561118434121634,
+      "loss": 0.8897,
+      "step": 3264
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7371309995651245,
+      "learning_rate": 0.00015608696135381851,
+      "loss": 0.9725,
+      "step": 3265
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6809751391410828,
+      "learning_rate": 0.00015606207422833425,
+      "loss": 0.9137,
+      "step": 3266
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6739715933799744,
+      "learning_rate": 0.00015603718203795902,
+      "loss": 1.0297,
+      "step": 3267
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7235797047615051,
+      "learning_rate": 0.00015601228478494166,
+      "loss": 0.8851,
+      "step": 3268
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7177626490592957,
+      "learning_rate": 0.00015598738247153156,
+      "loss": 0.8884,
+      "step": 3269
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7195912599563599,
+      "learning_rate": 0.00015596247509997842,
+      "loss": 0.992,
+      "step": 3270
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6846754550933838,
+      "learning_rate": 0.0001559375626725325,
+      "loss": 1.073,
+      "step": 3271
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.8329934477806091,
+      "learning_rate": 0.0001559126451914445,
+      "loss": 1.0911,
+      "step": 3272
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7074479460716248,
+      "learning_rate": 0.0001558877226589656,
+      "loss": 0.991,
+      "step": 3273
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7118126749992371,
+      "learning_rate": 0.00015586279507734743,
+      "loss": 1.0068,
+      "step": 3274
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7480494379997253,
+      "learning_rate": 0.00015583786244884198,
+      "loss": 1.0171,
+      "step": 3275
+    },
+    {
+      "epoch": 1.23,
+      "eval_loss": 0.9602957367897034,
+      "eval_runtime": 17.812,
+      "eval_samples_per_second": 153.211,
+      "eval_steps_per_second": 38.345,
+      "step": 3275
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6866698265075684,
+      "learning_rate": 0.0001558129247757018,
+      "loss": 1.0152,
+      "step": 3276
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7157483696937561,
+      "learning_rate": 0.00015578798206017987,
+      "loss": 0.9652,
+      "step": 3277
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.8190251588821411,
+      "learning_rate": 0.00015576303430452968,
+      "loss": 1.0241,
+      "step": 3278
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7425306439399719,
+      "learning_rate": 0.00015573808151100505,
+      "loss": 1.0053,
+      "step": 3279
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6862152218818665,
+      "learning_rate": 0.00015571312368186036,
+      "loss": 0.9581,
+      "step": 3280
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7101152539253235,
+      "learning_rate": 0.00015568816081935034,
+      "loss": 0.8594,
+      "step": 3281
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6727707386016846,
+      "learning_rate": 0.00015566319292573035,
+      "loss": 0.9591,
+      "step": 3282
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6215928792953491,
+      "learning_rate": 0.000155638220003256,
+      "loss": 0.8791,
+      "step": 3283
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.8467756509780884,
+      "learning_rate": 0.00015561324205418353,
+      "loss": 1.0264,
+      "step": 3284
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7936529517173767,
+      "learning_rate": 0.0001555882590807695,
+      "loss": 0.8712,
+      "step": 3285
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7405680418014526,
+      "learning_rate": 0.000155563271085271,
+      "loss": 0.8966,
+      "step": 3286
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7553184032440186,
+      "learning_rate": 0.00015553827806994556,
+      "loss": 0.9128,
+      "step": 3287
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7074669599533081,
+      "learning_rate": 0.00015551328003705117,
+      "loss": 0.7814,
+      "step": 3288
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7011464834213257,
+      "learning_rate": 0.00015548827698884626,
+      "loss": 1.133,
+      "step": 3289
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.71796715259552,
+      "learning_rate": 0.00015546326892758964,
+      "loss": 0.919,
+      "step": 3290
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6808836460113525,
+      "learning_rate": 0.00015543825585554073,
+      "loss": 0.835,
+      "step": 3291
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7186148166656494,
+      "learning_rate": 0.0001554132377749593,
+      "loss": 0.8625,
+      "step": 3292
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7094897031784058,
+      "learning_rate": 0.0001553882146881056,
+      "loss": 0.8783,
+      "step": 3293
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7546431422233582,
+      "learning_rate": 0.0001553631865972403,
+      "loss": 0.9686,
+      "step": 3294
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6863462924957275,
+      "learning_rate": 0.00015533815350462455,
+      "loss": 1.0179,
+      "step": 3295
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6790794134140015,
+      "learning_rate": 0.00015531311541251995,
+      "loss": 0.9417,
+      "step": 3296
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7436459064483643,
+      "learning_rate": 0.00015528807232318857,
+      "loss": 0.9276,
+      "step": 3297
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.732758641242981,
+      "learning_rate": 0.00015526302423889287,
+      "loss": 1.0102,
+      "step": 3298
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7761924266815186,
+      "learning_rate": 0.00015523797116189587,
+      "loss": 1.1556,
+      "step": 3299
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7570998668670654,
+      "learning_rate": 0.00015521291309446089,
+      "loss": 0.8907,
+      "step": 3300
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7217000722885132,
+      "learning_rate": 0.00015518785003885185,
+      "loss": 0.9827,
+      "step": 3301
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.9155686497688293,
+      "learning_rate": 0.00015516278199733302,
+      "loss": 0.8761,
+      "step": 3302
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.8354843854904175,
+      "learning_rate": 0.00015513770897216918,
+      "loss": 1.1249,
+      "step": 3303
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7380690574645996,
+      "learning_rate": 0.0001551126309656255,
+      "loss": 0.8494,
+      "step": 3304
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.73750239610672,
+      "learning_rate": 0.0001550875479799677,
+      "loss": 1.0285,
+      "step": 3305
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.793064296245575,
+      "learning_rate": 0.0001550624600174618,
+      "loss": 0.8748,
+      "step": 3306
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7800728678703308,
+      "learning_rate": 0.0001550373670803744,
+      "loss": 0.9082,
+      "step": 3307
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6863796710968018,
+      "learning_rate": 0.00015501226917097257,
+      "loss": 0.9241,
+      "step": 3308
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.8050414323806763,
+      "learning_rate": 0.00015498716629152367,
+      "loss": 0.9243,
+      "step": 3309
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6615169048309326,
+      "learning_rate": 0.0001549620584442956,
+      "loss": 0.8398,
+      "step": 3310
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7062418460845947,
+      "learning_rate": 0.00015493694563155675,
+      "loss": 1.0246,
+      "step": 3311
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6960569024085999,
+      "learning_rate": 0.00015491182785557593,
+      "loss": 0.9849,
+      "step": 3312
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7825234532356262,
+      "learning_rate": 0.00015488670511862237,
+      "loss": 1.0409,
+      "step": 3313
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7394260764122009,
+      "learning_rate": 0.00015486157742296576,
+      "loss": 0.8847,
+      "step": 3314
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7210456728935242,
+      "learning_rate": 0.00015483644477087626,
+      "loss": 1.0684,
+      "step": 3315
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.786108672618866,
+      "learning_rate": 0.00015481130716462447,
+      "loss": 0.7895,
+      "step": 3316
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.639926552772522,
+      "learning_rate": 0.00015478616460648143,
+      "loss": 0.86,
+      "step": 3317
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7433844208717346,
+      "learning_rate": 0.0001547610170987186,
+      "loss": 0.7851,
+      "step": 3318
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7296815514564514,
+      "learning_rate": 0.0001547358646436079,
+      "loss": 0.9552,
+      "step": 3319
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6938995122909546,
+      "learning_rate": 0.00015471070724342177,
+      "loss": 1.0096,
+      "step": 3320
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8119737505912781,
+      "learning_rate": 0.00015468554490043302,
+      "loss": 1.0975,
+      "step": 3321
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6842712759971619,
+      "learning_rate": 0.0001546603776169149,
+      "loss": 1.0019,
+      "step": 3322
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7045173048973083,
+      "learning_rate": 0.00015463520539514116,
+      "loss": 0.9034,
+      "step": 3323
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6672970652580261,
+      "learning_rate": 0.00015461002823738597,
+      "loss": 0.8547,
+      "step": 3324
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7053340673446655,
+      "learning_rate": 0.00015458484614592393,
+      "loss": 0.9211,
+      "step": 3325
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6476091146469116,
+      "learning_rate": 0.00015455965912303005,
+      "loss": 0.8302,
+      "step": 3326
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7886658310890198,
+      "learning_rate": 0.00015453446717097994,
+      "loss": 1.0586,
+      "step": 3327
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6936331987380981,
+      "learning_rate": 0.00015450927029204948,
+      "loss": 0.9953,
+      "step": 3328
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8038539290428162,
+      "learning_rate": 0.00015448406848851508,
+      "loss": 1.0148,
+      "step": 3329
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7686813473701477,
+      "learning_rate": 0.00015445886176265354,
+      "loss": 0.9441,
+      "step": 3330
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7950506210327148,
+      "learning_rate": 0.00015443365011674222,
+      "loss": 1.0075,
+      "step": 3331
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7895020246505737,
+      "learning_rate": 0.0001544084335530588,
+      "loss": 0.9223,
+      "step": 3332
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7187060713768005,
+      "learning_rate": 0.0001543832120738815,
+      "loss": 0.9758,
+      "step": 3333
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.771331787109375,
+      "learning_rate": 0.0001543579856814889,
+      "loss": 1.0886,
+      "step": 3334
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7543261051177979,
+      "learning_rate": 0.00015433275437816004,
+      "loss": 0.9031,
+      "step": 3335
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7562270164489746,
+      "learning_rate": 0.00015430751816617445,
+      "loss": 1.0196,
+      "step": 3336
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.688003659248352,
+      "learning_rate": 0.0001542822770478121,
+      "loss": 1.0277,
+      "step": 3337
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7344198226928711,
+      "learning_rate": 0.00015425703102535339,
+      "loss": 1.0609,
+      "step": 3338
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.8227985501289368,
+      "learning_rate": 0.0001542317801010791,
+      "loss": 0.9561,
+      "step": 3339
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6553336977958679,
+      "learning_rate": 0.00015420652427727054,
+      "loss": 0.9462,
+      "step": 3340
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7725002765655518,
+      "learning_rate": 0.00015418126355620944,
+      "loss": 0.8813,
+      "step": 3341
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.687086284160614,
+      "learning_rate": 0.00015415599794017793,
+      "loss": 0.9903,
+      "step": 3342
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.735305905342102,
+      "learning_rate": 0.00015413072743145865,
+      "loss": 1.0012,
+      "step": 3343
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.800947368144989,
+      "learning_rate": 0.0001541054520323346,
+      "loss": 0.7199,
+      "step": 3344
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7069570422172546,
+      "learning_rate": 0.00015408017174508936,
+      "loss": 0.9401,
+      "step": 3345
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7075111865997314,
+      "learning_rate": 0.00015405488657200674,
+      "loss": 0.8181,
+      "step": 3346
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7980069518089294,
+      "learning_rate": 0.00015402959651537125,
+      "loss": 1.0082,
+      "step": 3347
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7631432414054871,
+      "learning_rate": 0.00015400430157746757,
+      "loss": 0.9643,
+      "step": 3348
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.8499820828437805,
+      "learning_rate": 0.00015397900176058107,
+      "loss": 0.967,
+      "step": 3349
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7591486573219299,
+      "learning_rate": 0.00015395369706699733,
+      "loss": 0.981,
+      "step": 3350
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7172725200653076,
+      "learning_rate": 0.00015392838749900257,
+      "loss": 1.0516,
+      "step": 3351
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7486289739608765,
+      "learning_rate": 0.00015390307305888335,
+      "loss": 0.9728,
+      "step": 3352
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7924482822418213,
+      "learning_rate": 0.00015387775374892665,
+      "loss": 0.7647,
+      "step": 3353
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.9388603568077087,
+      "learning_rate": 0.00015385242957142,
+      "loss": 1.1205,
+      "step": 3354
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6928844451904297,
+      "learning_rate": 0.0001538271005286512,
+      "loss": 0.8248,
+      "step": 3355
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7360419034957886,
+      "learning_rate": 0.00015380176662290867,
+      "loss": 1.0013,
+      "step": 3356
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.706224262714386,
+      "learning_rate": 0.00015377642785648115,
+      "loss": 1.0004,
+      "step": 3357
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7525461316108704,
+      "learning_rate": 0.00015375108423165785,
+      "loss": 1.0026,
+      "step": 3358
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.732796847820282,
+      "learning_rate": 0.00015372573575072842,
+      "loss": 0.9005,
+      "step": 3359
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6761058568954468,
+      "learning_rate": 0.00015370038241598294,
+      "loss": 0.9114,
+      "step": 3360
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7725770473480225,
+      "learning_rate": 0.000153675024229712,
+      "loss": 0.9592,
+      "step": 3361
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7360583543777466,
+      "learning_rate": 0.0001536496611942065,
+      "loss": 1.0003,
+      "step": 3362
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7047659754753113,
+      "learning_rate": 0.0001536242933117579,
+      "loss": 0.8064,
+      "step": 3363
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8691027760505676,
+      "learning_rate": 0.000153598920584658,
+      "loss": 1.0726,
+      "step": 3364
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7493014931678772,
+      "learning_rate": 0.00015357354301519908,
+      "loss": 0.9652,
+      "step": 3365
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6685104370117188,
+      "learning_rate": 0.00015354816060567393,
+      "loss": 0.9319,
+      "step": 3366
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7175832390785217,
+      "learning_rate": 0.00015352277335837561,
+      "loss": 1.0431,
+      "step": 3367
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7429420948028564,
+      "learning_rate": 0.00015349738127559782,
+      "loss": 0.8698,
+      "step": 3368
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7226133346557617,
+      "learning_rate": 0.0001534719843596345,
+      "loss": 1.0228,
+      "step": 3369
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8891973495483398,
+      "learning_rate": 0.0001534465826127801,
+      "loss": 0.9278,
+      "step": 3370
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7273642420768738,
+      "learning_rate": 0.00015342117603732965,
+      "loss": 0.9188,
+      "step": 3371
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7487305402755737,
+      "learning_rate": 0.00015339576463557837,
+      "loss": 0.9189,
+      "step": 3372
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7615191340446472,
+      "learning_rate": 0.00015337034840982213,
+      "loss": 1.0485,
+      "step": 3373
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8446441292762756,
+      "learning_rate": 0.00015334492736235705,
+      "loss": 1.0208,
+      "step": 3374
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7321614623069763,
+      "learning_rate": 0.0001533195014954798,
+      "loss": 0.9412,
+      "step": 3375
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7862521409988403,
+      "learning_rate": 0.00015329407081148753,
+      "loss": 0.8936,
+      "step": 3376
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8918995261192322,
+      "learning_rate": 0.00015326863531267773,
+      "loss": 0.978,
+      "step": 3377
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6822420358657837,
+      "learning_rate": 0.0001532431950013483,
+      "loss": 0.9463,
+      "step": 3378
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7396413087844849,
+      "learning_rate": 0.00015321774987979767,
+      "loss": 0.9282,
+      "step": 3379
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7108849883079529,
+      "learning_rate": 0.00015319229995032464,
+      "loss": 0.8528,
+      "step": 3380
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.813988983631134,
+      "learning_rate": 0.0001531668452152285,
+      "loss": 0.9853,
+      "step": 3381
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.717553973197937,
+      "learning_rate": 0.00015314138567680896,
+      "loss": 0.8755,
+      "step": 3382
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6788370609283447,
+      "learning_rate": 0.00015311592133736605,
+      "loss": 0.835,
+      "step": 3383
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7784591913223267,
+      "learning_rate": 0.0001530904521992004,
+      "loss": 1.0843,
+      "step": 3384
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.9008314609527588,
+      "learning_rate": 0.00015306497826461302,
+      "loss": 0.9313,
+      "step": 3385
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8545159101486206,
+      "learning_rate": 0.00015303949953590529,
+      "loss": 0.9582,
+      "step": 3386
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8266891241073608,
+      "learning_rate": 0.0001530140160153791,
+      "loss": 0.9513,
+      "step": 3387
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7323997616767883,
+      "learning_rate": 0.00015298852770533668,
+      "loss": 0.9853,
+      "step": 3388
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7101405262947083,
+      "learning_rate": 0.00015296303460808085,
+      "loss": 0.9666,
+      "step": 3389
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6381584405899048,
+      "learning_rate": 0.0001529375367259147,
+      "loss": 0.8627,
+      "step": 3390
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8005196452140808,
+      "learning_rate": 0.00015291203406114185,
+      "loss": 1.1132,
+      "step": 3391
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7641215324401855,
+      "learning_rate": 0.00015288652661606632,
+      "loss": 1.067,
+      "step": 3392
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7402655482292175,
+      "learning_rate": 0.0001528610143929925,
+      "loss": 1.0194,
+      "step": 3393
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.088860273361206,
+      "learning_rate": 0.0001528354973942254,
+      "loss": 0.8712,
+      "step": 3394
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7720997333526611,
+      "learning_rate": 0.00015280997562207023,
+      "loss": 0.9561,
+      "step": 3395
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7261041402816772,
+      "learning_rate": 0.00015278444907883273,
+      "loss": 0.8256,
+      "step": 3396
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6564508676528931,
+      "learning_rate": 0.00015275891776681918,
+      "loss": 0.9009,
+      "step": 3397
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7378062009811401,
+      "learning_rate": 0.00015273338168833613,
+      "loss": 1.0495,
+      "step": 3398
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8059569597244263,
+      "learning_rate": 0.0001527078408456906,
+      "loss": 0.8445,
+      "step": 3399
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.649719774723053,
+      "learning_rate": 0.00015268229524119007,
+      "loss": 0.8538,
+      "step": 3400
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7831336259841919,
+      "learning_rate": 0.00015265674487714245,
+      "loss": 0.9864,
+      "step": 3401
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7160272002220154,
+      "learning_rate": 0.0001526311897558561,
+      "loss": 0.9808,
+      "step": 3402
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7628503441810608,
+      "learning_rate": 0.00015260562987963976,
+      "loss": 0.9573,
+      "step": 3403
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.760503888130188,
+      "learning_rate": 0.00015258006525080257,
+      "loss": 0.9998,
+      "step": 3404
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6774016618728638,
+      "learning_rate": 0.00015255449587165423,
+      "loss": 0.9838,
+      "step": 3405
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7273324728012085,
+      "learning_rate": 0.00015252892174450473,
+      "loss": 0.9831,
+      "step": 3406
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7090844511985779,
+      "learning_rate": 0.00015250334287166453,
+      "loss": 1.0137,
+      "step": 3407
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7482985258102417,
+      "learning_rate": 0.00015247775925544466,
+      "loss": 0.8267,
+      "step": 3408
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6855625510215759,
+      "learning_rate": 0.0001524521708981563,
+      "loss": 0.8884,
+      "step": 3409
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.720038890838623,
+      "learning_rate": 0.00015242657780211133,
+      "loss": 1.009,
+      "step": 3410
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8395721912384033,
+      "learning_rate": 0.00015240097996962185,
+      "loss": 0.8667,
+      "step": 3411
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7716408967971802,
+      "learning_rate": 0.00015237537740300056,
+      "loss": 0.9672,
+      "step": 3412
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6766606569290161,
+      "learning_rate": 0.00015234977010456047,
+      "loss": 0.8397,
+      "step": 3413
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8159101605415344,
+      "learning_rate": 0.00015232415807661504,
+      "loss": 0.9371,
+      "step": 3414
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.797480046749115,
+      "learning_rate": 0.00015229854132147816,
+      "loss": 1.0038,
+      "step": 3415
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7471280097961426,
+      "learning_rate": 0.0001522729198414642,
+      "loss": 1.0894,
+      "step": 3416
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7765840291976929,
+      "learning_rate": 0.00015224729363888794,
+      "loss": 0.9695,
+      "step": 3417
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.674380362033844,
+      "learning_rate": 0.0001522216627160645,
+      "loss": 1.0249,
+      "step": 3418
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7781017422676086,
+      "learning_rate": 0.0001521960270753095,
+      "loss": 1.0102,
+      "step": 3419
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8835538625717163,
+      "learning_rate": 0.00015217038671893896,
+      "loss": 1.1093,
+      "step": 3420
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7141196131706238,
+      "learning_rate": 0.00015214474164926945,
+      "loss": 0.894,
+      "step": 3421
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7608664631843567,
+      "learning_rate": 0.00015211909186861773,
+      "loss": 0.9484,
+      "step": 3422
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7789210677146912,
+      "learning_rate": 0.00015209343737930116,
+      "loss": 1.1107,
+      "step": 3423
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.701990008354187,
+      "learning_rate": 0.00015206777818363746,
+      "loss": 0.7809,
+      "step": 3424
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8689419031143188,
+      "learning_rate": 0.00015204211428394482,
+      "loss": 1.1252,
+      "step": 3425
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6336739659309387,
+      "learning_rate": 0.0001520164456825418,
+      "loss": 0.8921,
+      "step": 3426
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6421981453895569,
+      "learning_rate": 0.00015199077238174747,
+      "loss": 0.8899,
+      "step": 3427
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8911098837852478,
+      "learning_rate": 0.0001519650943838812,
+      "loss": 0.9314,
+      "step": 3428
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8368533253669739,
+      "learning_rate": 0.00015193941169126286,
+      "loss": 0.9066,
+      "step": 3429
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6978997588157654,
+      "learning_rate": 0.00015191372430621277,
+      "loss": 0.9714,
+      "step": 3430
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7743934988975525,
+      "learning_rate": 0.0001518880322310516,
+      "loss": 1.0117,
+      "step": 3431
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7560721635818481,
+      "learning_rate": 0.00015186233546810057,
+      "loss": 0.9588,
+      "step": 3432
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6989509463310242,
+      "learning_rate": 0.0001518366340196811,
+      "loss": 0.9755,
+      "step": 3433
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7402550578117371,
+      "learning_rate": 0.00015181092788811533,
+      "loss": 0.9117,
+      "step": 3434
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8155473470687866,
+      "learning_rate": 0.00015178521707572552,
+      "loss": 0.9887,
+      "step": 3435
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7658118009567261,
+      "learning_rate": 0.00015175950158483455,
+      "loss": 0.9122,
+      "step": 3436
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7669534087181091,
+      "learning_rate": 0.00015173378141776568,
+      "loss": 0.9987,
+      "step": 3437
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8077985048294067,
+      "learning_rate": 0.0001517080565768426,
+      "loss": 0.8834,
+      "step": 3438
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.717319667339325,
+      "learning_rate": 0.00015168232706438937,
+      "loss": 0.9681,
+      "step": 3439
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.802055299282074,
+      "learning_rate": 0.0001516565928827305,
+      "loss": 1.1436,
+      "step": 3440
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7889139652252197,
+      "learning_rate": 0.00015163085403419095,
+      "loss": 0.9313,
+      "step": 3441
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7818065285682678,
+      "learning_rate": 0.00015160511052109604,
+      "loss": 1.1377,
+      "step": 3442
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7210289239883423,
+      "learning_rate": 0.00015157936234577163,
+      "loss": 0.8287,
+      "step": 3443
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7757744193077087,
+      "learning_rate": 0.00015155360951054383,
+      "loss": 0.8272,
+      "step": 3444
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7870873212814331,
+      "learning_rate": 0.0001515278520177393,
+      "loss": 0.902,
+      "step": 3445
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7317755818367004,
+      "learning_rate": 0.00015150208986968515,
+      "loss": 1.0121,
+      "step": 3446
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7527960538864136,
+      "learning_rate": 0.00015147632306870875,
+      "loss": 0.9836,
+      "step": 3447
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7288510203361511,
+      "learning_rate": 0.00015145055161713803,
+      "loss": 0.9623,
+      "step": 3448
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7237223386764526,
+      "learning_rate": 0.00015142477551730128,
+      "loss": 1.0682,
+      "step": 3449
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6942408680915833,
+      "learning_rate": 0.0001513989947715272,
+      "loss": 0.8924,
+      "step": 3450
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7724366188049316,
+      "learning_rate": 0.000151373209382145,
+      "loss": 1.0604,
+      "step": 3451
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7728193402290344,
+      "learning_rate": 0.0001513474193514842,
+      "loss": 0.9177,
+      "step": 3452
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.8020523190498352,
+      "learning_rate": 0.0001513216246818748,
+      "loss": 1.1552,
+      "step": 3453
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7387242913246155,
+      "learning_rate": 0.00015129582537564716,
+      "loss": 0.9259,
+      "step": 3454
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7924624681472778,
+      "learning_rate": 0.00015127002143513213,
+      "loss": 0.9954,
+      "step": 3455
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7585887312889099,
+      "learning_rate": 0.000151244212862661,
+      "loss": 0.7944,
+      "step": 3456
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7654167413711548,
+      "learning_rate": 0.00015121839966056536,
+      "loss": 1.0601,
+      "step": 3457
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.8087634444236755,
+      "learning_rate": 0.00015119258183117728,
+      "loss": 0.8418,
+      "step": 3458
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7145108580589294,
+      "learning_rate": 0.00015116675937682934,
+      "loss": 0.9512,
+      "step": 3459
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.755843997001648,
+      "learning_rate": 0.00015114093229985436,
+      "loss": 0.9324,
+      "step": 3460
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.744723379611969,
+      "learning_rate": 0.0001511151006025857,
+      "loss": 1.0098,
+      "step": 3461
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": NaN,
+      "learning_rate": 0.0001511151006025857,
+      "loss": 0.9791,
+      "step": 3462
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7100603580474854,
+      "learning_rate": 0.0001510892642873571,
+      "loss": 1.0333,
+      "step": 3463
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7299081683158875,
+      "learning_rate": 0.0001510634233565028,
+      "loss": 0.8616,
+      "step": 3464
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7373127937316895,
+      "learning_rate": 0.00015103757781235726,
+      "loss": 0.9998,
+      "step": 3465
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7837493419647217,
+      "learning_rate": 0.00015101172765725554,
+      "loss": 1.0791,
+      "step": 3466
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.722700297832489,
+      "learning_rate": 0.00015098587289353308,
+      "loss": 0.856,
+      "step": 3467
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6745336055755615,
+      "learning_rate": 0.0001509600135235257,
+      "loss": 0.8963,
+      "step": 3468
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6792166233062744,
+      "learning_rate": 0.0001509341495495696,
+      "loss": 0.991,
+      "step": 3469
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7579882740974426,
+      "learning_rate": 0.0001509082809740015,
+      "loss": 0.9236,
+      "step": 3470
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6892275214195251,
+      "learning_rate": 0.00015088240779915842,
+      "loss": 0.8991,
+      "step": 3471
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7705352306365967,
+      "learning_rate": 0.0001508565300273779,
+      "loss": 1.0342,
+      "step": 3472
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.811631977558136,
+      "learning_rate": 0.00015083064766099788,
+      "loss": 0.8619,
+      "step": 3473
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8345324397087097,
+      "learning_rate": 0.00015080476070235662,
+      "loss": 1.0333,
+      "step": 3474
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6630562543869019,
+      "learning_rate": 0.0001507788691537929,
+      "loss": 0.9507,
+      "step": 3475
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7931245565414429,
+      "learning_rate": 0.00015075297301764585,
+      "loss": 0.9996,
+      "step": 3476
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7066725492477417,
+      "learning_rate": 0.00015072707229625505,
+      "loss": 0.9775,
+      "step": 3477
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7714331746101379,
+      "learning_rate": 0.00015070116699196045,
+      "loss": 0.972,
+      "step": 3478
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.854289174079895,
+      "learning_rate": 0.00015067525710710252,
+      "loss": 1.0122,
+      "step": 3479
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6653081178665161,
+      "learning_rate": 0.00015064934264402203,
+      "loss": 0.9286,
+      "step": 3480
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.698539137840271,
+      "learning_rate": 0.00015062342360506022,
+      "loss": 0.8626,
+      "step": 3481
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.84158855676651,
+      "learning_rate": 0.0001505974999925587,
+      "loss": 0.8943,
+      "step": 3482
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8350929021835327,
+      "learning_rate": 0.00015057157180885952,
+      "loss": 0.8051,
+      "step": 3483
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8504432439804077,
+      "learning_rate": 0.00015054563905630522,
+      "loss": 1.1219,
+      "step": 3484
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7879734039306641,
+      "learning_rate": 0.00015051970173723858,
+      "loss": 0.8582,
+      "step": 3485
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7625810503959656,
+      "learning_rate": 0.0001504937598540029,
+      "loss": 1.0164,
+      "step": 3486
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7224019765853882,
+      "learning_rate": 0.00015046781340894195,
+      "loss": 1.0259,
+      "step": 3487
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.738791286945343,
+      "learning_rate": 0.0001504418624043998,
+      "loss": 1.0126,
+      "step": 3488
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6853582262992859,
+      "learning_rate": 0.000150415906842721,
+      "loss": 0.965,
+      "step": 3489
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7412823438644409,
+      "learning_rate": 0.00015038994672625047,
+      "loss": 0.878,
+      "step": 3490
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7869895100593567,
+      "learning_rate": 0.0001503639820573335,
+      "loss": 0.7748,
+      "step": 3491
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8981345891952515,
+      "learning_rate": 0.00015033801283831596,
+      "loss": 0.8713,
+      "step": 3492
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.9000715613365173,
+      "learning_rate": 0.00015031203907154395,
+      "loss": 1.0349,
+      "step": 3493
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7573916912078857,
+      "learning_rate": 0.00015028606075936414,
+      "loss": 1.0724,
+      "step": 3494
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6647499799728394,
+      "learning_rate": 0.0001502600779041234,
+      "loss": 0.9733,
+      "step": 3495
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8138927817344666,
+      "learning_rate": 0.00015023409050816922,
+      "loss": 1.1055,
+      "step": 3496
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7094717621803284,
+      "learning_rate": 0.00015020809857384937,
+      "loss": 0.9432,
+      "step": 3497
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7191406488418579,
+      "learning_rate": 0.0001501821021035121,
+      "loss": 1.1021,
+      "step": 3498
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8311622142791748,
+      "learning_rate": 0.00015015610109950602,
+      "loss": 0.9388,
+      "step": 3499
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7128794193267822,
+      "learning_rate": 0.00015013009556418024,
+      "loss": 0.8522,
+      "step": 3500
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7016494274139404,
+      "learning_rate": 0.00015010408549988408,
+      "loss": 0.9804,
+      "step": 3501
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7902352809906006,
+      "learning_rate": 0.00015007807090896752,
+      "loss": 0.99,
+      "step": 3502
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7388083934783936,
+      "learning_rate": 0.00015005205179378082,
+      "loss": 1.0261,
+      "step": 3503
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8480693101882935,
+      "learning_rate": 0.00015002602815667463,
+      "loss": 1.006,
+      "step": 3504
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7673492431640625,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 1.026,
+      "step": 3505
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8567508459091187,
+      "learning_rate": 0.0001499739673261085,
+      "loss": 0.8649,
+      "step": 3506
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8326045274734497,
+      "learning_rate": 0.000149947930137352,
+      "loss": 1.001,
+      "step": 3507
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7344638705253601,
+      "learning_rate": 0.0001499218884360828,
+      "loss": 0.9197,
+      "step": 3508
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.816310703754425,
+      "learning_rate": 0.00014989584222465367,
+      "loss": 1.0347,
+      "step": 3509
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8038501143455505,
+      "learning_rate": 0.0001498697915054177,
+      "loss": 1.0826,
+      "step": 3510
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7537353038787842,
+      "learning_rate": 0.0001498437362807284,
+      "loss": 1.0289,
+      "step": 3511
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7652748823165894,
+      "learning_rate": 0.00014981767655293976,
+      "loss": 0.7611,
+      "step": 3512
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.732870876789093,
+      "learning_rate": 0.00014979161232440613,
+      "loss": 1.0268,
+      "step": 3513
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.725989818572998,
+      "learning_rate": 0.00014976554359748224,
+      "loss": 0.9618,
+      "step": 3514
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8126868009567261,
+      "learning_rate": 0.00014973947037452324,
+      "loss": 0.9914,
+      "step": 3515
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8361872434616089,
+      "learning_rate": 0.0001497133926578847,
+      "loss": 1.1645,
+      "step": 3516
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8904275298118591,
+      "learning_rate": 0.00014968731044992262,
+      "loss": 0.9246,
+      "step": 3517
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7571832537651062,
+      "learning_rate": 0.0001496612237529934,
+      "loss": 0.9037,
+      "step": 3518
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.8950560092926025,
+      "learning_rate": 0.00014963513256945377,
+      "loss": 1.1312,
+      "step": 3519
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7928481101989746,
+      "learning_rate": 0.00014960903690166094,
+      "loss": 0.9981,
+      "step": 3520
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.676289439201355,
+      "learning_rate": 0.0001495829367519725,
+      "loss": 0.9821,
+      "step": 3521
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7975377440452576,
+      "learning_rate": 0.0001495568321227465,
+      "loss": 1.0974,
+      "step": 3522
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7216330766677856,
+      "learning_rate": 0.00014953072301634128,
+      "loss": 1.0396,
+      "step": 3523
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7259735465049744,
+      "learning_rate": 0.00014950460943511572,
+      "loss": 0.958,
+      "step": 3524
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6637389659881592,
+      "learning_rate": 0.00014947849138142894,
+      "loss": 0.8469,
+      "step": 3525
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6828493475914001,
+      "learning_rate": 0.00014945236885764062,
+      "loss": 0.9702,
+      "step": 3526
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.9553461074829102,
+      "learning_rate": 0.0001494262418661108,
+      "loss": 1.12,
+      "step": 3527
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6893486976623535,
+      "learning_rate": 0.00014940011040919985,
+      "loss": 0.8297,
+      "step": 3528
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6896955370903015,
+      "learning_rate": 0.00014937397448926864,
+      "loss": 0.8091,
+      "step": 3529
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.769710898399353,
+      "learning_rate": 0.0001493478341086784,
+      "loss": 0.9944,
+      "step": 3530
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6986170411109924,
+      "learning_rate": 0.00014932168926979074,
+      "loss": 0.8848,
+      "step": 3531
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7402127981185913,
+      "learning_rate": 0.00014929553997496773,
+      "loss": 0.9583,
+      "step": 3532
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.8498134016990662,
+      "learning_rate": 0.00014926938622657177,
+      "loss": 0.9529,
+      "step": 3533
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7263255715370178,
+      "learning_rate": 0.00014924322802696574,
+      "loss": 0.966,
+      "step": 3534
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6980369091033936,
+      "learning_rate": 0.00014921706537851287,
+      "loss": 0.9146,
+      "step": 3535
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7615999579429626,
+      "learning_rate": 0.0001491908982835768,
+      "loss": 0.9918,
+      "step": 3536
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7459874153137207,
+      "learning_rate": 0.0001491647267445216,
+      "loss": 0.8988,
+      "step": 3537
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.8014723062515259,
+      "learning_rate": 0.0001491385507637117,
+      "loss": 1.0638,
+      "step": 3538
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7640689611434937,
+      "learning_rate": 0.00014911237034351195,
+      "loss": 1.0857,
+      "step": 3539
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7542036771774292,
+      "learning_rate": 0.00014908618548628767,
+      "loss": 0.8638,
+      "step": 3540
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.8518773317337036,
+      "learning_rate": 0.00014905999619440442,
+      "loss": 0.9615,
+      "step": 3541
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7619187831878662,
+      "learning_rate": 0.00014903380247022826,
+      "loss": 0.9976,
+      "step": 3542
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.8794530034065247,
+      "learning_rate": 0.0001490076043161257,
+      "loss": 0.7971,
+      "step": 3543
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.8578647375106812,
+      "learning_rate": 0.00014898140173446356,
+      "loss": 1.0534,
+      "step": 3544
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7493592500686646,
+      "learning_rate": 0.00014895519472760913,
+      "loss": 1.0227,
+      "step": 3545
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.8536798357963562,
+      "learning_rate": 0.00014892898329793003,
+      "loss": 0.9422,
+      "step": 3546
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7785354256629944,
+      "learning_rate": 0.0001489027674477943,
+      "loss": 1.0192,
+      "step": 3547
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7300075888633728,
+      "learning_rate": 0.00014887654717957044,
+      "loss": 0.9542,
+      "step": 3548
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7428839802742004,
+      "learning_rate": 0.0001488503224956273,
+      "loss": 1.0701,
+      "step": 3549
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7662368416786194,
+      "learning_rate": 0.0001488240933983341,
+      "loss": 1.0985,
+      "step": 3550
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7324904799461365,
+      "learning_rate": 0.0001487978598900605,
+      "loss": 0.8729,
+      "step": 3551
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7264313697814941,
+      "learning_rate": 0.00014877162197317654,
+      "loss": 0.9747,
+      "step": 3552
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7613300085067749,
+      "learning_rate": 0.00014874537965005267,
+      "loss": 1.1532,
+      "step": 3553
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7279550433158875,
+      "learning_rate": 0.00014871913292305978,
+      "loss": 1.0743,
+      "step": 3554
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7823167443275452,
+      "learning_rate": 0.00014869288179456905,
+      "loss": 0.9377,
+      "step": 3555
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.854567289352417,
+      "learning_rate": 0.00014866662626695213,
+      "loss": 0.9287,
+      "step": 3556
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7282840609550476,
+      "learning_rate": 0.00014864036634258112,
+      "loss": 0.9809,
+      "step": 3557
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6796576976776123,
+      "learning_rate": 0.0001486141020238284,
+      "loss": 0.803,
+      "step": 3558
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7584293484687805,
+      "learning_rate": 0.00014858783331306682,
+      "loss": 1.0061,
+      "step": 3559
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7804273366928101,
+      "learning_rate": 0.00014856156021266956,
+      "loss": 1.0346,
+      "step": 3560
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7761217951774597,
+      "learning_rate": 0.00014853528272501033,
+      "loss": 0.8941,
+      "step": 3561
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7115616798400879,
+      "learning_rate": 0.0001485090008524631,
+      "loss": 0.9078,
+      "step": 3562
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.8151125311851501,
+      "learning_rate": 0.0001484827145974023,
+      "loss": 1.135,
+      "step": 3563
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7691593170166016,
+      "learning_rate": 0.00014845642396220276,
+      "loss": 0.995,
+      "step": 3564
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7243972420692444,
+      "learning_rate": 0.00014843012894923967,
+      "loss": 0.9317,
+      "step": 3565
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.714561402797699,
+      "learning_rate": 0.00014840382956088862,
+      "loss": 1.0863,
+      "step": 3566
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7968424558639526,
+      "learning_rate": 0.00014837752579952563,
+      "loss": 0.9808,
+      "step": 3567
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6983405351638794,
+      "learning_rate": 0.00014835121766752714,
+      "loss": 0.9849,
+      "step": 3568
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6801846027374268,
+      "learning_rate": 0.00014832490516726986,
+      "loss": 0.9781,
+      "step": 3569
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7692368030548096,
+      "learning_rate": 0.000148298588301131,
+      "loss": 0.8748,
+      "step": 3570
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.885966956615448,
+      "learning_rate": 0.00014827226707148818,
+      "loss": 1.1406,
+      "step": 3571
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6834810972213745,
+      "learning_rate": 0.00014824594148071934,
+      "loss": 0.8532,
+      "step": 3572
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8407056331634521,
+      "learning_rate": 0.00014821961153120287,
+      "loss": 0.923,
+      "step": 3573
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8173869848251343,
+      "learning_rate": 0.00014819327722531749,
+      "loss": 0.9967,
+      "step": 3574
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8253581523895264,
+      "learning_rate": 0.0001481669385654424,
+      "loss": 1.1102,
+      "step": 3575
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7063455581665039,
+      "learning_rate": 0.00014814059555395712,
+      "loss": 0.9345,
+      "step": 3576
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7028665542602539,
+      "learning_rate": 0.0001481142481932416,
+      "loss": 0.9175,
+      "step": 3577
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7196552753448486,
+      "learning_rate": 0.0001480878964856762,
+      "loss": 0.9596,
+      "step": 3578
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.9650554656982422,
+      "learning_rate": 0.0001480615404336416,
+      "loss": 1.0165,
+      "step": 3579
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7212446928024292,
+      "learning_rate": 0.00014803518003951897,
+      "loss": 0.9302,
+      "step": 3580
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8186121582984924,
+      "learning_rate": 0.00014800881530568978,
+      "loss": 0.7887,
+      "step": 3581
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7647170424461365,
+      "learning_rate": 0.00014798244623453596,
+      "loss": 1.0245,
+      "step": 3582
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7086816430091858,
+      "learning_rate": 0.0001479560728284398,
+      "loss": 0.969,
+      "step": 3583
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7693114280700684,
+      "learning_rate": 0.00014792969508978398,
+      "loss": 1.0047,
+      "step": 3584
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7128672003746033,
+      "learning_rate": 0.0001479033130209516,
+      "loss": 0.7372,
+      "step": 3585
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7224681973457336,
+      "learning_rate": 0.00014787692662432612,
+      "loss": 0.9198,
+      "step": 3586
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7356724739074707,
+      "learning_rate": 0.00014785053590229134,
+      "loss": 0.812,
+      "step": 3587
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7544524669647217,
+      "learning_rate": 0.00014782414085723165,
+      "loss": 0.9964,
+      "step": 3588
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7893449664115906,
+      "learning_rate": 0.00014779774149153155,
+      "loss": 1.086,
+      "step": 3589
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7163593769073486,
+      "learning_rate": 0.00014777133780757617,
+      "loss": 0.9591,
+      "step": 3590
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7546105980873108,
+      "learning_rate": 0.0001477449298077509,
+      "loss": 0.9707,
+      "step": 3591
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7793951630592346,
+      "learning_rate": 0.00014771851749444156,
+      "loss": 1.0296,
+      "step": 3592
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7393033504486084,
+      "learning_rate": 0.0001476921008700343,
+      "loss": 0.9886,
+      "step": 3593
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7126401662826538,
+      "learning_rate": 0.00014766567993691583,
+      "loss": 0.9804,
+      "step": 3594
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6830233931541443,
+      "learning_rate": 0.000147639254697473,
+      "loss": 0.8706,
+      "step": 3595
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8743071556091309,
+      "learning_rate": 0.00014761282515409326,
+      "loss": 0.9761,
+      "step": 3596
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7935216426849365,
+      "learning_rate": 0.00014758639130916436,
+      "loss": 1.0794,
+      "step": 3597
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8059981465339661,
+      "learning_rate": 0.00014755995316507442,
+      "loss": 1.0446,
+      "step": 3598
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7111148834228516,
+      "learning_rate": 0.000147533510724212,
+      "loss": 0.9263,
+      "step": 3599
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7591990828514099,
+      "learning_rate": 0.00014750706398896604,
+      "loss": 1.0087,
+      "step": 3600
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7582071423530579,
+      "learning_rate": 0.0001474806129617258,
+      "loss": 0.8304,
+      "step": 3601
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8318824172019958,
+      "learning_rate": 0.00014745415764488105,
+      "loss": 0.9744,
+      "step": 3602
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6948772072792053,
+      "learning_rate": 0.00014742769804082178,
+      "loss": 1.0361,
+      "step": 3603
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7453570365905762,
+      "learning_rate": 0.0001474012341519386,
+      "loss": 0.8178,
+      "step": 3604
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7137439250946045,
+      "learning_rate": 0.0001473747659806223,
+      "loss": 0.9621,
+      "step": 3605
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.860538899898529,
+      "learning_rate": 0.00014734829352926405,
+      "loss": 1.0802,
+      "step": 3606
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7084441781044006,
+      "learning_rate": 0.00014732181680025564,
+      "loss": 0.9248,
+      "step": 3607
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6707698702812195,
+      "learning_rate": 0.000147295335795989,
+      "loss": 0.8148,
+      "step": 3608
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6666585803031921,
+      "learning_rate": 0.00014726885051885653,
+      "loss": 0.9949,
+      "step": 3609
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7207210659980774,
+      "learning_rate": 0.00014724236097125108,
+      "loss": 0.964,
+      "step": 3610
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.838354229927063,
+      "learning_rate": 0.0001472158671555658,
+      "loss": 0.9344,
+      "step": 3611
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7047592401504517,
+      "learning_rate": 0.00014718936907419424,
+      "loss": 0.7781,
+      "step": 3612
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8515235781669617,
+      "learning_rate": 0.00014716286672953038,
+      "loss": 0.9691,
+      "step": 3613
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6779474020004272,
+      "learning_rate": 0.00014713636012396854,
+      "loss": 0.8669,
+      "step": 3614
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7808057069778442,
+      "learning_rate": 0.00014710984925990345,
+      "loss": 0.9782,
+      "step": 3615
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6925843954086304,
+      "learning_rate": 0.0001470833341397302,
+      "loss": 0.8911,
+      "step": 3616
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7705125212669373,
+      "learning_rate": 0.00014705681476584432,
+      "loss": 1.0675,
+      "step": 3617
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8819014430046082,
+      "learning_rate": 0.00014703029114064166,
+      "loss": 0.9161,
+      "step": 3618
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7916482090950012,
+      "learning_rate": 0.00014700376326651847,
+      "loss": 0.8913,
+      "step": 3619
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7155230641365051,
+      "learning_rate": 0.00014697723114587142,
+      "loss": 0.9563,
+      "step": 3620
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7341004610061646,
+      "learning_rate": 0.0001469506947810975,
+      "loss": 1.0403,
+      "step": 3621
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7908432483673096,
+      "learning_rate": 0.00014692415417459414,
+      "loss": 0.9515,
+      "step": 3622
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7485032081604004,
+      "learning_rate": 0.00014689760932875916,
+      "loss": 0.9666,
+      "step": 3623
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.8104244470596313,
+      "learning_rate": 0.00014687106024599067,
+      "loss": 1.049,
+      "step": 3624
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6753594279289246,
+      "learning_rate": 0.00014684450692868732,
+      "loss": 0.9839,
+      "step": 3625
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.9177693724632263,
+      "learning_rate": 0.00014681794937924795,
+      "loss": 1.0646,
+      "step": 3626
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7891800999641418,
+      "learning_rate": 0.00014679138760007198,
+      "loss": 0.9064,
+      "step": 3627
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.728115439414978,
+      "learning_rate": 0.00014676482159355905,
+      "loss": 1.0293,
+      "step": 3628
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7733248472213745,
+      "learning_rate": 0.00014673825136210929,
+      "loss": 0.9834,
+      "step": 3629
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7536121606826782,
+      "learning_rate": 0.0001467116769081231,
+      "loss": 0.8856,
+      "step": 3630
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7316053509712219,
+      "learning_rate": 0.00014668509823400144,
+      "loss": 0.9515,
+      "step": 3631
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6929289698600769,
+      "learning_rate": 0.00014665851534214545,
+      "loss": 1.1022,
+      "step": 3632
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.8146265745162964,
+      "learning_rate": 0.0001466319282349568,
+      "loss": 0.9602,
+      "step": 3633
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7151821851730347,
+      "learning_rate": 0.00014660533691483746,
+      "loss": 0.995,
+      "step": 3634
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7196069955825806,
+      "learning_rate": 0.0001465787413841898,
+      "loss": 0.981,
+      "step": 3635
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.0193686485290527,
+      "learning_rate": 0.00014655214164541658,
+      "loss": 1.1679,
+      "step": 3636
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6898132562637329,
+      "learning_rate": 0.00014652553770092095,
+      "loss": 0.9004,
+      "step": 3637
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7551093697547913,
+      "learning_rate": 0.0001464989295531064,
+      "loss": 1.0437,
+      "step": 3638
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7888858914375305,
+      "learning_rate": 0.00014647231720437686,
+      "loss": 0.9664,
+      "step": 3639
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7758256196975708,
+      "learning_rate": 0.0001464457006571366,
+      "loss": 0.864,
+      "step": 3640
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7542037963867188,
+      "learning_rate": 0.0001464190799137902,
+      "loss": 1.0199,
+      "step": 3641
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7568319439888,
+      "learning_rate": 0.00014639245497674282,
+      "loss": 0.9234,
+      "step": 3642
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.776556134223938,
+      "learning_rate": 0.00014636582584839978,
+      "loss": 1.0185,
+      "step": 3643
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7782880067825317,
+      "learning_rate": 0.0001463391925311669,
+      "loss": 1.205,
+      "step": 3644
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.8340494632720947,
+      "learning_rate": 0.00014631255502745036,
+      "loss": 1.0501,
+      "step": 3645
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7521584630012512,
+      "learning_rate": 0.00014628591333965663,
+      "loss": 1.0301,
+      "step": 3646
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.80910724401474,
+      "learning_rate": 0.00014625926747019276,
+      "loss": 1.1471,
+      "step": 3647
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7709221243858337,
+      "learning_rate": 0.00014623261742146602,
+      "loss": 0.9235,
+      "step": 3648
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.8370350003242493,
+      "learning_rate": 0.000146205963195884,
+      "loss": 0.8984,
+      "step": 3649
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8251968026161194,
+      "learning_rate": 0.00014617930479585487,
+      "loss": 1.0234,
+      "step": 3650
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8307152986526489,
+      "learning_rate": 0.00014615264222378696,
+      "loss": 1.0159,
+      "step": 3651
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8258378505706787,
+      "learning_rate": 0.0001461259754820892,
+      "loss": 1.1309,
+      "step": 3652
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.661595344543457,
+      "learning_rate": 0.0001460993045731707,
+      "loss": 0.7346,
+      "step": 3653
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8245034217834473,
+      "learning_rate": 0.00014607262949944104,
+      "loss": 1.0182,
+      "step": 3654
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7197308540344238,
+      "learning_rate": 0.00014604595026331015,
+      "loss": 0.9962,
+      "step": 3655
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8030989170074463,
+      "learning_rate": 0.0001460192668671884,
+      "loss": 0.9241,
+      "step": 3656
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8018397688865662,
+      "learning_rate": 0.00014599257931348643,
+      "loss": 0.9935,
+      "step": 3657
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7212077975273132,
+      "learning_rate": 0.00014596588760461534,
+      "loss": 0.8948,
+      "step": 3658
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7394177317619324,
+      "learning_rate": 0.00014593919174298655,
+      "loss": 1.0604,
+      "step": 3659
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7774109244346619,
+      "learning_rate": 0.00014591249173101192,
+      "loss": 0.8841,
+      "step": 3660
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.781120240688324,
+      "learning_rate": 0.00014588578757110358,
+      "loss": 1.0129,
+      "step": 3661
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8257413506507874,
+      "learning_rate": 0.00014585907926567416,
+      "loss": 1.0161,
+      "step": 3662
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7354034185409546,
+      "learning_rate": 0.00014583236681713657,
+      "loss": 0.9752,
+      "step": 3663
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7330094575881958,
+      "learning_rate": 0.00014580565022790414,
+      "loss": 0.9434,
+      "step": 3664
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7110805511474609,
+      "learning_rate": 0.00014577892950039058,
+      "loss": 1.0484,
+      "step": 3665
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8536974191665649,
+      "learning_rate": 0.00014575220463700992,
+      "loss": 0.903,
+      "step": 3666
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7048803567886353,
+      "learning_rate": 0.00014572547564017664,
+      "loss": 0.8204,
+      "step": 3667
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7898951172828674,
+      "learning_rate": 0.0001456987425123055,
+      "loss": 0.9791,
+      "step": 3668
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8470849394798279,
+      "learning_rate": 0.00014567200525581175,
+      "loss": 1.0017,
+      "step": 3669
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8138737082481384,
+      "learning_rate": 0.0001456452638731109,
+      "loss": 0.9451,
+      "step": 3670
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7108284831047058,
+      "learning_rate": 0.00014561851836661892,
+      "loss": 0.83,
+      "step": 3671
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8926004767417908,
+      "learning_rate": 0.00014559176873875208,
+      "loss": 0.8786,
+      "step": 3672
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7158498764038086,
+      "learning_rate": 0.00014556501499192709,
+      "loss": 1.097,
+      "step": 3673
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7436313033103943,
+      "learning_rate": 0.00014553825712856097,
+      "loss": 0.8637,
+      "step": 3674
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.723731517791748,
+      "learning_rate": 0.00014551149515107116,
+      "loss": 0.947,
+      "step": 3675
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.784081757068634,
+      "learning_rate": 0.00014548472906187544,
+      "loss": 0.9916,
+      "step": 3676
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.789717435836792,
+      "learning_rate": 0.00014545795886339202,
+      "loss": 1.1036,
+      "step": 3677
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7132088541984558,
+      "learning_rate": 0.00014543118455803935,
+      "loss": 0.9898,
+      "step": 3678
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.0612138509750366,
+      "learning_rate": 0.00014540440614823643,
+      "loss": 1.0354,
+      "step": 3679
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7618119120597839,
+      "learning_rate": 0.00014537762363640248,
+      "loss": 1.0126,
+      "step": 3680
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6965414881706238,
+      "learning_rate": 0.00014535083702495714,
+      "loss": 1.0952,
+      "step": 3681
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7453968524932861,
+      "learning_rate": 0.00014532404631632047,
+      "loss": 1.3152,
+      "step": 3682
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.8747276663780212,
+      "learning_rate": 0.00014529725151291283,
+      "loss": 1.0533,
+      "step": 3683
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7281939387321472,
+      "learning_rate": 0.000145270452617155,
+      "loss": 0.9864,
+      "step": 3684
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7459536790847778,
+      "learning_rate": 0.0001452436496314681,
+      "loss": 1.047,
+      "step": 3685
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.8038800358772278,
+      "learning_rate": 0.0001452168425582736,
+      "loss": 0.965,
+      "step": 3686
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.8129239678382874,
+      "learning_rate": 0.00014519003139999337,
+      "loss": 0.8938,
+      "step": 3687
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.784221887588501,
+      "learning_rate": 0.00014516321615904972,
+      "loss": 1.0522,
+      "step": 3688
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7523348331451416,
+      "learning_rate": 0.0001451363968378652,
+      "loss": 0.9661,
+      "step": 3689
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7557814717292786,
+      "learning_rate": 0.00014510957343886278,
+      "loss": 0.9508,
+      "step": 3690
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7756308913230896,
+      "learning_rate": 0.00014508274596446578,
+      "loss": 0.9669,
+      "step": 3691
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.799447774887085,
+      "learning_rate": 0.00014505591441709795,
+      "loss": 0.9815,
+      "step": 3692
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7798731327056885,
+      "learning_rate": 0.0001450290787991834,
+      "loss": 1.0101,
+      "step": 3693
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7313754558563232,
+      "learning_rate": 0.0001450022391131465,
+      "loss": 0.8448,
+      "step": 3694
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7849476933479309,
+      "learning_rate": 0.0001449753953614121,
+      "loss": 1.0423,
+      "step": 3695
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.8375855088233948,
+      "learning_rate": 0.0001449485475464054,
+      "loss": 1.1203,
+      "step": 3696
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.79655522108078,
+      "learning_rate": 0.0001449216956705519,
+      "loss": 1.0499,
+      "step": 3697
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7769049406051636,
+      "learning_rate": 0.00014489483973627754,
+      "loss": 1.0213,
+      "step": 3698
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7109012603759766,
+      "learning_rate": 0.00014486797974600865,
+      "loss": 0.9309,
+      "step": 3699
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7079145908355713,
+      "learning_rate": 0.0001448411157021718,
+      "loss": 0.8,
+      "step": 3700
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7436159253120422,
+      "learning_rate": 0.000144814247607194,
+      "loss": 1.0672,
+      "step": 3701
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7641329169273376,
+      "learning_rate": 0.00014478737546350272,
+      "loss": 0.9365,
+      "step": 3702
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.9353510737419128,
+      "learning_rate": 0.00014476049927352564,
+      "loss": 1.0433,
+      "step": 3703
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.757537305355072,
+      "learning_rate": 0.00014473361903969087,
+      "loss": 1.058,
+      "step": 3704
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6940644383430481,
+      "learning_rate": 0.0001447067347644269,
+      "loss": 0.8961,
+      "step": 3705
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7819743752479553,
+      "learning_rate": 0.00014467984645016258,
+      "loss": 0.9163,
+      "step": 3706
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7648773193359375,
+      "learning_rate": 0.0001446529540993271,
+      "loss": 0.8864,
+      "step": 3707
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.770237386226654,
+      "learning_rate": 0.00014462605771435008,
+      "loss": 0.8647,
+      "step": 3708
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7355459332466125,
+      "learning_rate": 0.0001445991572976614,
+      "loss": 0.9652,
+      "step": 3709
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7374024391174316,
+      "learning_rate": 0.00014457225285169134,
+      "loss": 0.906,
+      "step": 3710
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8194604516029358,
+      "learning_rate": 0.00014454534437887063,
+      "loss": 1.0381,
+      "step": 3711
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.762771487236023,
+      "learning_rate": 0.00014451843188163029,
+      "loss": 0.9989,
+      "step": 3712
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.881404459476471,
+      "learning_rate": 0.00014449151536240166,
+      "loss": 0.9936,
+      "step": 3713
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8181387782096863,
+      "learning_rate": 0.00014446459482361654,
+      "loss": 0.9665,
+      "step": 3714
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7393097877502441,
+      "learning_rate": 0.00014443767026770706,
+      "loss": 0.9013,
+      "step": 3715
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7258290648460388,
+      "learning_rate": 0.00014441074169710562,
+      "loss": 0.94,
+      "step": 3716
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.738956868648529,
+      "learning_rate": 0.00014438380911424516,
+      "loss": 0.8634,
+      "step": 3717
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8704954981803894,
+      "learning_rate": 0.00014435687252155885,
+      "loss": 1.1659,
+      "step": 3718
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7542216777801514,
+      "learning_rate": 0.00014432993192148023,
+      "loss": 0.9391,
+      "step": 3719
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6866111755371094,
+      "learning_rate": 0.0001443029873164433,
+      "loss": 0.8937,
+      "step": 3720
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.715201735496521,
+      "learning_rate": 0.00014427603870888225,
+      "loss": 0.9683,
+      "step": 3721
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.9084267020225525,
+      "learning_rate": 0.00014424908610123182,
+      "loss": 0.9476,
+      "step": 3722
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7752277851104736,
+      "learning_rate": 0.000144222129495927,
+      "loss": 0.8479,
+      "step": 3723
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7288671731948853,
+      "learning_rate": 0.00014419516889540315,
+      "loss": 0.9883,
+      "step": 3724
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8041513562202454,
+      "learning_rate": 0.00014416820430209607,
+      "loss": 0.9846,
+      "step": 3725
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7570642828941345,
+      "learning_rate": 0.00014414123571844178,
+      "loss": 0.8865,
+      "step": 3726
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8212285041809082,
+      "learning_rate": 0.00014411426314687678,
+      "loss": 1.0912,
+      "step": 3727
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7754359245300293,
+      "learning_rate": 0.00014408728658983784,
+      "loss": 1.1325,
+      "step": 3728
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.8131769299507141,
+      "learning_rate": 0.00014406030604976222,
+      "loss": 1.0475,
+      "step": 3729
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6917926073074341,
+      "learning_rate": 0.0001440333215290874,
+      "loss": 0.8188,
+      "step": 3730
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7929239273071289,
+      "learning_rate": 0.00014400633303025131,
+      "loss": 1.0119,
+      "step": 3731
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7956200838088989,
+      "learning_rate": 0.00014397934055569218,
+      "loss": 1.1565,
+      "step": 3732
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.9268775582313538,
+      "learning_rate": 0.00014395234410784867,
+      "loss": 0.9625,
+      "step": 3733
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7056459188461304,
+      "learning_rate": 0.00014392534368915972,
+      "loss": 0.9952,
+      "step": 3734
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6934985518455505,
+      "learning_rate": 0.00014389833930206463,
+      "loss": 0.9829,
+      "step": 3735
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7250593304634094,
+      "learning_rate": 0.00014387133094900315,
+      "loss": 0.9692,
+      "step": 3736
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7265465259552002,
+      "learning_rate": 0.0001438443186324153,
+      "loss": 1.033,
+      "step": 3737
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7638846039772034,
+      "learning_rate": 0.00014381730235474155,
+      "loss": 0.95,
+      "step": 3738
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.8740837574005127,
+      "learning_rate": 0.00014379028211842265,
+      "loss": 1.1423,
+      "step": 3739
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7578192949295044,
+      "learning_rate": 0.00014376325792589964,
+      "loss": 1.0836,
+      "step": 3740
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7851478457450867,
+      "learning_rate": 0.00014373622977961409,
+      "loss": 1.0269,
+      "step": 3741
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7320329546928406,
+      "learning_rate": 0.0001437091976820078,
+      "loss": 0.8852,
+      "step": 3742
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7559787631034851,
+      "learning_rate": 0.00014368216163552298,
+      "loss": 1.0108,
+      "step": 3743
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7940683960914612,
+      "learning_rate": 0.0001436551216426022,
+      "loss": 1.0229,
+      "step": 3744
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7840447425842285,
+      "learning_rate": 0.00014362807770568835,
+      "loss": 1.049,
+      "step": 3745
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6953481435775757,
+      "learning_rate": 0.0001436010298272247,
+      "loss": 0.9457,
+      "step": 3746
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7428427934646606,
+      "learning_rate": 0.00014357397800965487,
+      "loss": 0.7076,
+      "step": 3747
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7457131147384644,
+      "learning_rate": 0.00014354692225542284,
+      "loss": 0.9413,
+      "step": 3748
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6950200200080872,
+      "learning_rate": 0.00014351986256697297,
+      "loss": 0.7335,
+      "step": 3749
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7549350261688232,
+      "learning_rate": 0.0001434927989467499,
+      "loss": 0.9044,
+      "step": 3750
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7664708495140076,
+      "learning_rate": 0.00014346573139719867,
+      "loss": 1.1704,
+      "step": 3751
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6717768907546997,
+      "learning_rate": 0.00014343865992076478,
+      "loss": 0.8295,
+      "step": 3752
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.701734721660614,
+      "learning_rate": 0.00014341158451989386,
+      "loss": 0.9549,
+      "step": 3753
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.932205080986023,
+      "learning_rate": 0.00014338450519703214,
+      "loss": 0.926,
+      "step": 3754
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.9582824110984802,
+      "learning_rate": 0.00014335742195462596,
+      "loss": 1.0883,
+      "step": 3755
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7160861492156982,
+      "learning_rate": 0.00014333033479512218,
+      "loss": 0.9851,
+      "step": 3756
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7445812821388245,
+      "learning_rate": 0.00014330324372096803,
+      "loss": 0.8497,
+      "step": 3757
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6866586804389954,
+      "learning_rate": 0.000143276148734611,
+      "loss": 0.9647,
+      "step": 3758
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8147525191307068,
+      "learning_rate": 0.0001432490498384989,
+      "loss": 1.0837,
+      "step": 3759
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7915096879005432,
+      "learning_rate": 0.0001432219470350801,
+      "loss": 0.975,
+      "step": 3760
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7530489563941956,
+      "learning_rate": 0.00014319484032680307,
+      "loss": 0.7848,
+      "step": 3761
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7870637774467468,
+      "learning_rate": 0.00014316772971611675,
+      "loss": 0.7446,
+      "step": 3762
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7144870758056641,
+      "learning_rate": 0.00014314061520547054,
+      "loss": 1.0712,
+      "step": 3763
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8082242608070374,
+      "learning_rate": 0.00014311349679731395,
+      "loss": 0.9206,
+      "step": 3764
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7679134607315063,
+      "learning_rate": 0.00014308637449409706,
+      "loss": 0.909,
+      "step": 3765
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7546840310096741,
+      "learning_rate": 0.00014305924829827017,
+      "loss": 1.0257,
+      "step": 3766
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7072516679763794,
+      "learning_rate": 0.000143032118212284,
+      "loss": 1.0949,
+      "step": 3767
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.9368565082550049,
+      "learning_rate": 0.00014300498423858962,
+      "loss": 0.8724,
+      "step": 3768
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7789929509162903,
+      "learning_rate": 0.00014297784637963838,
+      "loss": 1.0801,
+      "step": 3769
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8462679982185364,
+      "learning_rate": 0.00014295070463788207,
+      "loss": 0.9488,
+      "step": 3770
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7308972477912903,
+      "learning_rate": 0.00014292355901577276,
+      "loss": 1.1541,
+      "step": 3771
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7259547710418701,
+      "learning_rate": 0.00014289640951576296,
+      "loss": 0.9442,
+      "step": 3772
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7557168006896973,
+      "learning_rate": 0.00014286925614030542,
+      "loss": 1.1086,
+      "step": 3773
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7561401128768921,
+      "learning_rate": 0.00014284209889185335,
+      "loss": 1.1196,
+      "step": 3774
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8499369621276855,
+      "learning_rate": 0.00014281493777286016,
+      "loss": 0.8272,
+      "step": 3775
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6794196367263794,
+      "learning_rate": 0.00014278777278577977,
+      "loss": 1.0419,
+      "step": 3776
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7229602336883545,
+      "learning_rate": 0.0001427606039330664,
+      "loss": 0.8417,
+      "step": 3777
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7260411977767944,
+      "learning_rate": 0.00014273343121717457,
+      "loss": 0.9832,
+      "step": 3778
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.9786548614501953,
+      "learning_rate": 0.00014270625464055917,
+      "loss": 1.0393,
+      "step": 3779
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.816220223903656,
+      "learning_rate": 0.00014267907420567548,
+      "loss": 0.9016,
+      "step": 3780
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.738017201423645,
+      "learning_rate": 0.00014265188991497907,
+      "loss": 1.0055,
+      "step": 3781
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7251632213592529,
+      "learning_rate": 0.0001426247017709259,
+      "loss": 0.9767,
+      "step": 3782
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7959434986114502,
+      "learning_rate": 0.00014259750977597228,
+      "loss": 1.03,
+      "step": 3783
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7960846424102783,
+      "learning_rate": 0.00014257031393257485,
+      "loss": 1.0113,
+      "step": 3784
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8744489550590515,
+      "learning_rate": 0.00014254311424319056,
+      "loss": 1.0178,
+      "step": 3785
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8908901214599609,
+      "learning_rate": 0.0001425159107102768,
+      "loss": 0.9478,
+      "step": 3786
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7419928312301636,
+      "learning_rate": 0.0001424887033362912,
+      "loss": 0.8289,
+      "step": 3787
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7056096792221069,
+      "learning_rate": 0.00014246149212369184,
+      "loss": 0.9822,
+      "step": 3788
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8795596361160278,
+      "learning_rate": 0.0001424342770749371,
+      "loss": 1.1753,
+      "step": 3789
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8885317444801331,
+      "learning_rate": 0.00014240705819248568,
+      "loss": 1.1536,
+      "step": 3790
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7233941555023193,
+      "learning_rate": 0.00014237983547879662,
+      "loss": 0.9257,
+      "step": 3791
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.823823869228363,
+      "learning_rate": 0.00014235260893632944,
+      "loss": 1.181,
+      "step": 3792
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8890931606292725,
+      "learning_rate": 0.00014232537856754383,
+      "loss": 0.9468,
+      "step": 3793
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7318471074104309,
+      "learning_rate": 0.00014229814437489988,
+      "loss": 0.9542,
+      "step": 3794
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.9147783517837524,
+      "learning_rate": 0.0001422709063608581,
+      "loss": 1.0679,
+      "step": 3795
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.813987672328949,
+      "learning_rate": 0.00014224366452787924,
+      "loss": 1.0211,
+      "step": 3796
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6452633738517761,
+      "learning_rate": 0.0001422164188784245,
+      "loss": 0.8243,
+      "step": 3797
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7558382153511047,
+      "learning_rate": 0.00014218916941495535,
+      "loss": 0.8988,
+      "step": 3798
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7191858887672424,
+      "learning_rate": 0.0001421619161399336,
+      "loss": 0.8812,
+      "step": 3799
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7513072490692139,
+      "learning_rate": 0.00014213465905582145,
+      "loss": 1.0562,
+      "step": 3800
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.671654999256134,
+      "learning_rate": 0.00014210739816508143,
+      "loss": 0.9391,
+      "step": 3801
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7372080683708191,
+      "learning_rate": 0.00014208013347017637,
+      "loss": 0.9444,
+      "step": 3802
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8163162469863892,
+      "learning_rate": 0.00014205286497356956,
+      "loss": 1.0991,
+      "step": 3803
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8253856301307678,
+      "learning_rate": 0.00014202559267772444,
+      "loss": 0.9762,
+      "step": 3804
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7247931361198425,
+      "learning_rate": 0.00014199831658510503,
+      "loss": 0.9654,
+      "step": 3805
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.76895672082901,
+      "learning_rate": 0.00014197103669817546,
+      "loss": 1.1187,
+      "step": 3806
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6907783150672913,
+      "learning_rate": 0.00014194375301940037,
+      "loss": 1.0392,
+      "step": 3807
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7909299731254578,
+      "learning_rate": 0.0001419164655512447,
+      "loss": 0.9487,
+      "step": 3808
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.804953396320343,
+      "learning_rate": 0.0001418891742961737,
+      "loss": 0.9633,
+      "step": 3809
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7446818947792053,
+      "learning_rate": 0.00014186187925665296,
+      "loss": 0.8268,
+      "step": 3810
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7549101114273071,
+      "learning_rate": 0.00014183458043514848,
+      "loss": 1.0275,
+      "step": 3811
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6599337458610535,
+      "learning_rate": 0.00014180727783412648,
+      "loss": 0.974,
+      "step": 3812
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.8352928757667542,
+      "learning_rate": 0.00014177997145605367,
+      "loss": 1.1111,
+      "step": 3813
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7139683961868286,
+      "learning_rate": 0.000141752661303397,
+      "loss": 1.1222,
+      "step": 3814
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.8132135272026062,
+      "learning_rate": 0.0001417253473786238,
+      "loss": 0.9324,
+      "step": 3815
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7724325060844421,
+      "learning_rate": 0.00014169802968420168,
+      "loss": 0.9621,
+      "step": 3816
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.903314471244812,
+      "learning_rate": 0.00014167070822259867,
+      "loss": 0.9272,
+      "step": 3817
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.809247612953186,
+      "learning_rate": 0.00014164338299628312,
+      "loss": 0.8033,
+      "step": 3818
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.8909575343132019,
+      "learning_rate": 0.0001416160540077237,
+      "loss": 1.0189,
+      "step": 3819
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.689109206199646,
+      "learning_rate": 0.00014158872125938943,
+      "loss": 0.96,
+      "step": 3820
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.8286650776863098,
+      "learning_rate": 0.00014156138475374968,
+      "loss": 1.1713,
+      "step": 3821
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7234490513801575,
+      "learning_rate": 0.00014153404449327413,
+      "loss": 0.8662,
+      "step": 3822
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6873946785926819,
+      "learning_rate": 0.0001415067004804328,
+      "loss": 0.8921,
+      "step": 3823
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.8717650175094604,
+      "learning_rate": 0.00014147935271769613,
+      "loss": 1.0844,
+      "step": 3824
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.9551481008529663,
+      "learning_rate": 0.00014145200120753478,
+      "loss": 0.9103,
+      "step": 3825
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.8124105930328369,
+      "learning_rate": 0.00014142464595241983,
+      "loss": 1.1163,
+      "step": 3826
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7727875709533691,
+      "learning_rate": 0.00014139728695482267,
+      "loss": 0.8512,
+      "step": 3827
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.707558274269104,
+      "learning_rate": 0.000141369924217215,
+      "loss": 1.0252,
+      "step": 3828
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7345936298370361,
+      "learning_rate": 0.00014134255774206892,
+      "loss": 0.9271,
+      "step": 3829
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.781466543674469,
+      "learning_rate": 0.00014131518753185684,
+      "loss": 0.9791,
+      "step": 3830
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.8418498039245605,
+      "learning_rate": 0.00014128781358905147,
+      "loss": 0.9273,
+      "step": 3831
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7959011793136597,
+      "learning_rate": 0.00014126043591612594,
+      "loss": 0.9947,
+      "step": 3832
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7734252214431763,
+      "learning_rate": 0.00014123305451555363,
+      "loss": 0.9747,
+      "step": 3833
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7216122150421143,
+      "learning_rate": 0.0001412056693898083,
+      "loss": 0.968,
+      "step": 3834
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7466599345207214,
+      "learning_rate": 0.0001411782805413641,
+      "loss": 0.8883,
+      "step": 3835
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8146680593490601,
+      "learning_rate": 0.0001411508879726953,
+      "loss": 1.0624,
+      "step": 3836
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6900213360786438,
+      "learning_rate": 0.00014112349168627684,
+      "loss": 0.9781,
+      "step": 3837
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6831471920013428,
+      "learning_rate": 0.00014109609168458373,
+      "loss": 0.9465,
+      "step": 3838
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7674217224121094,
+      "learning_rate": 0.00014106868797009143,
+      "loss": 0.9436,
+      "step": 3839
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7712320685386658,
+      "learning_rate": 0.0001410412805452757,
+      "loss": 0.9189,
+      "step": 3840
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.806429386138916,
+      "learning_rate": 0.00014101386941261262,
+      "loss": 0.8869,
+      "step": 3841
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7063391208648682,
+      "learning_rate": 0.0001409864545745787,
+      "loss": 0.9804,
+      "step": 3842
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.841386616230011,
+      "learning_rate": 0.00014095903603365066,
+      "loss": 1.1257,
+      "step": 3843
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.9420889616012573,
+      "learning_rate": 0.0001409316137923056,
+      "loss": 1.0279,
+      "step": 3844
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.9218207597732544,
+      "learning_rate": 0.00014090418785302103,
+      "loss": 1.1393,
+      "step": 3845
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8026069402694702,
+      "learning_rate": 0.00014087675821827464,
+      "loss": 0.9735,
+      "step": 3846
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7738460898399353,
+      "learning_rate": 0.0001408493248905446,
+      "loss": 0.9638,
+      "step": 3847
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6933304071426392,
+      "learning_rate": 0.00014082188787230938,
+      "loss": 0.8385,
+      "step": 3848
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7282217741012573,
+      "learning_rate": 0.0001407944471660477,
+      "loss": 1.0095,
+      "step": 3849
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7486565113067627,
+      "learning_rate": 0.00014076700277423869,
+      "loss": 0.9736,
+      "step": 3850
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8711503148078918,
+      "learning_rate": 0.0001407395546993618,
+      "loss": 1.0308,
+      "step": 3851
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7639390826225281,
+      "learning_rate": 0.00014071210294389682,
+      "loss": 1.0158,
+      "step": 3852
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7631992101669312,
+      "learning_rate": 0.00014068464751032384,
+      "loss": 1.0516,
+      "step": 3853
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7923541069030762,
+      "learning_rate": 0.00014065718840112335,
+      "loss": 0.9994,
+      "step": 3854
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7509554028511047,
+      "learning_rate": 0.00014062972561877603,
+      "loss": 0.9913,
+      "step": 3855
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7766771912574768,
+      "learning_rate": 0.00014060225916576305,
+      "loss": 0.984,
+      "step": 3856
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7930575609207153,
+      "learning_rate": 0.0001405747890445659,
+      "loss": 0.9921,
+      "step": 3857
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7377687692642212,
+      "learning_rate": 0.00014054731525766628,
+      "loss": 0.9928,
+      "step": 3858
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8305283188819885,
+      "learning_rate": 0.00014051983780754627,
+      "loss": 1.0663,
+      "step": 3859
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.838887095451355,
+      "learning_rate": 0.00014049235669668834,
+      "loss": 1.049,
+      "step": 3860
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7463981509208679,
+      "learning_rate": 0.00014046487192757528,
+      "loss": 0.8271,
+      "step": 3861
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7429585456848145,
+      "learning_rate": 0.00014043738350269015,
+      "loss": 1.0368,
+      "step": 3862
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7435213327407837,
+      "learning_rate": 0.00014040989142451632,
+      "loss": 0.9739,
+      "step": 3863
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6512870192527771,
+      "learning_rate": 0.00014038239569553765,
+      "loss": 0.7547,
+      "step": 3864
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7357031106948853,
+      "learning_rate": 0.0001403548963182382,
+      "loss": 0.8476,
+      "step": 3865
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7592230439186096,
+      "learning_rate": 0.00014032739329510228,
+      "loss": 1.045,
+      "step": 3866
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.751270592212677,
+      "learning_rate": 0.00014029988662861477,
+      "loss": 1.1283,
+      "step": 3867
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8035612106323242,
+      "learning_rate": 0.00014027237632126068,
+      "loss": 0.9893,
+      "step": 3868
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7580231428146362,
+      "learning_rate": 0.00014024486237552538,
+      "loss": 0.9774,
+      "step": 3869
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7281249165534973,
+      "learning_rate": 0.00014021734479389465,
+      "loss": 0.95,
+      "step": 3870
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7264341711997986,
+      "learning_rate": 0.0001401898235788545,
+      "loss": 1.0655,
+      "step": 3871
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7836666703224182,
+      "learning_rate": 0.00014016229873289136,
+      "loss": 0.8045,
+      "step": 3872
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8120614886283875,
+      "learning_rate": 0.00014013477025849195,
+      "loss": 0.9924,
+      "step": 3873
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6951356530189514,
+      "learning_rate": 0.00014010723815814326,
+      "loss": 1.031,
+      "step": 3874
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7149147987365723,
+      "learning_rate": 0.00014007970243433272,
+      "loss": 0.8887,
+      "step": 3875
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7184805274009705,
+      "learning_rate": 0.00014005216308954796,
+      "loss": 0.9276,
+      "step": 3876
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.768484354019165,
+      "learning_rate": 0.00014002462012627703,
+      "loss": 1.0148,
+      "step": 3877
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7808959484100342,
+      "learning_rate": 0.00013999707354700835,
+      "loss": 0.9402,
+      "step": 3878
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8177810311317444,
+      "learning_rate": 0.0001399695233542305,
+      "loss": 1.0563,
+      "step": 3879
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.800896942615509,
+      "learning_rate": 0.00013994196955043254,
+      "loss": 1.0257,
+      "step": 3880
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8390647172927856,
+      "learning_rate": 0.00013991441213810374,
+      "loss": 1.0125,
+      "step": 3881
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6802325248718262,
+      "learning_rate": 0.00013988685111973384,
+      "loss": 0.8674,
+      "step": 3882
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8133691549301147,
+      "learning_rate": 0.00013985928649781277,
+      "loss": 0.9342,
+      "step": 3883
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.736500084400177,
+      "learning_rate": 0.00013983171827483085,
+      "loss": 0.9786,
+      "step": 3884
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7998164892196655,
+      "learning_rate": 0.0001398041464532787,
+      "loss": 1.0348,
+      "step": 3885
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.8324884176254272,
+      "learning_rate": 0.00013977657103564727,
+      "loss": 1.1051,
+      "step": 3886
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7612653374671936,
+      "learning_rate": 0.0001397489920244279,
+      "loss": 0.9836,
+      "step": 3887
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.8343334794044495,
+      "learning_rate": 0.00013972140942211214,
+      "loss": 1.0232,
+      "step": 3888
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7189517617225647,
+      "learning_rate": 0.00013969382323119196,
+      "loss": 0.9571,
+      "step": 3889
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7711721062660217,
+      "learning_rate": 0.00013966623345415956,
+      "loss": 0.9215,
+      "step": 3890
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.769665002822876,
+      "learning_rate": 0.00013963864009350758,
+      "loss": 0.8627,
+      "step": 3891
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7339462041854858,
+      "learning_rate": 0.0001396110431517289,
+      "loss": 0.9453,
+      "step": 3892
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.8329907655715942,
+      "learning_rate": 0.00013958344263131673,
+      "loss": 0.8773,
+      "step": 3893
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7495461702346802,
+      "learning_rate": 0.00013955583853476468,
+      "loss": 0.9406,
+      "step": 3894
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7086147665977478,
+      "learning_rate": 0.00013952823086456654,
+      "loss": 0.6748,
+      "step": 3895
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7977877855300903,
+      "learning_rate": 0.00013950061962321656,
+      "loss": 0.9603,
+      "step": 3896
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7677198052406311,
+      "learning_rate": 0.00013947300481320925,
+      "loss": 1.0029,
+      "step": 3897
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7576024532318115,
+      "learning_rate": 0.0001394453864370395,
+      "loss": 1.1522,
+      "step": 3898
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7466287612915039,
+      "learning_rate": 0.00013941776449720236,
+      "loss": 0.8189,
+      "step": 3899
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6966318488121033,
+      "learning_rate": 0.0001393901389961934,
+      "loss": 0.9375,
+      "step": 3900
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7954654097557068,
+      "learning_rate": 0.00013936250993650838,
+      "loss": 1.0475,
+      "step": 3901
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6510365009307861,
+      "learning_rate": 0.00013933487732064353,
+      "loss": 0.8256,
+      "step": 3902
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7718179225921631,
+      "learning_rate": 0.00013930724115109516,
+      "loss": 1.1221,
+      "step": 3903
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.747000515460968,
+      "learning_rate": 0.00013927960143036015,
+      "loss": 0.9645,
+      "step": 3904
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7126678824424744,
+      "learning_rate": 0.00013925195816093553,
+      "loss": 0.851,
+      "step": 3905
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7508659362792969,
+      "learning_rate": 0.00013922431134531874,
+      "loss": 0.7579,
+      "step": 3906
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.8252602815628052,
+      "learning_rate": 0.00013919666098600753,
+      "loss": 0.9258,
+      "step": 3907
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.8027669787406921,
+      "learning_rate": 0.00013916900708549988,
+      "loss": 1.1709,
+      "step": 3908
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7662482261657715,
+      "learning_rate": 0.00013914134964629427,
+      "loss": 0.7885,
+      "step": 3909
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7759925127029419,
+      "learning_rate": 0.00013911368867088928,
+      "loss": 1.0471,
+      "step": 3910
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.829074501991272,
+      "learning_rate": 0.00013908602416178403,
+      "loss": 1.0349,
+      "step": 3911
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7995732426643372,
+      "learning_rate": 0.00013905835612147778,
+      "loss": 1.0179,
+      "step": 3912
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6526477932929993,
+      "learning_rate": 0.0001390306845524702,
+      "loss": 0.8083,
+      "step": 3913
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.791499674320221,
+      "learning_rate": 0.00013900300945726127,
+      "loss": 1.0309,
+      "step": 3914
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.8595665693283081,
+      "learning_rate": 0.0001389753308383513,
+      "loss": 0.9313,
+      "step": 3915
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6814246773719788,
+      "learning_rate": 0.0001389476486982408,
+      "loss": 0.8139,
+      "step": 3916
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.8005412220954895,
+      "learning_rate": 0.00013891996303943077,
+      "loss": 0.8702,
+      "step": 3917
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.819680392742157,
+      "learning_rate": 0.00013889227386442248,
+      "loss": 0.7957,
+      "step": 3918
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7297448515892029,
+      "learning_rate": 0.00013886458117571743,
+      "loss": 0.9389,
+      "step": 3919
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7080660462379456,
+      "learning_rate": 0.00013883688497581755,
+      "loss": 1.0576,
+      "step": 3920
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.9082624316215515,
+      "learning_rate": 0.00013880918526722497,
+      "loss": 1.178,
+      "step": 3921
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.8355306386947632,
+      "learning_rate": 0.00013878148205244225,
+      "loss": 0.9952,
+      "step": 3922
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7118585705757141,
+      "learning_rate": 0.00013875377533397221,
+      "loss": 0.9223,
+      "step": 3923
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.8073044419288635,
+      "learning_rate": 0.00013872606511431803,
+      "loss": 0.941,
+      "step": 3924
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7692440152168274,
+      "learning_rate": 0.0001386983513959831,
+      "loss": 0.9574,
+      "step": 3925
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.798399806022644,
+      "learning_rate": 0.0001386706341814712,
+      "loss": 1.0244,
+      "step": 3926
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.9476202130317688,
+      "learning_rate": 0.0001386429134732865,
+      "loss": 0.9288,
+      "step": 3927
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7193227410316467,
+      "learning_rate": 0.00013861518927393333,
+      "loss": 0.9651,
+      "step": 3928
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.0366324186325073,
+      "learning_rate": 0.00013858746158591648,
+      "loss": 1.0158,
+      "step": 3929
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6833821535110474,
+      "learning_rate": 0.00013855973041174092,
+      "loss": 0.7529,
+      "step": 3930
+    },
+    {
+      "epoch": 1.48,
+      "eval_loss": 0.9561265707015991,
+      "eval_runtime": 17.7845,
+      "eval_samples_per_second": 153.448,
+      "eval_steps_per_second": 38.404,
+      "step": 3930
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7655655741691589,
+      "learning_rate": 0.00013853199575391206,
+      "loss": 0.9809,
+      "step": 3931
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.702148973941803,
+      "learning_rate": 0.00013850425761493558,
+      "loss": 0.8815,
+      "step": 3932
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.8169935345649719,
+      "learning_rate": 0.0001384765159973174,
+      "loss": 1.0864,
+      "step": 3933
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6614169478416443,
+      "learning_rate": 0.00013844877090356389,
+      "loss": 0.8591,
+      "step": 3934
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.8012004494667053,
+      "learning_rate": 0.00013842102233618158,
+      "loss": 0.9003,
+      "step": 3935
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7351922988891602,
+      "learning_rate": 0.00013839327029767744,
+      "loss": 1.0095,
+      "step": 3936
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.792578399181366,
+      "learning_rate": 0.00013836551479055875,
+      "loss": 0.8775,
+      "step": 3937
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7366359233856201,
+      "learning_rate": 0.000138337755817333,
+      "loss": 1.0098,
+      "step": 3938
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7572357654571533,
+      "learning_rate": 0.0001383099933805081,
+      "loss": 1.0691,
+      "step": 3939
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.792985200881958,
+      "learning_rate": 0.00013828222748259219,
+      "loss": 0.9662,
+      "step": 3940
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.791918933391571,
+      "learning_rate": 0.00013825445812609374,
+      "loss": 1.1858,
+      "step": 3941
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7129440903663635,
+      "learning_rate": 0.00013822668531352164,
+      "loss": 0.9818,
+      "step": 3942
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7407063841819763,
+      "learning_rate": 0.000138198909047385,
+      "loss": 1.0905,
+      "step": 3943
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7492382526397705,
+      "learning_rate": 0.00013817112933019315,
+      "loss": 1.1537,
+      "step": 3944
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7281045913696289,
+      "learning_rate": 0.0001381433461644559,
+      "loss": 0.9421,
+      "step": 3945
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7278364896774292,
+      "learning_rate": 0.00013811555955268328,
+      "loss": 0.8511,
+      "step": 3946
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.734602689743042,
+      "learning_rate": 0.00013808776949738569,
+      "loss": 0.9815,
+      "step": 3947
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.8731375932693481,
+      "learning_rate": 0.0001380599760010738,
+      "loss": 1.0916,
+      "step": 3948
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.718619704246521,
+      "learning_rate": 0.00013803217906625852,
+      "loss": 0.8546,
+      "step": 3949
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.8179419040679932,
+      "learning_rate": 0.00013800437869545125,
+      "loss": 0.9959,
+      "step": 3950
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7585844993591309,
+      "learning_rate": 0.00013797657489116348,
+      "loss": 0.9139,
+      "step": 3951
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.8072126507759094,
+      "learning_rate": 0.00013794876765590723,
+      "loss": 1.0179,
+      "step": 3952
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7540550827980042,
+      "learning_rate": 0.0001379209569921947,
+      "loss": 0.9438,
+      "step": 3953
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7537872791290283,
+      "learning_rate": 0.0001378931429025384,
+      "loss": 0.8506,
+      "step": 3954
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7491480112075806,
+      "learning_rate": 0.00013786532538945118,
+      "loss": 0.9545,
+      "step": 3955
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7633241415023804,
+      "learning_rate": 0.00013783750445544622,
+      "loss": 0.9078,
+      "step": 3956
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7701250910758972,
+      "learning_rate": 0.00013780968010303695,
+      "loss": 0.891,
+      "step": 3957
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7452569007873535,
+      "learning_rate": 0.00013778185233473717,
+      "loss": 1.0401,
+      "step": 3958
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.8602326512336731,
+      "learning_rate": 0.000137754021153061,
+      "loss": 1.0892,
+      "step": 3959
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6995726823806763,
+      "learning_rate": 0.0001377261865605227,
+      "loss": 0.8855,
+      "step": 3960
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.8985742330551147,
+      "learning_rate": 0.00013769834855963708,
+      "loss": 0.9417,
+      "step": 3961
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7605254650115967,
+      "learning_rate": 0.00013767050715291914,
+      "loss": 0.9318,
+      "step": 3962
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7218833565711975,
+      "learning_rate": 0.00013764266234288412,
+      "loss": 1.017,
+      "step": 3963
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8716464638710022,
+      "learning_rate": 0.00013761481413204773,
+      "loss": 0.9225,
+      "step": 3964
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8877992033958435,
+      "learning_rate": 0.00013758696252292584,
+      "loss": 1.04,
+      "step": 3965
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6978724598884583,
+      "learning_rate": 0.0001375591075180347,
+      "loss": 0.8764,
+      "step": 3966
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8825734853744507,
+      "learning_rate": 0.00013753124911989088,
+      "loss": 1.0465,
+      "step": 3967
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8109865188598633,
+      "learning_rate": 0.00013750338733101115,
+      "loss": 1.1471,
+      "step": 3968
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7337487936019897,
+      "learning_rate": 0.0001374755221539128,
+      "loss": 0.942,
+      "step": 3969
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7397893071174622,
+      "learning_rate": 0.00013744765359111313,
+      "loss": 1.1048,
+      "step": 3970
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.773961067199707,
+      "learning_rate": 0.00013741978164513,
+      "loss": 1.0013,
+      "step": 3971
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7440094947814941,
+      "learning_rate": 0.0001373919063184815,
+      "loss": 1.0297,
+      "step": 3972
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8709926009178162,
+      "learning_rate": 0.00013736402761368598,
+      "loss": 1.0916,
+      "step": 3973
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8232678174972534,
+      "learning_rate": 0.00013733614553326212,
+      "loss": 0.9848,
+      "step": 3974
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7813726663589478,
+      "learning_rate": 0.0001373082600797289,
+      "loss": 1.0767,
+      "step": 3975
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7393433451652527,
+      "learning_rate": 0.0001372803712556056,
+      "loss": 1.0539,
+      "step": 3976
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7628100514411926,
+      "learning_rate": 0.00013725247906341186,
+      "loss": 0.9012,
+      "step": 3977
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7220436334609985,
+      "learning_rate": 0.00013722458350566753,
+      "loss": 0.8679,
+      "step": 3978
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6969638466835022,
+      "learning_rate": 0.0001371966845848929,
+      "loss": 0.7855,
+      "step": 3979
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8976461887359619,
+      "learning_rate": 0.00013716878230360838,
+      "loss": 1.0124,
+      "step": 3980
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8752234578132629,
+      "learning_rate": 0.00013714087666433483,
+      "loss": 1.0404,
+      "step": 3981
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7346514463424683,
+      "learning_rate": 0.00013711296766959335,
+      "loss": 1.0326,
+      "step": 3982
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8979107737541199,
+      "learning_rate": 0.0001370850553219054,
+      "loss": 0.9406,
+      "step": 3983
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8028544783592224,
+      "learning_rate": 0.00013705713962379267,
+      "loss": 1.1347,
+      "step": 3984
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.9713441133499146,
+      "learning_rate": 0.00013702922057777723,
+      "loss": 0.9485,
+      "step": 3985
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6593590974807739,
+      "learning_rate": 0.00013700129818638133,
+      "loss": 0.9953,
+      "step": 3986
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6654981374740601,
+      "learning_rate": 0.00013697337245212763,
+      "loss": 0.9089,
+      "step": 3987
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.9414458870887756,
+      "learning_rate": 0.0001369454433775391,
+      "loss": 1.244,
+      "step": 3988
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7171873450279236,
+      "learning_rate": 0.00013691751096513892,
+      "loss": 0.8293,
+      "step": 3989
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8503627181053162,
+      "learning_rate": 0.00013688957521745067,
+      "loss": 1.1873,
+      "step": 3990
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8794218897819519,
+      "learning_rate": 0.00013686163613699814,
+      "loss": 1.0791,
+      "step": 3991
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7562376856803894,
+      "learning_rate": 0.00013683369372630554,
+      "loss": 0.9468,
+      "step": 3992
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.747326135635376,
+      "learning_rate": 0.00013680574798789722,
+      "loss": 1.0067,
+      "step": 3993
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6801347732543945,
+      "learning_rate": 0.00013677779892429798,
+      "loss": 0.8768,
+      "step": 3994
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.9982435703277588,
+      "learning_rate": 0.00013674984653803284,
+      "loss": 0.9128,
+      "step": 3995
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7373642921447754,
+      "learning_rate": 0.00013672189083162715,
+      "loss": 0.9162,
+      "step": 3996
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6660494804382324,
+      "learning_rate": 0.0001366939318076065,
+      "loss": 0.9756,
+      "step": 3997
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7656930685043335,
+      "learning_rate": 0.00013666596946849694,
+      "loss": 0.9965,
+      "step": 3998
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8281304836273193,
+      "learning_rate": 0.00013663800381682464,
+      "loss": 0.9022,
+      "step": 3999
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7213367223739624,
+      "learning_rate": 0.0001366100348551161,
+      "loss": 1.0319,
+      "step": 4000
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8114079833030701,
+      "learning_rate": 0.0001365820625858982,
+      "loss": 1.0008,
+      "step": 4001
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7172978520393372,
+      "learning_rate": 0.0001365540870116981,
+      "loss": 0.9841,
+      "step": 4002
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7489575743675232,
+      "learning_rate": 0.00013652610813504317,
+      "loss": 1.0965,
+      "step": 4003
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8985177874565125,
+      "learning_rate": 0.0001364981259584612,
+      "loss": 1.0878,
+      "step": 4004
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7306704521179199,
+      "learning_rate": 0.0001364701404844802,
+      "loss": 1.0132,
+      "step": 4005
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8897285461425781,
+      "learning_rate": 0.0001364421517156285,
+      "loss": 0.8861,
+      "step": 4006
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7293336987495422,
+      "learning_rate": 0.0001364141596544347,
+      "loss": 1.024,
+      "step": 4007
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6985815167427063,
+      "learning_rate": 0.00013638616430342777,
+      "loss": 0.902,
+      "step": 4008
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7957209348678589,
+      "learning_rate": 0.00013635816566513696,
+      "loss": 1.1111,
+      "step": 4009
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.756161630153656,
+      "learning_rate": 0.00013633016374209168,
+      "loss": 0.9724,
+      "step": 4010
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8828256130218506,
+      "learning_rate": 0.0001363021585368218,
+      "loss": 0.8165,
+      "step": 4011
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7583345174789429,
+      "learning_rate": 0.0001362741500518575,
+      "loss": 1.1074,
+      "step": 4012
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8206169605255127,
+      "learning_rate": 0.00013624613828972908,
+      "loss": 0.9496,
+      "step": 4013
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6681649088859558,
+      "learning_rate": 0.0001362181232529673,
+      "loss": 0.9365,
+      "step": 4014
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7639154195785522,
+      "learning_rate": 0.0001361901049441032,
+      "loss": 0.8674,
+      "step": 4015
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8410853147506714,
+      "learning_rate": 0.00013616208336566794,
+      "loss": 1.0192,
+      "step": 4016
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7662779092788696,
+      "learning_rate": 0.00013613405852019324,
+      "loss": 0.8263,
+      "step": 4017
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7114266753196716,
+      "learning_rate": 0.00013610603041021096,
+      "loss": 0.8868,
+      "step": 4018
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7133528590202332,
+      "learning_rate": 0.00013607799903825324,
+      "loss": 1.0178,
+      "step": 4019
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7552312612533569,
+      "learning_rate": 0.0001360499644068526,
+      "loss": 0.9497,
+      "step": 4020
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8263150453567505,
+      "learning_rate": 0.00013602192651854174,
+      "loss": 1.154,
+      "step": 4021
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7287723422050476,
+      "learning_rate": 0.00013599388537585382,
+      "loss": 0.957,
+      "step": 4022
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8263999819755554,
+      "learning_rate": 0.00013596584098132217,
+      "loss": 1.1154,
+      "step": 4023
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7667343020439148,
+      "learning_rate": 0.00013593779333748038,
+      "loss": 0.9197,
+      "step": 4024
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7314568758010864,
+      "learning_rate": 0.00013590974244686248,
+      "loss": 0.9844,
+      "step": 4025
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7331525683403015,
+      "learning_rate": 0.00013588168831200263,
+      "loss": 0.999,
+      "step": 4026
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7766103148460388,
+      "learning_rate": 0.00013585363093543542,
+      "loss": 0.99,
+      "step": 4027
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7825350165367126,
+      "learning_rate": 0.00013582557031969567,
+      "loss": 0.9469,
+      "step": 4028
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7960454821586609,
+      "learning_rate": 0.00013579750646731847,
+      "loss": 0.965,
+      "step": 4029
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7917805314064026,
+      "learning_rate": 0.00013576943938083927,
+      "loss": 0.982,
+      "step": 4030
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7845766544342041,
+      "learning_rate": 0.0001357413690627937,
+      "loss": 0.9755,
+      "step": 4031
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7146531343460083,
+      "learning_rate": 0.00013571329551571783,
+      "loss": 0.8617,
+      "step": 4032
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7622401118278503,
+      "learning_rate": 0.00013568521874214796,
+      "loss": 0.7809,
+      "step": 4033
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7676883339881897,
+      "learning_rate": 0.00013565713874462058,
+      "loss": 1.0348,
+      "step": 4034
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.9790372252464294,
+      "learning_rate": 0.0001356290555256726,
+      "loss": 1.0473,
+      "step": 4035
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.9019134640693665,
+      "learning_rate": 0.00013560096908784119,
+      "loss": 1.0282,
+      "step": 4036
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8120095133781433,
+      "learning_rate": 0.0001355728794336638,
+      "loss": 1.0385,
+      "step": 4037
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8038890957832336,
+      "learning_rate": 0.00013554478656567818,
+      "loss": 1.0467,
+      "step": 4038
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6963586807250977,
+      "learning_rate": 0.00013551669048642232,
+      "loss": 0.993,
+      "step": 4039
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.797791063785553,
+      "learning_rate": 0.00013548859119843457,
+      "loss": 1.0253,
+      "step": 4040
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7721526026725769,
+      "learning_rate": 0.00013546048870425356,
+      "loss": 1.0582,
+      "step": 4041
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8010839223861694,
+      "learning_rate": 0.0001354323830064182,
+      "loss": 1.0061,
+      "step": 4042
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.723960816860199,
+      "learning_rate": 0.00013540427410746762,
+      "loss": 0.8673,
+      "step": 4043
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7292559742927551,
+      "learning_rate": 0.0001353761620099413,
+      "loss": 1.277,
+      "step": 4044
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7340478897094727,
+      "learning_rate": 0.0001353480467163791,
+      "loss": 0.9326,
+      "step": 4045
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6826416850090027,
+      "learning_rate": 0.00013531992822932097,
+      "loss": 0.96,
+      "step": 4046
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.9213247299194336,
+      "learning_rate": 0.00013529180655130735,
+      "loss": 0.8988,
+      "step": 4047
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.8114209175109863,
+      "learning_rate": 0.0001352636816848788,
+      "loss": 1.0367,
+      "step": 4048
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7611256241798401,
+      "learning_rate": 0.0001352355536325763,
+      "loss": 0.9916,
+      "step": 4049
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7742871046066284,
+      "learning_rate": 0.00013520742239694103,
+      "loss": 0.9578,
+      "step": 4050
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7500746846199036,
+      "learning_rate": 0.00013517928798051444,
+      "loss": 1.0355,
+      "step": 4051
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.9206206202507019,
+      "learning_rate": 0.0001351511503858384,
+      "loss": 1.1098,
+      "step": 4052
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7646648287773132,
+      "learning_rate": 0.00013512300961545496,
+      "loss": 1.1684,
+      "step": 4053
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7009553909301758,
+      "learning_rate": 0.00013509486567190646,
+      "loss": 0.8637,
+      "step": 4054
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7195073366165161,
+      "learning_rate": 0.00013506671855773554,
+      "loss": 0.8753,
+      "step": 4055
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7679314613342285,
+      "learning_rate": 0.00013503856827548513,
+      "loss": 1.0202,
+      "step": 4056
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.9173232316970825,
+      "learning_rate": 0.0001350104148276985,
+      "loss": 0.911,
+      "step": 4057
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.8022065758705139,
+      "learning_rate": 0.0001349822582169191,
+      "loss": 1.0949,
+      "step": 4058
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7112104296684265,
+      "learning_rate": 0.00013495409844569076,
+      "loss": 1.1556,
+      "step": 4059
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7301331162452698,
+      "learning_rate": 0.00013492593551655752,
+      "loss": 0.8969,
+      "step": 4060
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.8814001083374023,
+      "learning_rate": 0.00013489776943206375,
+      "loss": 1.1629,
+      "step": 4061
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6501629948616028,
+      "learning_rate": 0.0001348696001947541,
+      "loss": 0.8705,
+      "step": 4062
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.78933185338974,
+      "learning_rate": 0.00013484142780717355,
+      "loss": 1.1239,
+      "step": 4063
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.8425937294960022,
+      "learning_rate": 0.0001348132522718672,
+      "loss": 1.0504,
+      "step": 4064
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.729695737361908,
+      "learning_rate": 0.00013478507359138066,
+      "loss": 0.8882,
+      "step": 4065
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7847967147827148,
+      "learning_rate": 0.00013475689176825967,
+      "loss": 0.9521,
+      "step": 4066
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.8310196995735168,
+      "learning_rate": 0.00013472870680505028,
+      "loss": 0.8892,
+      "step": 4067
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7683903574943542,
+      "learning_rate": 0.00013470051870429888,
+      "loss": 1.0349,
+      "step": 4068
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.8151059746742249,
+      "learning_rate": 0.00013467232746855213,
+      "loss": 0.861,
+      "step": 4069
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.9114918112754822,
+      "learning_rate": 0.00013464413310035686,
+      "loss": 0.9156,
+      "step": 4070
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.7894443869590759,
+      "learning_rate": 0.0001346159356022603,
+      "loss": 1.0753,
+      "step": 4071
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.7362051010131836,
+      "learning_rate": 0.00013458773497680997,
+      "loss": 0.8435,
+      "step": 4072
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.7690173387527466,
+      "learning_rate": 0.00013455953122655366,
+      "loss": 1.1125,
+      "step": 4073
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.7989892959594727,
+      "learning_rate": 0.00013453132435403936,
+      "loss": 0.8798,
+      "step": 4074
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.709227979183197,
+      "learning_rate": 0.00013450311436181538,
+      "loss": 0.8941,
+      "step": 4075
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.8580532073974609,
+      "learning_rate": 0.00013447490125243042,
+      "loss": 0.9188,
+      "step": 4076
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.7872312068939209,
+      "learning_rate": 0.0001344466850284333,
+      "loss": 0.8544,
+      "step": 4077
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.7518987655639648,
+      "learning_rate": 0.00013441846569237323,
+      "loss": 0.8517,
+      "step": 4078
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.7955414056777954,
+      "learning_rate": 0.00013439024324679968,
+      "loss": 0.9751,
+      "step": 4079
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.8094600439071655,
+      "learning_rate": 0.00013436201769426235,
+      "loss": 1.0051,
+      "step": 4080
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.8085797429084778,
+      "learning_rate": 0.00013433378903731125,
+      "loss": 1.0915,
+      "step": 4081
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.8019757270812988,
+      "learning_rate": 0.0001343055572784968,
+      "loss": 0.9156,
+      "step": 4082
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.7250171899795532,
+      "learning_rate": 0.0001342773224203694,
+      "loss": 0.883,
+      "step": 4083
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.8753831386566162,
+      "learning_rate": 0.00013424908446548004,
+      "loss": 0.983,
+      "step": 4084
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.7171947956085205,
+      "learning_rate": 0.00013422084341637979,
+      "loss": 0.9899,
+      "step": 4085
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.7305735349655151,
+      "learning_rate": 0.00013419259927562011,
+      "loss": 0.8964,
+      "step": 4086
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.8032180070877075,
+      "learning_rate": 0.0001341643520457527,
+      "loss": 0.9373,
+      "step": 4087
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.7876158356666565,
+      "learning_rate": 0.0001341361017293295,
+      "loss": 0.967,
+      "step": 4088
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.679061770439148,
+      "learning_rate": 0.00013410784832890277,
+      "loss": 0.7776,
+      "step": 4089
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.8126422166824341,
+      "learning_rate": 0.0001340795918470251,
+      "loss": 0.9163,
+      "step": 4090
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.8398410677909851,
+      "learning_rate": 0.0001340513322862492,
+      "loss": 0.9656,
+      "step": 4091
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.8889121413230896,
+      "learning_rate": 0.0001340230696491283,
+      "loss": 0.9531,
+      "step": 4092
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.7200760245323181,
+      "learning_rate": 0.00013399480393821569,
+      "loss": 0.9501,
+      "step": 4093
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.826648473739624,
+      "learning_rate": 0.00013396653515606498,
+      "loss": 1.053,
+      "step": 4094
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.7926040887832642,
+      "learning_rate": 0.0001339382633052302,
+      "loss": 1.0978,
+      "step": 4095
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.788530170917511,
+      "learning_rate": 0.00013390998838826542,
+      "loss": 1.0361,
+      "step": 4096
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.7253285050392151,
+      "learning_rate": 0.0001338817104077252,
+      "loss": 0.6804,
+      "step": 4097
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.8385868668556213,
+      "learning_rate": 0.0001338534293661643,
+      "loss": 0.8149,
+      "step": 4098
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6631655693054199,
+      "learning_rate": 0.00013382514526613773,
+      "loss": 0.9524,
+      "step": 4099
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.7763095498085022,
+      "learning_rate": 0.0001337968581102008,
+      "loss": 0.9904,
+      "step": 4100
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6638621687889099,
+      "learning_rate": 0.0001337685679009091,
+      "loss": 0.9542,
+      "step": 4101
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.7378938794136047,
+      "learning_rate": 0.00013374027464081847,
+      "loss": 0.9329,
+      "step": 4102
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.7555968761444092,
+      "learning_rate": 0.00013371197833248507,
+      "loss": 0.8773,
+      "step": 4103
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.8379522562026978,
+      "learning_rate": 0.00013368367897846533,
+      "loss": 1.0052,
+      "step": 4104
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.990310788154602,
+      "learning_rate": 0.0001336553765813159,
+      "loss": 1.2306,
+      "step": 4105
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6684477925300598,
+      "learning_rate": 0.00013362707114359374,
+      "loss": 0.8162,
+      "step": 4106
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.7488822340965271,
+      "learning_rate": 0.0001335987626678561,
+      "loss": 1.0563,
+      "step": 4107
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.8354619145393372,
+      "learning_rate": 0.0001335704511566605,
+      "loss": 1.0533,
+      "step": 4108
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.829838752746582,
+      "learning_rate": 0.00013354213661256472,
+      "loss": 0.9526,
+      "step": 4109
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.7552241683006287,
+      "learning_rate": 0.0001335138190381268,
+      "loss": 0.8768,
+      "step": 4110
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6395108103752136,
+      "learning_rate": 0.00013348549843590504,
+      "loss": 0.9264,
+      "step": 4111
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.7724140882492065,
+      "learning_rate": 0.00013345717480845817,
+      "loss": 1.0384,
+      "step": 4112
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.7144059538841248,
+      "learning_rate": 0.00013342884815834493,
+      "loss": 1.069,
+      "step": 4113
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.7421502470970154,
+      "learning_rate": 0.00013340051848812458,
+      "loss": 1.0402,
+      "step": 4114
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.7333446741104126,
+      "learning_rate": 0.00013337218580035647,
+      "loss": 1.0038,
+      "step": 4115
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.8908722400665283,
+      "learning_rate": 0.00013334385009760032,
+      "loss": 1.1055,
+      "step": 4116
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6802178621292114,
+      "learning_rate": 0.0001333155113824161,
+      "loss": 0.8414,
+      "step": 4117
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.8234414458274841,
+      "learning_rate": 0.00013328716965736408,
+      "loss": 1.1654,
+      "step": 4118
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.749435305595398,
+      "learning_rate": 0.00013325882492500477,
+      "loss": 1.0812,
+      "step": 4119
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.7330403923988342,
+      "learning_rate": 0.0001332304771878989,
+      "loss": 0.9709,
+      "step": 4120
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.794902503490448,
+      "learning_rate": 0.00013320212644860757,
+      "loss": 0.9516,
+      "step": 4121
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.7441633939743042,
+      "learning_rate": 0.00013317377270969212,
+      "loss": 1.0171,
+      "step": 4122
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8326237201690674,
+      "learning_rate": 0.00013314541597371412,
+      "loss": 0.9255,
+      "step": 4123
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8376749753952026,
+      "learning_rate": 0.0001331170562432355,
+      "loss": 1.0207,
+      "step": 4124
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.7114330530166626,
+      "learning_rate": 0.00013308869352081834,
+      "loss": 0.79,
+      "step": 4125
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.7856300473213196,
+      "learning_rate": 0.00013306032780902502,
+      "loss": 0.9169,
+      "step": 4126
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.7197442650794983,
+      "learning_rate": 0.00013303195911041832,
+      "loss": 0.9439,
+      "step": 4127
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.7740024924278259,
+      "learning_rate": 0.0001330035874275611,
+      "loss": 1.0669,
+      "step": 4128
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8333525061607361,
+      "learning_rate": 0.00013297521276301664,
+      "loss": 1.1763,
+      "step": 4129
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8492178320884705,
+      "learning_rate": 0.00013294683511934845,
+      "loss": 0.9247,
+      "step": 4130
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8128172755241394,
+      "learning_rate": 0.0001329184544991202,
+      "loss": 1.0977,
+      "step": 4131
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.7545269131660461,
+      "learning_rate": 0.000132890070904896,
+      "loss": 0.8532,
+      "step": 4132
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.7314281463623047,
+      "learning_rate": 0.0001328616843392401,
+      "loss": 0.8558,
+      "step": 4133
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6871530413627625,
+      "learning_rate": 0.0001328332948047171,
+      "loss": 0.6922,
+      "step": 4134
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.7285220623016357,
+      "learning_rate": 0.0001328049023038918,
+      "loss": 0.9591,
+      "step": 4135
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8840282559394836,
+      "learning_rate": 0.00013277650683932933,
+      "loss": 0.9727,
+      "step": 4136
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.7239083647727966,
+      "learning_rate": 0.00013274810841359502,
+      "loss": 0.9342,
+      "step": 4137
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.9012027978897095,
+      "learning_rate": 0.00013271970702925456,
+      "loss": 1.0467,
+      "step": 4138
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.7457629442214966,
+      "learning_rate": 0.00013269130268887382,
+      "loss": 0.9247,
+      "step": 4139
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.7632551193237305,
+      "learning_rate": 0.000132662895395019,
+      "loss": 1.1305,
+      "step": 4140
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8433321118354797,
+      "learning_rate": 0.00013263448515025647,
+      "loss": 1.1017,
+      "step": 4141
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.7884944081306458,
+      "learning_rate": 0.00013260607195715303,
+      "loss": 0.9197,
+      "step": 4142
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8757911920547485,
+      "learning_rate": 0.0001325776558182756,
+      "loss": 1.1378,
+      "step": 4143
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.985319197177887,
+      "learning_rate": 0.0001325492367361914,
+      "loss": 1.2242,
+      "step": 4144
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6358724236488342,
+      "learning_rate": 0.00013252081471346795,
+      "loss": 0.9055,
+      "step": 4145
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.7014467716217041,
+      "learning_rate": 0.00013249238975267304,
+      "loss": 1.0029,
+      "step": 4146
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8960198760032654,
+      "learning_rate": 0.00013246396185637467,
+      "loss": 1.0953,
+      "step": 4147
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7448418140411377,
+      "learning_rate": 0.00013243553102714118,
+      "loss": 0.8569,
+      "step": 4148
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.8383636474609375,
+      "learning_rate": 0.0001324070972675411,
+      "loss": 1.1298,
+      "step": 4149
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6508424282073975,
+      "learning_rate": 0.00013237866058014326,
+      "loss": 0.7927,
+      "step": 4150
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7935487627983093,
+      "learning_rate": 0.00013235022096751676,
+      "loss": 0.9801,
+      "step": 4151
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7764027118682861,
+      "learning_rate": 0.000132321778432231,
+      "loss": 1.1115,
+      "step": 4152
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.8405836820602417,
+      "learning_rate": 0.00013229333297685552,
+      "loss": 0.9208,
+      "step": 4153
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7989316582679749,
+      "learning_rate": 0.00013226488460396028,
+      "loss": 1.0406,
+      "step": 4154
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7648982405662537,
+      "learning_rate": 0.00013223643331611537,
+      "loss": 0.96,
+      "step": 4155
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7140777111053467,
+      "learning_rate": 0.00013220797911589127,
+      "loss": 0.8974,
+      "step": 4156
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7468279004096985,
+      "learning_rate": 0.00013217952200585858,
+      "loss": 0.8749,
+      "step": 4157
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7843912839889526,
+      "learning_rate": 0.0001321510619885883,
+      "loss": 0.9856,
+      "step": 4158
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7949358820915222,
+      "learning_rate": 0.0001321225990666516,
+      "loss": 1.0912,
+      "step": 4159
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.824992835521698,
+      "learning_rate": 0.00013209413324261998,
+      "loss": 0.9953,
+      "step": 4160
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7797181606292725,
+      "learning_rate": 0.00013206566451906508,
+      "loss": 0.9361,
+      "step": 4161
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7983461022377014,
+      "learning_rate": 0.000132037192898559,
+      "loss": 0.9637,
+      "step": 4162
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7876800894737244,
+      "learning_rate": 0.0001320087183836739,
+      "loss": 0.9159,
+      "step": 4163
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7618224024772644,
+      "learning_rate": 0.00013198024097698237,
+      "loss": 0.7168,
+      "step": 4164
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6901776790618896,
+      "learning_rate": 0.00013195176068105712,
+      "loss": 0.8032,
+      "step": 4165
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.8126145601272583,
+      "learning_rate": 0.00013192327749847116,
+      "loss": 0.9431,
+      "step": 4166
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7916374802589417,
+      "learning_rate": 0.00013189479143179789,
+      "loss": 0.9993,
+      "step": 4167
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.8331419825553894,
+      "learning_rate": 0.0001318663024836108,
+      "loss": 0.9779,
+      "step": 4168
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.8191587328910828,
+      "learning_rate": 0.00013183781065648364,
+      "loss": 1.0442,
+      "step": 4169
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7728171944618225,
+      "learning_rate": 0.0001318093159529906,
+      "loss": 0.9568,
+      "step": 4170
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7042739987373352,
+      "learning_rate": 0.00013178081837570598,
+      "loss": 0.7985,
+      "step": 4171
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7546452879905701,
+      "learning_rate": 0.0001317523179272043,
+      "loss": 0.9792,
+      "step": 4172
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.9695934057235718,
+      "learning_rate": 0.00013172381461006057,
+      "loss": 0.9767,
+      "step": 4173
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.9317048788070679,
+      "learning_rate": 0.00013169530842684974,
+      "loss": 1.0918,
+      "step": 4174
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6831107139587402,
+      "learning_rate": 0.00013166679938014726,
+      "loss": 0.7801,
+      "step": 4175
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.8045737147331238,
+      "learning_rate": 0.00013163828747252878,
+      "loss": 0.9696,
+      "step": 4176
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.7625781297683716,
+      "learning_rate": 0.00013160977270657012,
+      "loss": 0.9746,
+      "step": 4177
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.7749671936035156,
+      "learning_rate": 0.0001315812550848475,
+      "loss": 0.8871,
+      "step": 4178
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.8254308104515076,
+      "learning_rate": 0.00013155273460993728,
+      "loss": 0.9818,
+      "step": 4179
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.8079273700714111,
+      "learning_rate": 0.00013152421128441617,
+      "loss": 1.0846,
+      "step": 4180
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.8321089744567871,
+      "learning_rate": 0.00013149568511086103,
+      "loss": 1.1869,
+      "step": 4181
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.8857249617576599,
+      "learning_rate": 0.00013146715609184907,
+      "loss": 1.1094,
+      "step": 4182
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.8076190948486328,
+      "learning_rate": 0.00013143862422995776,
+      "loss": 1.0152,
+      "step": 4183
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.800880491733551,
+      "learning_rate": 0.00013141008952776475,
+      "loss": 0.9568,
+      "step": 4184
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.7508581876754761,
+      "learning_rate": 0.00013138155198784795,
+      "loss": 0.8768,
+      "step": 4185
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.8762493133544922,
+      "learning_rate": 0.00013135301161278564,
+      "loss": 1.1144,
+      "step": 4186
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.7529940605163574,
+      "learning_rate": 0.00013132446840515623,
+      "loss": 0.8161,
+      "step": 4187
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6504919528961182,
+      "learning_rate": 0.00013129592236753846,
+      "loss": 0.8698,
+      "step": 4188
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.801974892616272,
+      "learning_rate": 0.00013126737350251135,
+      "loss": 0.9568,
+      "step": 4189
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.737905740737915,
+      "learning_rate": 0.00013123882181265403,
+      "loss": 0.9602,
+      "step": 4190
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.8118032813072205,
+      "learning_rate": 0.00013121026730054605,
+      "loss": 0.9315,
+      "step": 4191
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.8885903358459473,
+      "learning_rate": 0.00013118170996876714,
+      "loss": 0.9292,
+      "step": 4192
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.7553448677062988,
+      "learning_rate": 0.00013115314981989725,
+      "loss": 1.0179,
+      "step": 4193
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.8398468494415283,
+      "learning_rate": 0.00013112458685651668,
+      "loss": 0.9042,
+      "step": 4194
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.761009693145752,
+      "learning_rate": 0.00013109602108120593,
+      "loss": 0.9415,
+      "step": 4195
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.7048000693321228,
+      "learning_rate": 0.0001310674524965457,
+      "loss": 0.8545,
+      "step": 4196
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.7960367202758789,
+      "learning_rate": 0.0001310388811051171,
+      "loss": 1.0841,
+      "step": 4197
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.8713595867156982,
+      "learning_rate": 0.0001310103069095013,
+      "loss": 1.1152,
+      "step": 4198
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.7115532159805298,
+      "learning_rate": 0.00013098172991227985,
+      "loss": 1.083,
+      "step": 4199
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.7147960662841797,
+      "learning_rate": 0.00013095315011603454,
+      "loss": 0.9406,
+      "step": 4200
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8112908005714417,
+      "learning_rate": 0.00013092456752334734,
+      "loss": 1.0151,
+      "step": 4201
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8009018898010254,
+      "learning_rate": 0.00013089598213680058,
+      "loss": 1.0362,
+      "step": 4202
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8003485798835754,
+      "learning_rate": 0.00013086739395897677,
+      "loss": 0.7262,
+      "step": 4203
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.7411230206489563,
+      "learning_rate": 0.00013083880299245867,
+      "loss": 0.9487,
+      "step": 4204
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.7706307768821716,
+      "learning_rate": 0.00013081020923982935,
+      "loss": 0.8444,
+      "step": 4205
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8083503842353821,
+      "learning_rate": 0.00013078161270367207,
+      "loss": 0.9118,
+      "step": 4206
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.750612199306488,
+      "learning_rate": 0.00013075301338657036,
+      "loss": 1.0934,
+      "step": 4207
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8122290968894958,
+      "learning_rate": 0.00013072441129110803,
+      "loss": 1.0062,
+      "step": 4208
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6777164340019226,
+      "learning_rate": 0.00013069580641986907,
+      "loss": 0.9177,
+      "step": 4209
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.9212057590484619,
+      "learning_rate": 0.00013066719877543782,
+      "loss": 1.1174,
+      "step": 4210
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.75261390209198,
+      "learning_rate": 0.00013063858836039876,
+      "loss": 0.9974,
+      "step": 4211
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8800492286682129,
+      "learning_rate": 0.00013060997517733676,
+      "loss": 0.9127,
+      "step": 4212
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.7661362290382385,
+      "learning_rate": 0.0001305813592288368,
+      "loss": 0.9915,
+      "step": 4213
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8467798829078674,
+      "learning_rate": 0.00013055274051748417,
+      "loss": 1.2046,
+      "step": 4214
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.7771646976470947,
+      "learning_rate": 0.00013052411904586445,
+      "loss": 0.9882,
+      "step": 4215
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8530020117759705,
+      "learning_rate": 0.0001304954948165634,
+      "loss": 1.0174,
+      "step": 4216
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.7605085968971252,
+      "learning_rate": 0.00013046686783216703,
+      "loss": 1.0307,
+      "step": 4217
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6596463322639465,
+      "learning_rate": 0.0001304382380952617,
+      "loss": 0.7798,
+      "step": 4218
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.74962317943573,
+      "learning_rate": 0.0001304096056084339,
+      "loss": 0.8796,
+      "step": 4219
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.7627604603767395,
+      "learning_rate": 0.00013038097037427037,
+      "loss": 0.9478,
+      "step": 4220
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.7824596762657166,
+      "learning_rate": 0.00013035233239535817,
+      "loss": 1.0287,
+      "step": 4221
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8189258575439453,
+      "learning_rate": 0.0001303236916742846,
+      "loss": 0.9816,
+      "step": 4222
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.7686831951141357,
+      "learning_rate": 0.00013029504821363724,
+      "loss": 0.9055,
+      "step": 4223
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8042637705802917,
+      "learning_rate": 0.00013026640201600377,
+      "loss": 0.9434,
+      "step": 4224
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8106065988540649,
+      "learning_rate": 0.00013023775308397222,
+      "loss": 1.1275,
+      "step": 4225
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.8299038410186768,
+      "learning_rate": 0.00013020910142013092,
+      "loss": 1.0951,
+      "step": 4226
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.8075307607650757,
+      "learning_rate": 0.00013018044702706836,
+      "loss": 0.931,
+      "step": 4227
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7008748650550842,
+      "learning_rate": 0.00013015178990737324,
+      "loss": 1.0001,
+      "step": 4228
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.8621682524681091,
+      "learning_rate": 0.00013012313006363467,
+      "loss": 0.9809,
+      "step": 4229
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7828312516212463,
+      "learning_rate": 0.00013009446749844182,
+      "loss": 0.8595,
+      "step": 4230
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7661519646644592,
+      "learning_rate": 0.00013006580221438423,
+      "loss": 0.7974,
+      "step": 4231
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7845397591590881,
+      "learning_rate": 0.00013003713421405166,
+      "loss": 0.9607,
+      "step": 4232
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7443758249282837,
+      "learning_rate": 0.0001300084635000341,
+      "loss": 1.0433,
+      "step": 4233
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7668888568878174,
+      "learning_rate": 0.00012997979007492173,
+      "loss": 1.0791,
+      "step": 4234
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7785104513168335,
+      "learning_rate": 0.0001299511139413051,
+      "loss": 0.881,
+      "step": 4235
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.8156974911689758,
+      "learning_rate": 0.00012992243510177487,
+      "loss": 1.1101,
+      "step": 4236
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7256063222885132,
+      "learning_rate": 0.0001298937535589221,
+      "loss": 0.9809,
+      "step": 4237
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7718733549118042,
+      "learning_rate": 0.0001298650693153379,
+      "loss": 0.9407,
+      "step": 4238
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.72328120470047,
+      "learning_rate": 0.00012983638237361382,
+      "loss": 0.9191,
+      "step": 4239
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.978679895401001,
+      "learning_rate": 0.00012980769273634156,
+      "loss": 1.1024,
+      "step": 4240
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.767125129699707,
+      "learning_rate": 0.00012977900040611295,
+      "loss": 1.1594,
+      "step": 4241
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7211596965789795,
+      "learning_rate": 0.00012975030538552032,
+      "loss": 0.8163,
+      "step": 4242
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.8155632019042969,
+      "learning_rate": 0.00012972160767715603,
+      "loss": 1.1148,
+      "step": 4243
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7436785101890564,
+      "learning_rate": 0.00012969290728361273,
+      "loss": 1.105,
+      "step": 4244
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7024710178375244,
+      "learning_rate": 0.00012966420420748345,
+      "loss": 0.6854,
+      "step": 4245
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7749850153923035,
+      "learning_rate": 0.00012963549845136123,
+      "loss": 0.9937,
+      "step": 4246
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6667187213897705,
+      "learning_rate": 0.00012960679001783952,
+      "loss": 0.7582,
+      "step": 4247
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7609478831291199,
+      "learning_rate": 0.000129578078909512,
+      "loss": 0.8994,
+      "step": 4248
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6955171823501587,
+      "learning_rate": 0.00012954936512897248,
+      "loss": 1.1101,
+      "step": 4249
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6980825066566467,
+      "learning_rate": 0.00012952064867881514,
+      "loss": 1.1454,
+      "step": 4250
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7012463212013245,
+      "learning_rate": 0.00012949192956163437,
+      "loss": 0.8899,
+      "step": 4251
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8292000889778137,
+      "learning_rate": 0.0001294632077800247,
+      "loss": 0.9554,
+      "step": 4252
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.7803111672401428,
+      "learning_rate": 0.00012943448333658107,
+      "loss": 0.8801,
+      "step": 4253
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.9052737355232239,
+      "learning_rate": 0.00012940575623389847,
+      "loss": 0.9664,
+      "step": 4254
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.7098575234413147,
+      "learning_rate": 0.00012937702647457234,
+      "loss": 1.0257,
+      "step": 4255
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.682171106338501,
+      "learning_rate": 0.00012934829406119818,
+      "loss": 0.897,
+      "step": 4256
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.7598903775215149,
+      "learning_rate": 0.00012931955899637181,
+      "loss": 0.9512,
+      "step": 4257
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.7627248167991638,
+      "learning_rate": 0.00012929082128268932,
+      "loss": 0.9523,
+      "step": 4258
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.7255799174308777,
+      "learning_rate": 0.000129262080922747,
+      "loss": 1.0527,
+      "step": 4259
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.7689884305000305,
+      "learning_rate": 0.0001292333379191413,
+      "loss": 1.1442,
+      "step": 4260
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8066941499710083,
+      "learning_rate": 0.00012920459227446903,
+      "loss": 0.9867,
+      "step": 4261
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.775963544845581,
+      "learning_rate": 0.00012917584399132722,
+      "loss": 1.027,
+      "step": 4262
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.7826582789421082,
+      "learning_rate": 0.0001291470930723131,
+      "loss": 0.978,
+      "step": 4263
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.7322339415550232,
+      "learning_rate": 0.00012911833952002416,
+      "loss": 1.0189,
+      "step": 4264
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.7850020527839661,
+      "learning_rate": 0.00012908958333705808,
+      "loss": 1.1508,
+      "step": 4265
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.7325538992881775,
+      "learning_rate": 0.00012906082452601288,
+      "loss": 0.8523,
+      "step": 4266
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8257113099098206,
+      "learning_rate": 0.00012903206308948674,
+      "loss": 0.9526,
+      "step": 4267
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.7576756477355957,
+      "learning_rate": 0.00012900329903007805,
+      "loss": 0.9564,
+      "step": 4268
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.7235482931137085,
+      "learning_rate": 0.00012897453235038552,
+      "loss": 1.044,
+      "step": 4269
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.7639951705932617,
+      "learning_rate": 0.00012894576305300806,
+      "loss": 1.1088,
+      "step": 4270
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8137654066085815,
+      "learning_rate": 0.00012891699114054477,
+      "loss": 1.0389,
+      "step": 4271
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.7565323710441589,
+      "learning_rate": 0.00012888821661559508,
+      "loss": 0.9217,
+      "step": 4272
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6858435869216919,
+      "learning_rate": 0.00012885943948075856,
+      "loss": 0.9392,
+      "step": 4273
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.820617139339447,
+      "learning_rate": 0.0001288306597386351,
+      "loss": 1.0266,
+      "step": 4274
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8873162269592285,
+      "learning_rate": 0.00012880187739182478,
+      "loss": 0.8399,
+      "step": 4275
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.7443687915802002,
+      "learning_rate": 0.00012877309244292787,
+      "loss": 0.8388,
+      "step": 4276
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.858851969242096,
+      "learning_rate": 0.00012874430489454502,
+      "loss": 0.9356,
+      "step": 4277
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.850151002407074,
+      "learning_rate": 0.00012871551474927698,
+      "loss": 1.021,
+      "step": 4278
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7548375725746155,
+      "learning_rate": 0.00012868672200972475,
+      "loss": 0.7736,
+      "step": 4279
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6958070397377014,
+      "learning_rate": 0.0001286579266784896,
+      "loss": 0.9859,
+      "step": 4280
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7235026359558105,
+      "learning_rate": 0.00012862912875817305,
+      "loss": 0.9848,
+      "step": 4281
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7634601593017578,
+      "learning_rate": 0.00012860032825137677,
+      "loss": 0.9571,
+      "step": 4282
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7626282572746277,
+      "learning_rate": 0.00012857152516070283,
+      "loss": 1.0399,
+      "step": 4283
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7731013298034668,
+      "learning_rate": 0.00012854271948875332,
+      "loss": 0.9295,
+      "step": 4284
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.8193756341934204,
+      "learning_rate": 0.00012851391123813074,
+      "loss": 1.1184,
+      "step": 4285
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7630685567855835,
+      "learning_rate": 0.0001284851004114377,
+      "loss": 0.9509,
+      "step": 4286
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7839036583900452,
+      "learning_rate": 0.00012845628701127712,
+      "loss": 0.8557,
+      "step": 4287
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7505083084106445,
+      "learning_rate": 0.00012842747104025215,
+      "loss": 0.8715,
+      "step": 4288
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.8192002773284912,
+      "learning_rate": 0.00012839865250096613,
+      "loss": 1.1744,
+      "step": 4289
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7832188606262207,
+      "learning_rate": 0.00012836983139602264,
+      "loss": 1.0416,
+      "step": 4290
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.8175313472747803,
+      "learning_rate": 0.00012834100772802552,
+      "loss": 1.043,
+      "step": 4291
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6733172535896301,
+      "learning_rate": 0.00012831218149957877,
+      "loss": 1.066,
+      "step": 4292
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.775862991809845,
+      "learning_rate": 0.00012828335271328678,
+      "loss": 0.949,
+      "step": 4293
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7226769328117371,
+      "learning_rate": 0.000128254521371754,
+      "loss": 0.9923,
+      "step": 4294
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.762204647064209,
+      "learning_rate": 0.00012822568747758518,
+      "loss": 0.9009,
+      "step": 4295
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.8002398014068604,
+      "learning_rate": 0.00012819685103338533,
+      "loss": 1.013,
+      "step": 4296
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7634932398796082,
+      "learning_rate": 0.00012816801204175964,
+      "loss": 0.9535,
+      "step": 4297
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7286591529846191,
+      "learning_rate": 0.00012813917050531352,
+      "loss": 0.8235,
+      "step": 4298
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7977073788642883,
+      "learning_rate": 0.00012811032642665268,
+      "loss": 0.86,
+      "step": 4299
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.737250030040741,
+      "learning_rate": 0.00012808147980838302,
+      "loss": 1.0641,
+      "step": 4300
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7920666337013245,
+      "learning_rate": 0.00012805263065311063,
+      "loss": 0.8847,
+      "step": 4301
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7638289332389832,
+      "learning_rate": 0.00012802377896344193,
+      "loss": 1.1462,
+      "step": 4302
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7117271423339844,
+      "learning_rate": 0.00012799492474198344,
+      "loss": 0.9799,
+      "step": 4303
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.8018771409988403,
+      "learning_rate": 0.00012796606799134202,
+      "loss": 1.0583,
+      "step": 4304
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.8100774884223938,
+      "learning_rate": 0.00012793720871412468,
+      "loss": 0.9696,
+      "step": 4305
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.7184478044509888,
+      "learning_rate": 0.00012790834691293873,
+      "loss": 0.9018,
+      "step": 4306
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.8160077333450317,
+      "learning_rate": 0.00012787948259039165,
+      "loss": 1.1151,
+      "step": 4307
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.670409083366394,
+      "learning_rate": 0.0001278506157490912,
+      "loss": 0.9515,
+      "step": 4308
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.730998694896698,
+      "learning_rate": 0.0001278217463916453,
+      "loss": 1.0486,
+      "step": 4309
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.7451099753379822,
+      "learning_rate": 0.0001277928745206621,
+      "loss": 0.8497,
+      "step": 4310
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.7183091640472412,
+      "learning_rate": 0.00012776400013875006,
+      "loss": 0.9432,
+      "step": 4311
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.7617359161376953,
+      "learning_rate": 0.0001277351232485178,
+      "loss": 0.8215,
+      "step": 4312
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.7619341015815735,
+      "learning_rate": 0.00012770624385257423,
+      "loss": 1.071,
+      "step": 4313
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.9530133008956909,
+      "learning_rate": 0.0001276773619535284,
+      "loss": 0.9284,
+      "step": 4314
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.8487468957901001,
+      "learning_rate": 0.0001276484775539896,
+      "loss": 1.1118,
+      "step": 4315
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.8442698121070862,
+      "learning_rate": 0.00012761959065656737,
+      "loss": 0.8761,
+      "step": 4316
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.7341495156288147,
+      "learning_rate": 0.00012759070126387156,
+      "loss": 0.9816,
+      "step": 4317
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.900442361831665,
+      "learning_rate": 0.0001275618093785121,
+      "loss": 0.9645,
+      "step": 4318
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.8728200793266296,
+      "learning_rate": 0.00012753291500309922,
+      "loss": 0.9267,
+      "step": 4319
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.7861907482147217,
+      "learning_rate": 0.00012750401814024334,
+      "loss": 1.1425,
+      "step": 4320
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.8056981563568115,
+      "learning_rate": 0.00012747511879255513,
+      "loss": 0.959,
+      "step": 4321
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.8388335704803467,
+      "learning_rate": 0.00012744621696264555,
+      "loss": 1.0679,
+      "step": 4322
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.7195658087730408,
+      "learning_rate": 0.00012741731265312564,
+      "loss": 0.8373,
+      "step": 4323
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.7100113034248352,
+      "learning_rate": 0.00012738840586660677,
+      "loss": 0.9086,
+      "step": 4324
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.7771344780921936,
+      "learning_rate": 0.0001273594966057005,
+      "loss": 1.0846,
+      "step": 4325
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.748335599899292,
+      "learning_rate": 0.00012733058487301863,
+      "loss": 0.9785,
+      "step": 4326
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.8741627335548401,
+      "learning_rate": 0.00012730167067117313,
+      "loss": 0.9909,
+      "step": 4327
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.8887857794761658,
+      "learning_rate": 0.0001272727540027763,
+      "loss": 0.951,
+      "step": 4328
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.7612056136131287,
+      "learning_rate": 0.00012724383487044055,
+      "loss": 1.0076,
+      "step": 4329
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.7381542921066284,
+      "learning_rate": 0.00012721491327677862,
+      "loss": 0.9325,
+      "step": 4330
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.77471524477005,
+      "learning_rate": 0.0001271859892244033,
+      "loss": 0.9169,
+      "step": 4331
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.8204985857009888,
+      "learning_rate": 0.0001271570627159278,
+      "loss": 1.063,
+      "step": 4332
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.745324432849884,
+      "learning_rate": 0.00012712813375396546,
+      "loss": 0.9889,
+      "step": 4333
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.7239276170730591,
+      "learning_rate": 0.00012709920234112984,
+      "loss": 0.973,
+      "step": 4334
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6909604668617249,
+      "learning_rate": 0.00012707026848003471,
+      "loss": 0.8158,
+      "step": 4335
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.9055747389793396,
+      "learning_rate": 0.00012704133217329407,
+      "loss": 0.9749,
+      "step": 4336
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.8274331092834473,
+      "learning_rate": 0.00012701239342352222,
+      "loss": 0.939,
+      "step": 4337
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.8642345070838928,
+      "learning_rate": 0.00012698345223333355,
+      "loss": 0.9019,
+      "step": 4338
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.8090344667434692,
+      "learning_rate": 0.00012695450860534277,
+      "loss": 0.9395,
+      "step": 4339
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.7602019906044006,
+      "learning_rate": 0.00012692556254216471,
+      "loss": 0.9288,
+      "step": 4340
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.7767679691314697,
+      "learning_rate": 0.00012689661404641456,
+      "loss": 1.0532,
+      "step": 4341
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6911072134971619,
+      "learning_rate": 0.0001268676631207076,
+      "loss": 0.8661,
+      "step": 4342
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.7428793907165527,
+      "learning_rate": 0.0001268387097676594,
+      "loss": 0.953,
+      "step": 4343
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.7641132473945618,
+      "learning_rate": 0.00012680975398988577,
+      "loss": 0.8437,
+      "step": 4344
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.7458558082580566,
+      "learning_rate": 0.00012678079579000264,
+      "loss": 0.8762,
+      "step": 4345
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.8327491283416748,
+      "learning_rate": 0.0001267518351706262,
+      "loss": 1.0132,
+      "step": 4346
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.7638497948646545,
+      "learning_rate": 0.000126722872134373,
+      "loss": 0.803,
+      "step": 4347
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.720806896686554,
+      "learning_rate": 0.00012669390668385954,
+      "loss": 1.0511,
+      "step": 4348
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.7823967933654785,
+      "learning_rate": 0.0001266649388217028,
+      "loss": 0.9631,
+      "step": 4349
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.7093958258628845,
+      "learning_rate": 0.0001266359685505198,
+      "loss": 0.9645,
+      "step": 4350
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.9044625759124756,
+      "learning_rate": 0.00012660699587292782,
+      "loss": 1.0589,
+      "step": 4351
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.8032203316688538,
+      "learning_rate": 0.00012657802079154445,
+      "loss": 1.1116,
+      "step": 4352
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.8076255917549133,
+      "learning_rate": 0.00012654904330898742,
+      "loss": 1.0103,
+      "step": 4353
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.7746655344963074,
+      "learning_rate": 0.00012652006342787458,
+      "loss": 0.8966,
+      "step": 4354
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.854537308216095,
+      "learning_rate": 0.00012649108115082418,
+      "loss": 0.9969,
+      "step": 4355
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6939969062805176,
+      "learning_rate": 0.0001264620964804546,
+      "loss": 0.8869,
+      "step": 4356
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7948812246322632,
+      "learning_rate": 0.00012643310941938443,
+      "loss": 0.9075,
+      "step": 4357
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7065130472183228,
+      "learning_rate": 0.00012640411997023247,
+      "loss": 1.0169,
+      "step": 4358
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7679563760757446,
+      "learning_rate": 0.00012637512813561778,
+      "loss": 0.9954,
+      "step": 4359
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7309630513191223,
+      "learning_rate": 0.0001263461339181596,
+      "loss": 0.924,
+      "step": 4360
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7593986988067627,
+      "learning_rate": 0.00012631713732047742,
+      "loss": 1.1259,
+      "step": 4361
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7429969906806946,
+      "learning_rate": 0.00012628813834519083,
+      "loss": 1.0467,
+      "step": 4362
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.8035417199134827,
+      "learning_rate": 0.00012625913699491984,
+      "loss": 1.0063,
+      "step": 4363
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7101812958717346,
+      "learning_rate": 0.00012623013327228447,
+      "loss": 1.0125,
+      "step": 4364
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7777599692344666,
+      "learning_rate": 0.0001262011271799051,
+      "loss": 0.9239,
+      "step": 4365
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.680126428604126,
+      "learning_rate": 0.0001261721187204022,
+      "loss": 0.8978,
+      "step": 4366
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7517068982124329,
+      "learning_rate": 0.0001261431078963966,
+      "loss": 1.0133,
+      "step": 4367
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7588438987731934,
+      "learning_rate": 0.00012611409471050924,
+      "loss": 1.0754,
+      "step": 4368
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7387205362319946,
+      "learning_rate": 0.00012608507916536126,
+      "loss": 1.0153,
+      "step": 4369
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7710445523262024,
+      "learning_rate": 0.00012605606126357403,
+      "loss": 0.8623,
+      "step": 4370
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7151059508323669,
+      "learning_rate": 0.00012602704100776922,
+      "loss": 1.0148,
+      "step": 4371
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7390320897102356,
+      "learning_rate": 0.00012599801840056863,
+      "loss": 1.0822,
+      "step": 4372
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7536359429359436,
+      "learning_rate": 0.00012596899344459429,
+      "loss": 0.969,
+      "step": 4373
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.8541299700737,
+      "learning_rate": 0.00012593996614246843,
+      "loss": 1.0275,
+      "step": 4374
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.8438847661018372,
+      "learning_rate": 0.00012591093649681344,
+      "loss": 0.9905,
+      "step": 4375
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7539075613021851,
+      "learning_rate": 0.00012588190451025207,
+      "loss": 1.0476,
+      "step": 4376
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.8668161630630493,
+      "learning_rate": 0.0001258528701854072,
+      "loss": 1.0088,
+      "step": 4377
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6120255589485168,
+      "learning_rate": 0.00012582383352490188,
+      "loss": 0.7642,
+      "step": 4378
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.8503530621528625,
+      "learning_rate": 0.00012579479453135941,
+      "loss": 1.1022,
+      "step": 4379
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7852493524551392,
+      "learning_rate": 0.00012576575320740328,
+      "loss": 1.0352,
+      "step": 4380
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.8241454362869263,
+      "learning_rate": 0.00012573670955565725,
+      "loss": 0.8224,
+      "step": 4381
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.8677917718887329,
+      "learning_rate": 0.00012570766357874525,
+      "loss": 0.7866,
+      "step": 4382
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7602413892745972,
+      "learning_rate": 0.00012567861527929138,
+      "loss": 0.9296,
+      "step": 4383
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7921785712242126,
+      "learning_rate": 0.00012564956465992003,
+      "loss": 0.9769,
+      "step": 4384
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7256081104278564,
+      "learning_rate": 0.00012562051172325576,
+      "loss": 1.0039,
+      "step": 4385
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7994236350059509,
+      "learning_rate": 0.00012559145647192326,
+      "loss": 1.1076,
+      "step": 4386
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7854109406471252,
+      "learning_rate": 0.00012556239890854762,
+      "loss": 1.0892,
+      "step": 4387
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6456579566001892,
+      "learning_rate": 0.00012553333903575394,
+      "loss": 0.8171,
+      "step": 4388
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7927770018577576,
+      "learning_rate": 0.00012550427685616765,
+      "loss": 0.9639,
+      "step": 4389
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7582933902740479,
+      "learning_rate": 0.0001254752123724144,
+      "loss": 0.9513,
+      "step": 4390
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7228808999061584,
+      "learning_rate": 0.00012544614558711987,
+      "loss": 0.8841,
+      "step": 4391
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7322767376899719,
+      "learning_rate": 0.0001254170765029102,
+      "loss": 0.8727,
+      "step": 4392
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.8463506102561951,
+      "learning_rate": 0.00012538800512241157,
+      "loss": 0.9998,
+      "step": 4393
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.8110516667366028,
+      "learning_rate": 0.00012535893144825043,
+      "loss": 1.0143,
+      "step": 4394
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6583903431892395,
+      "learning_rate": 0.00012532985548305346,
+      "loss": 0.7218,
+      "step": 4395
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7003604173660278,
+      "learning_rate": 0.0001253007772294474,
+      "loss": 0.8849,
+      "step": 4396
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7542617321014404,
+      "learning_rate": 0.0001252716966900594,
+      "loss": 0.9391,
+      "step": 4397
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.807867705821991,
+      "learning_rate": 0.00012524261386751668,
+      "loss": 0.89,
+      "step": 4398
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7230579257011414,
+      "learning_rate": 0.00012521352876444672,
+      "loss": 0.9678,
+      "step": 4399
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7677543759346008,
+      "learning_rate": 0.0001251844413834772,
+      "loss": 1.042,
+      "step": 4400
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7520354390144348,
+      "learning_rate": 0.000125155351727236,
+      "loss": 1.1399,
+      "step": 4401
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7441850304603577,
+      "learning_rate": 0.0001251262597983512,
+      "loss": 0.999,
+      "step": 4402
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7426802515983582,
+      "learning_rate": 0.0001250971655994511,
+      "loss": 1.1166,
+      "step": 4403
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.8040526509284973,
+      "learning_rate": 0.00012506806913316416,
+      "loss": 0.9493,
+      "step": 4404
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.8495841026306152,
+      "learning_rate": 0.00012503897040211913,
+      "loss": 1.1029,
+      "step": 4405
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7909674644470215,
+      "learning_rate": 0.0001250098694089449,
+      "loss": 1.1178,
+      "step": 4406
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7505643367767334,
+      "learning_rate": 0.00012498076615627055,
+      "loss": 0.9376,
+      "step": 4407
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7308511734008789,
+      "learning_rate": 0.00012495166064672547,
+      "loss": 1.059,
+      "step": 4408
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.7313070297241211,
+      "learning_rate": 0.00012492255288293908,
+      "loss": 0.9639,
+      "step": 4409
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.8354009985923767,
+      "learning_rate": 0.00012489344286754114,
+      "loss": 1.1099,
+      "step": 4410
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.8166077733039856,
+      "learning_rate": 0.0001248643306031616,
+      "loss": 0.9699,
+      "step": 4411
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.7767633199691772,
+      "learning_rate": 0.00012483521609243058,
+      "loss": 1.0304,
+      "step": 4412
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.750893235206604,
+      "learning_rate": 0.00012480609933797837,
+      "loss": 1.075,
+      "step": 4413
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.7821363210678101,
+      "learning_rate": 0.0001247769803424356,
+      "loss": 0.9453,
+      "step": 4414
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.7858138680458069,
+      "learning_rate": 0.00012474785910843286,
+      "loss": 0.921,
+      "step": 4415
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.7058123350143433,
+      "learning_rate": 0.00012471873563860123,
+      "loss": 0.9727,
+      "step": 4416
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.8942750692367554,
+      "learning_rate": 0.0001246896099355718,
+      "loss": 1.08,
+      "step": 4417
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.7779878973960876,
+      "learning_rate": 0.00012466048200197586,
+      "loss": 1.0273,
+      "step": 4418
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.8175771236419678,
+      "learning_rate": 0.00012463135184044504,
+      "loss": 0.9516,
+      "step": 4419
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.7624543905258179,
+      "learning_rate": 0.00012460221945361102,
+      "loss": 1.0565,
+      "step": 4420
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.7882060408592224,
+      "learning_rate": 0.00012457308484410576,
+      "loss": 1.0501,
+      "step": 4421
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.7546730637550354,
+      "learning_rate": 0.00012454394801456143,
+      "loss": 1.0056,
+      "step": 4422
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.0927858352661133,
+      "learning_rate": 0.00012451480896761037,
+      "loss": 1.0178,
+      "step": 4423
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.7927306890487671,
+      "learning_rate": 0.00012448566770588513,
+      "loss": 1.0552,
+      "step": 4424
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.8408909440040588,
+      "learning_rate": 0.00012445652423201845,
+      "loss": 1.1028,
+      "step": 4425
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.7818527221679688,
+      "learning_rate": 0.00012442737854864323,
+      "loss": 0.9129,
+      "step": 4426
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.8871914744377136,
+      "learning_rate": 0.0001243982306583927,
+      "loss": 1.0881,
+      "step": 4427
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.7353706955909729,
+      "learning_rate": 0.0001243690805639002,
+      "loss": 0.9453,
+      "step": 4428
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.7749044895172119,
+      "learning_rate": 0.00012433992826779922,
+      "loss": 1.053,
+      "step": 4429
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6286710500717163,
+      "learning_rate": 0.00012431077377272356,
+      "loss": 0.8711,
+      "step": 4430
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.8130338788032532,
+      "learning_rate": 0.00012428161708130707,
+      "loss": 0.9351,
+      "step": 4431
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.791189432144165,
+      "learning_rate": 0.000124252458196184,
+      "loss": 1.1517,
+      "step": 4432
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.8531875014305115,
+      "learning_rate": 0.00012422329711998868,
+      "loss": 1.031,
+      "step": 4433
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.9011879563331604,
+      "learning_rate": 0.00012419413385535556,
+      "loss": 0.9193,
+      "step": 4434
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.7443050742149353,
+      "learning_rate": 0.0001241649684049195,
+      "loss": 1.0149,
+      "step": 4435
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.885461151599884,
+      "learning_rate": 0.0001241358007713153,
+      "loss": 0.9496,
+      "step": 4436
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.707545816898346,
+      "learning_rate": 0.00012410663095717818,
+      "loss": 1.14,
+      "step": 4437
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.7754905223846436,
+      "learning_rate": 0.00012407745896514348,
+      "loss": 1.0194,
+      "step": 4438
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.7244762182235718,
+      "learning_rate": 0.00012404828479784663,
+      "loss": 0.9338,
+      "step": 4439
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.837539792060852,
+      "learning_rate": 0.00012401910845792345,
+      "loss": 0.9798,
+      "step": 4440
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.8418575525283813,
+      "learning_rate": 0.0001239899299480098,
+      "loss": 1.0648,
+      "step": 4441
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.7444628477096558,
+      "learning_rate": 0.0001239607492707418,
+      "loss": 1.0499,
+      "step": 4442
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6809917092323303,
+      "learning_rate": 0.0001239315664287558,
+      "loss": 0.8453,
+      "step": 4443
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.7794191837310791,
+      "learning_rate": 0.00012390238142468826,
+      "loss": 0.9172,
+      "step": 4444
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.7316984534263611,
+      "learning_rate": 0.00012387319426117587,
+      "loss": 0.9582,
+      "step": 4445
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.7308700084686279,
+      "learning_rate": 0.00012384400494085557,
+      "loss": 0.9009,
+      "step": 4446
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.7057145237922668,
+      "learning_rate": 0.00012381481346636442,
+      "loss": 1.0447,
+      "step": 4447
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.9074232578277588,
+      "learning_rate": 0.0001237856198403397,
+      "loss": 0.8815,
+      "step": 4448
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.8144044876098633,
+      "learning_rate": 0.00012375642406541894,
+      "loss": 1.1888,
+      "step": 4449
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.901542603969574,
+      "learning_rate": 0.00012372722614423974,
+      "loss": 0.9081,
+      "step": 4450
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.7050389051437378,
+      "learning_rate": 0.00012369802607944,
+      "loss": 0.9428,
+      "step": 4451
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.9148438572883606,
+      "learning_rate": 0.0001236688238736578,
+      "loss": 1.0899,
+      "step": 4452
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6836315393447876,
+      "learning_rate": 0.00012363961952953137,
+      "loss": 0.8702,
+      "step": 4453
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.8242979049682617,
+      "learning_rate": 0.0001236104130496992,
+      "loss": 1.0303,
+      "step": 4454
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.7668551206588745,
+      "learning_rate": 0.00012358120443679984,
+      "loss": 1.0039,
+      "step": 4455
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6840111613273621,
+      "learning_rate": 0.0001235519936934722,
+      "loss": 0.9588,
+      "step": 4456
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.7306951284408569,
+      "learning_rate": 0.00012352278082235527,
+      "loss": 0.9726,
+      "step": 4457
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.7979377508163452,
+      "learning_rate": 0.00012349356582608826,
+      "loss": 0.9883,
+      "step": 4458
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6972531676292419,
+      "learning_rate": 0.00012346434870731066,
+      "loss": 0.9739,
+      "step": 4459
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.8911920785903931,
+      "learning_rate": 0.00012343512946866195,
+      "loss": 0.9426,
+      "step": 4460
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.7012267112731934,
+      "learning_rate": 0.00012340590811278198,
+      "loss": 0.8533,
+      "step": 4461
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.7518066167831421,
+      "learning_rate": 0.00012337668464231075,
+      "loss": 1.028,
+      "step": 4462
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.7549425959587097,
+      "learning_rate": 0.00012334745905988846,
+      "loss": 1.0232,
+      "step": 4463
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.9345147609710693,
+      "learning_rate": 0.0001233182313681554,
+      "loss": 0.876,
+      "step": 4464
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.8345490097999573,
+      "learning_rate": 0.00012328900156975217,
+      "loss": 0.9704,
+      "step": 4465
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.7318769693374634,
+      "learning_rate": 0.00012325976966731948,
+      "loss": 0.9204,
+      "step": 4466
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.7904888987541199,
+      "learning_rate": 0.00012323053566349834,
+      "loss": 1.0055,
+      "step": 4467
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.7308286428451538,
+      "learning_rate": 0.00012320129956092983,
+      "loss": 1.0266,
+      "step": 4468
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.8149561285972595,
+      "learning_rate": 0.00012317206136225524,
+      "loss": 1.0419,
+      "step": 4469
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.8579559922218323,
+      "learning_rate": 0.00012314282107011613,
+      "loss": 1.0645,
+      "step": 4470
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.7712220549583435,
+      "learning_rate": 0.00012311357868715416,
+      "loss": 0.8687,
+      "step": 4471
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.7389052510261536,
+      "learning_rate": 0.00012308433421601123,
+      "loss": 0.9033,
+      "step": 4472
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.7898889780044556,
+      "learning_rate": 0.0001230550876593294,
+      "loss": 1.0619,
+      "step": 4473
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.7135383486747742,
+      "learning_rate": 0.00012302583901975094,
+      "loss": 0.9028,
+      "step": 4474
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.8847342729568481,
+      "learning_rate": 0.0001229965882999183,
+      "loss": 1.1626,
+      "step": 4475
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.731102466583252,
+      "learning_rate": 0.0001229673355024741,
+      "loss": 0.9203,
+      "step": 4476
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6908726692199707,
+      "learning_rate": 0.0001229380806300612,
+      "loss": 0.9012,
+      "step": 4477
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.7428862452507019,
+      "learning_rate": 0.00012290882368532258,
+      "loss": 1.1055,
+      "step": 4478
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.8211758136749268,
+      "learning_rate": 0.00012287956467090148,
+      "loss": 1.1169,
+      "step": 4479
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.7881626486778259,
+      "learning_rate": 0.00012285030358944122,
+      "loss": 0.9172,
+      "step": 4480
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.8638648986816406,
+      "learning_rate": 0.0001228210404435854,
+      "loss": 1.0965,
+      "step": 4481
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.798534095287323,
+      "learning_rate": 0.00012279177523597779,
+      "loss": 0.9125,
+      "step": 4482
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6998365521430969,
+      "learning_rate": 0.00012276250796926236,
+      "loss": 0.8585,
+      "step": 4483
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.8411529064178467,
+      "learning_rate": 0.0001227332386460832,
+      "loss": 0.9901,
+      "step": 4484
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.812244176864624,
+      "learning_rate": 0.00012270396726908467,
+      "loss": 0.9175,
+      "step": 4485
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.7803433537483215,
+      "learning_rate": 0.00012267469384091118,
+      "loss": 0.9656,
+      "step": 4486
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.7464903593063354,
+      "learning_rate": 0.00012264541836420756,
+      "loss": 0.8366,
+      "step": 4487
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.7667901515960693,
+      "learning_rate": 0.00012261614084161857,
+      "loss": 0.9332,
+      "step": 4488
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.7276467084884644,
+      "learning_rate": 0.00012258686127578934,
+      "loss": 0.7042,
+      "step": 4489
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6728271842002869,
+      "learning_rate": 0.00012255757966936502,
+      "loss": 0.9618,
+      "step": 4490
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.8029244542121887,
+      "learning_rate": 0.00012252829602499114,
+      "loss": 1.1388,
+      "step": 4491
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.8557717204093933,
+      "learning_rate": 0.0001224990103453133,
+      "loss": 0.9178,
+      "step": 4492
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.7355091571807861,
+      "learning_rate": 0.0001224697226329772,
+      "loss": 1.0869,
+      "step": 4493
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.7244634628295898,
+      "learning_rate": 0.00012244043289062891,
+      "loss": 1.0627,
+      "step": 4494
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.8612757325172424,
+      "learning_rate": 0.00012241114112091457,
+      "loss": 1.1135,
+      "step": 4495
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.8004133105278015,
+      "learning_rate": 0.00012238184732648048,
+      "loss": 1.0435,
+      "step": 4496
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.7044484615325928,
+      "learning_rate": 0.00012235255150997327,
+      "loss": 0.904,
+      "step": 4497
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.8220788240432739,
+      "learning_rate": 0.00012232325367403957,
+      "loss": 0.9572,
+      "step": 4498
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.7570661306381226,
+      "learning_rate": 0.0001222939538213263,
+      "loss": 1.1492,
+      "step": 4499
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.7427449822425842,
+      "learning_rate": 0.00012226465195448055,
+      "loss": 0.8829,
+      "step": 4500
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.7731136679649353,
+      "learning_rate": 0.00012223534807614947,
+      "loss": 0.889,
+      "step": 4501
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.8267369270324707,
+      "learning_rate": 0.00012220604218898066,
+      "loss": 1.0526,
+      "step": 4502
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.8772794008255005,
+      "learning_rate": 0.00012217673429562167,
+      "loss": 1.0476,
+      "step": 4503
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.2616798877716064,
+      "learning_rate": 0.0001221474243987203,
+      "loss": 0.9864,
+      "step": 4504
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.7818945050239563,
+      "learning_rate": 0.0001221181125009245,
+      "loss": 1.0259,
+      "step": 4505
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.7399562001228333,
+      "learning_rate": 0.0001220887986048825,
+      "loss": 0.9789,
+      "step": 4506
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.7645268440246582,
+      "learning_rate": 0.00012205948271324259,
+      "loss": 1.1027,
+      "step": 4507
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.8672826290130615,
+      "learning_rate": 0.00012203016482865335,
+      "loss": 0.8492,
+      "step": 4508
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.7715070247650146,
+      "learning_rate": 0.00012200084495376341,
+      "loss": 1.1246,
+      "step": 4509
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.7514753937721252,
+      "learning_rate": 0.00012197152309122173,
+      "loss": 1.0028,
+      "step": 4510
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.721250057220459,
+      "learning_rate": 0.00012194219924367731,
+      "loss": 1.0471,
+      "step": 4511
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.7699748873710632,
+      "learning_rate": 0.00012191287341377943,
+      "loss": 0.9206,
+      "step": 4512
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.7520911693572998,
+      "learning_rate": 0.00012188354560417755,
+      "loss": 1.0336,
+      "step": 4513
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.839390754699707,
+      "learning_rate": 0.00012185421581752119,
+      "loss": 1.0853,
+      "step": 4514
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.8181332349777222,
+      "learning_rate": 0.0001218248840564602,
+      "loss": 1.0199,
+      "step": 4515
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.8653576374053955,
+      "learning_rate": 0.00012179555032364447,
+      "loss": 0.9699,
+      "step": 4516
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.7609573602676392,
+      "learning_rate": 0.00012176621462172417,
+      "loss": 0.8121,
+      "step": 4517
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.80059814453125,
+      "learning_rate": 0.00012173687695334966,
+      "loss": 0.9658,
+      "step": 4518
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.9648892879486084,
+      "learning_rate": 0.00012170753732117137,
+      "loss": 1.0759,
+      "step": 4519
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.7956419587135315,
+      "learning_rate": 0.00012167819572783998,
+      "loss": 0.9202,
+      "step": 4520
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.800743818283081,
+      "learning_rate": 0.00012164885217600637,
+      "loss": 1.0032,
+      "step": 4521
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.8416756391525269,
+      "learning_rate": 0.00012161950666832153,
+      "loss": 0.9712,
+      "step": 4522
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.8566890358924866,
+      "learning_rate": 0.00012159015920743666,
+      "loss": 1.1431,
+      "step": 4523
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.7481673955917358,
+      "learning_rate": 0.00012156080979600317,
+      "loss": 0.8806,
+      "step": 4524
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.707356870174408,
+      "learning_rate": 0.0001215314584366726,
+      "loss": 0.9942,
+      "step": 4525
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.739875078201294,
+      "learning_rate": 0.00012150210513209664,
+      "loss": 1.0685,
+      "step": 4526
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.770443320274353,
+      "learning_rate": 0.00012147274988492726,
+      "loss": 0.9136,
+      "step": 4527
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.9301662445068359,
+      "learning_rate": 0.00012144339269781647,
+      "loss": 0.9608,
+      "step": 4528
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.7839982509613037,
+      "learning_rate": 0.00012141403357341659,
+      "loss": 1.0218,
+      "step": 4529
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.7178850173950195,
+      "learning_rate": 0.00012138467251437999,
+      "loss": 0.9524,
+      "step": 4530
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.8009347915649414,
+      "learning_rate": 0.00012135530952335931,
+      "loss": 1.0965,
+      "step": 4531
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.8162659406661987,
+      "learning_rate": 0.00012132594460300737,
+      "loss": 1.0292,
+      "step": 4532
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.685593843460083,
+      "learning_rate": 0.00012129657775597705,
+      "loss": 1.0476,
+      "step": 4533
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.9040690064430237,
+      "learning_rate": 0.00012126720898492155,
+      "loss": 1.034,
+      "step": 4534
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.797566831111908,
+      "learning_rate": 0.00012123783829249411,
+      "loss": 1.0243,
+      "step": 4535
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.7130178213119507,
+      "learning_rate": 0.00012120846568134819,
+      "loss": 0.8894,
+      "step": 4536
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.7870303392410278,
+      "learning_rate": 0.0001211790911541375,
+      "loss": 1.0783,
+      "step": 4537
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.8182770013809204,
+      "learning_rate": 0.00012114971471351588,
+      "loss": 0.9987,
+      "step": 4538
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.8288918733596802,
+      "learning_rate": 0.00012112033636213725,
+      "loss": 1.0125,
+      "step": 4539
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.674735963344574,
+      "learning_rate": 0.00012109095610265582,
+      "loss": 0.9747,
+      "step": 4540
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.752810001373291,
+      "learning_rate": 0.00012106157393772588,
+      "loss": 1.0043,
+      "step": 4541
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.737766683101654,
+      "learning_rate": 0.00012103218987000205,
+      "loss": 0.9728,
+      "step": 4542
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.7705424427986145,
+      "learning_rate": 0.00012100280390213894,
+      "loss": 0.9582,
+      "step": 4543
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6810630559921265,
+      "learning_rate": 0.00012097341603679137,
+      "loss": 0.891,
+      "step": 4544
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6789401173591614,
+      "learning_rate": 0.00012094402627661447,
+      "loss": 0.926,
+      "step": 4545
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.7720015048980713,
+      "learning_rate": 0.00012091463462426334,
+      "loss": 1.0054,
+      "step": 4546
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.7323508262634277,
+      "learning_rate": 0.00012088524108239338,
+      "loss": 1.0275,
+      "step": 4547
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.7956165671348572,
+      "learning_rate": 0.0001208558456536602,
+      "loss": 1.1275,
+      "step": 4548
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.8188318014144897,
+      "learning_rate": 0.00012082644834071939,
+      "loss": 1.0057,
+      "step": 4549
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.8520888090133667,
+      "learning_rate": 0.00012079704914622691,
+      "loss": 0.986,
+      "step": 4550
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.7133857011795044,
+      "learning_rate": 0.0001207676480728388,
+      "loss": 0.9289,
+      "step": 4551
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.7838119268417358,
+      "learning_rate": 0.00012073824512321126,
+      "loss": 0.8795,
+      "step": 4552
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.8384571075439453,
+      "learning_rate": 0.00012070884030000071,
+      "loss": 1.0691,
+      "step": 4553
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.7315096855163574,
+      "learning_rate": 0.00012067943360586369,
+      "loss": 0.9933,
+      "step": 4554
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.9339639544487,
+      "learning_rate": 0.00012065002504345692,
+      "loss": 0.97,
+      "step": 4555
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.869044840335846,
+      "learning_rate": 0.0001206206146154373,
+      "loss": 1.0211,
+      "step": 4556
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.7361880540847778,
+      "learning_rate": 0.00012059120232446191,
+      "loss": 0.7521,
+      "step": 4557
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.767719030380249,
+      "learning_rate": 0.000120561788173188,
+      "loss": 1.1313,
+      "step": 4558
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.7440416812896729,
+      "learning_rate": 0.00012053237216427293,
+      "loss": 0.9252,
+      "step": 4559
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.8230945467948914,
+      "learning_rate": 0.00012050295430037428,
+      "loss": 1.0488,
+      "step": 4560
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.7725455164909363,
+      "learning_rate": 0.00012047353458414981,
+      "loss": 0.8682,
+      "step": 4561
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.7603073120117188,
+      "learning_rate": 0.00012044411301825742,
+      "loss": 0.9021,
+      "step": 4562
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.9435483813285828,
+      "learning_rate": 0.00012041468960535516,
+      "loss": 0.9668,
+      "step": 4563
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.7270300388336182,
+      "learning_rate": 0.00012038526434810129,
+      "loss": 0.9416,
+      "step": 4564
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.7929261326789856,
+      "learning_rate": 0.00012035583724915418,
+      "loss": 1.0635,
+      "step": 4565
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6248114109039307,
+      "learning_rate": 0.00012032640831117243,
+      "loss": 0.7921,
+      "step": 4566
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8861623406410217,
+      "learning_rate": 0.00012029697753681482,
+      "loss": 1.1924,
+      "step": 4567
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6975063681602478,
+      "learning_rate": 0.00012026754492874017,
+      "loss": 0.9724,
+      "step": 4568
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.7411131262779236,
+      "learning_rate": 0.00012023811048960763,
+      "loss": 0.902,
+      "step": 4569
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.7723414301872253,
+      "learning_rate": 0.00012020867422207637,
+      "loss": 0.96,
+      "step": 4570
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8154706358909607,
+      "learning_rate": 0.00012017923612880579,
+      "loss": 0.975,
+      "step": 4571
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.7440472841262817,
+      "learning_rate": 0.00012014979621245552,
+      "loss": 1.1156,
+      "step": 4572
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.014880895614624,
+      "learning_rate": 0.00012012035447568524,
+      "loss": 0.9663,
+      "step": 4573
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.7933245301246643,
+      "learning_rate": 0.00012009091092115486,
+      "loss": 0.9387,
+      "step": 4574
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8018094301223755,
+      "learning_rate": 0.00012006146555152442,
+      "loss": 0.9435,
+      "step": 4575
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.7780526280403137,
+      "learning_rate": 0.00012003201836945414,
+      "loss": 0.8845,
+      "step": 4576
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8302222490310669,
+      "learning_rate": 0.00012000256937760445,
+      "loss": 1.0314,
+      "step": 4577
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.7264339923858643,
+      "learning_rate": 0.00011997311857863587,
+      "loss": 1.0435,
+      "step": 4578
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8215857744216919,
+      "learning_rate": 0.0001199436659752091,
+      "loss": 1.1421,
+      "step": 4579
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.7523236274719238,
+      "learning_rate": 0.00011991421156998506,
+      "loss": 0.8928,
+      "step": 4580
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6750524044036865,
+      "learning_rate": 0.00011988475536562471,
+      "loss": 0.8169,
+      "step": 4581
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.7918744683265686,
+      "learning_rate": 0.00011985529736478931,
+      "loss": 1.1446,
+      "step": 4582
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.7331019639968872,
+      "learning_rate": 0.00011982583757014025,
+      "loss": 0.8463,
+      "step": 4583
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8649581670761108,
+      "learning_rate": 0.00011979637598433899,
+      "loss": 0.8682,
+      "step": 4584
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8099588751792908,
+      "learning_rate": 0.00011976691261004727,
+      "loss": 1.1053,
+      "step": 4585
+    },
+    {
+      "epoch": 1.73,
+      "eval_loss": 0.9523382782936096,
+      "eval_runtime": 17.757,
+      "eval_samples_per_second": 153.686,
+      "eval_steps_per_second": 38.464,
+      "step": 4585
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8093003630638123,
+      "learning_rate": 0.00011973744744992688,
+      "loss": 0.8328,
+      "step": 4586
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.7210092544555664,
+      "learning_rate": 0.00011970798050663987,
+      "loss": 0.8347,
+      "step": 4587
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8135833144187927,
+      "learning_rate": 0.00011967851178284843,
+      "loss": 1.0053,
+      "step": 4588
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.7191389203071594,
+      "learning_rate": 0.00011964904128121486,
+      "loss": 0.9883,
+      "step": 4589
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.7279300689697266,
+      "learning_rate": 0.00011961956900440168,
+      "loss": 0.9336,
+      "step": 4590
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6938570141792297,
+      "learning_rate": 0.0001195900949550715,
+      "loss": 0.9598,
+      "step": 4591
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.7962003350257874,
+      "learning_rate": 0.00011956061913588716,
+      "loss": 0.9078,
+      "step": 4592
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.7468202114105225,
+      "learning_rate": 0.00011953114154951166,
+      "loss": 0.7303,
+      "step": 4593
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.7891976237297058,
+      "learning_rate": 0.0001195016621986081,
+      "loss": 0.9259,
+      "step": 4594
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.0634568929672241,
+      "learning_rate": 0.00011947218108583973,
+      "loss": 0.5085,
+      "step": 4595
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.8258597254753113,
+      "learning_rate": 0.0001194426982138701,
+      "loss": 0.8869,
+      "step": 4596
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.7219350337982178,
+      "learning_rate": 0.00011941321358536276,
+      "loss": 0.8407,
+      "step": 4597
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.9141265749931335,
+      "learning_rate": 0.0001193837272029815,
+      "loss": 1.0333,
+      "step": 4598
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.7562088966369629,
+      "learning_rate": 0.00011935423906939023,
+      "loss": 0.9877,
+      "step": 4599
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.8154380321502686,
+      "learning_rate": 0.00011932474918725303,
+      "loss": 0.8455,
+      "step": 4600
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.8079681396484375,
+      "learning_rate": 0.00011929525755923417,
+      "loss": 1.0385,
+      "step": 4601
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.7577683925628662,
+      "learning_rate": 0.00011926576418799804,
+      "loss": 1.1408,
+      "step": 4602
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.8447030186653137,
+      "learning_rate": 0.0001192362690762092,
+      "loss": 0.9535,
+      "step": 4603
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6743953227996826,
+      "learning_rate": 0.00011920677222653236,
+      "loss": 0.7593,
+      "step": 4604
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.9185356497764587,
+      "learning_rate": 0.00011917727364163238,
+      "loss": 0.9552,
+      "step": 4605
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.8450148701667786,
+      "learning_rate": 0.00011914777332417433,
+      "loss": 0.9634,
+      "step": 4606
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.8802364468574524,
+      "learning_rate": 0.00011911827127682337,
+      "loss": 0.7896,
+      "step": 4607
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.8088971972465515,
+      "learning_rate": 0.00011908876750224485,
+      "loss": 0.8993,
+      "step": 4608
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.7739360332489014,
+      "learning_rate": 0.00011905926200310427,
+      "loss": 0.9611,
+      "step": 4609
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.9451324343681335,
+      "learning_rate": 0.00011902975478206727,
+      "loss": 1.0925,
+      "step": 4610
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.8480505347251892,
+      "learning_rate": 0.00011900024584179966,
+      "loss": 1.1357,
+      "step": 4611
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.7975031733512878,
+      "learning_rate": 0.00011897073518496744,
+      "loss": 1.024,
+      "step": 4612
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.7346519231796265,
+      "learning_rate": 0.00011894122281423672,
+      "loss": 0.7701,
+      "step": 4613
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.7548093795776367,
+      "learning_rate": 0.00011891170873227373,
+      "loss": 0.9828,
+      "step": 4614
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.7536081671714783,
+      "learning_rate": 0.00011888219294174494,
+      "loss": 0.9656,
+      "step": 4615
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.837548553943634,
+      "learning_rate": 0.00011885267544531695,
+      "loss": 1.024,
+      "step": 4616
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.865839421749115,
+      "learning_rate": 0.00011882315624565645,
+      "loss": 0.9148,
+      "step": 4617
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.7667220830917358,
+      "learning_rate": 0.0001187936353454304,
+      "loss": 0.9866,
+      "step": 4618
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6959596872329712,
+      "learning_rate": 0.00011876411274730577,
+      "loss": 0.9945,
+      "step": 4619
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.7026944160461426,
+      "learning_rate": 0.00011873458845394979,
+      "loss": 0.8613,
+      "step": 4620
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.736354649066925,
+      "learning_rate": 0.00011870506246802984,
+      "loss": 0.8945,
+      "step": 4621
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.9902938604354858,
+      "learning_rate": 0.00011867553479221339,
+      "loss": 0.9826,
+      "step": 4622
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.7676527500152588,
+      "learning_rate": 0.00011864600542916813,
+      "loss": 0.9257,
+      "step": 4623
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6731281280517578,
+      "learning_rate": 0.00011861647438156183,
+      "loss": 0.8324,
+      "step": 4624
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.8030645251274109,
+      "learning_rate": 0.00011858694165206248,
+      "loss": 1.0004,
+      "step": 4625
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.7600473761558533,
+      "learning_rate": 0.00011855740724333821,
+      "loss": 0.9411,
+      "step": 4626
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6938033699989319,
+      "learning_rate": 0.00011852787115805726,
+      "loss": 0.8006,
+      "step": 4627
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.7645683884620667,
+      "learning_rate": 0.00011849833339888805,
+      "loss": 0.9532,
+      "step": 4628
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.7587887644767761,
+      "learning_rate": 0.00011846879396849917,
+      "loss": 0.9675,
+      "step": 4629
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.9182590842247009,
+      "learning_rate": 0.00011843925286955929,
+      "loss": 0.9558,
+      "step": 4630
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.7563308477401733,
+      "learning_rate": 0.00011840971010473734,
+      "loss": 0.9988,
+      "step": 4631
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6797767281532288,
+      "learning_rate": 0.00011838016567670231,
+      "loss": 0.8188,
+      "step": 4632
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.7414708733558655,
+      "learning_rate": 0.00011835061958812339,
+      "loss": 0.9559,
+      "step": 4633
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.7425746321678162,
+      "learning_rate": 0.0001183210718416699,
+      "loss": 1.1321,
+      "step": 4634
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.7789604067802429,
+      "learning_rate": 0.00011829152244001127,
+      "loss": 0.8989,
+      "step": 4635
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.8125724196434021,
+      "learning_rate": 0.00011826197138581719,
+      "loss": 0.9971,
+      "step": 4636
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.7495322227478027,
+      "learning_rate": 0.00011823241868175738,
+      "loss": 0.9785,
+      "step": 4637
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.7192978858947754,
+      "learning_rate": 0.00011820286433050176,
+      "loss": 0.974,
+      "step": 4638
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.7956582307815552,
+      "learning_rate": 0.00011817330833472042,
+      "loss": 0.8802,
+      "step": 4639
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.8172240257263184,
+      "learning_rate": 0.00011814375069708357,
+      "loss": 1.0455,
+      "step": 4640
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.7698330283164978,
+      "learning_rate": 0.00011811419142026156,
+      "loss": 0.9873,
+      "step": 4641
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.8238794207572937,
+      "learning_rate": 0.00011808463050692495,
+      "loss": 1.0997,
+      "step": 4642
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.8221603035926819,
+      "learning_rate": 0.00011805506795974435,
+      "loss": 1.038,
+      "step": 4643
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.8607699871063232,
+      "learning_rate": 0.0001180255037813906,
+      "loss": 0.9641,
+      "step": 4644
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7579261064529419,
+      "learning_rate": 0.00011799593797453462,
+      "loss": 0.942,
+      "step": 4645
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.8301036357879639,
+      "learning_rate": 0.00011796637054184753,
+      "loss": 0.9381,
+      "step": 4646
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7816101908683777,
+      "learning_rate": 0.00011793680148600062,
+      "loss": 0.9926,
+      "step": 4647
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.8143907189369202,
+      "learning_rate": 0.00011790723080966524,
+      "loss": 1.1058,
+      "step": 4648
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7717945575714111,
+      "learning_rate": 0.00011787765851551296,
+      "loss": 1.13,
+      "step": 4649
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.8154736757278442,
+      "learning_rate": 0.00011784808460621541,
+      "loss": 1.0087,
+      "step": 4650
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7709436416625977,
+      "learning_rate": 0.0001178185090844445,
+      "loss": 0.869,
+      "step": 4651
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7582793831825256,
+      "learning_rate": 0.00011778893195287222,
+      "loss": 1.0573,
+      "step": 4652
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.8012235760688782,
+      "learning_rate": 0.00011775935321417062,
+      "loss": 0.8933,
+      "step": 4653
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7932279109954834,
+      "learning_rate": 0.000117729772871012,
+      "loss": 1.0786,
+      "step": 4654
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.9135048985481262,
+      "learning_rate": 0.00011770019092606879,
+      "loss": 1.0244,
+      "step": 4655
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.8108317852020264,
+      "learning_rate": 0.00011767060738201358,
+      "loss": 1.0483,
+      "step": 4656
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.808539867401123,
+      "learning_rate": 0.000117641022241519,
+      "loss": 0.8809,
+      "step": 4657
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.8646968603134155,
+      "learning_rate": 0.00011761143550725799,
+      "loss": 0.9518,
+      "step": 4658
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7806581258773804,
+      "learning_rate": 0.00011758184718190346,
+      "loss": 1.0036,
+      "step": 4659
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6390308737754822,
+      "learning_rate": 0.0001175522572681286,
+      "loss": 0.7685,
+      "step": 4660
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7053999304771423,
+      "learning_rate": 0.00011752266576860668,
+      "loss": 0.9902,
+      "step": 4661
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7349623441696167,
+      "learning_rate": 0.00011749307268601111,
+      "loss": 0.9056,
+      "step": 4662
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7815699577331543,
+      "learning_rate": 0.0001174634780230155,
+      "loss": 0.9977,
+      "step": 4663
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7745949029922485,
+      "learning_rate": 0.00011743388178229349,
+      "loss": 1.0299,
+      "step": 4664
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.8292572498321533,
+      "learning_rate": 0.00011740428396651896,
+      "loss": 1.0306,
+      "step": 4665
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.8445864319801331,
+      "learning_rate": 0.00011737468457836598,
+      "loss": 1.0174,
+      "step": 4666
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7877253890037537,
+      "learning_rate": 0.00011734508362050856,
+      "loss": 0.904,
+      "step": 4667
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.8767524361610413,
+      "learning_rate": 0.00011731548109562109,
+      "loss": 1.0872,
+      "step": 4668
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7433600425720215,
+      "learning_rate": 0.00011728587700637792,
+      "loss": 0.8279,
+      "step": 4669
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7911831736564636,
+      "learning_rate": 0.0001172562713554536,
+      "loss": 1.0507,
+      "step": 4670
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7265031933784485,
+      "learning_rate": 0.00011722666414552293,
+      "loss": 0.9152,
+      "step": 4671
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.757936954498291,
+      "learning_rate": 0.00011719705537926069,
+      "loss": 0.995,
+      "step": 4672
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7522801160812378,
+      "learning_rate": 0.00011716744505934182,
+      "loss": 0.7812,
+      "step": 4673
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6782301664352417,
+      "learning_rate": 0.00011713783318844154,
+      "loss": 0.9455,
+      "step": 4674
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7656162977218628,
+      "learning_rate": 0.000117108219769235,
+      "loss": 0.9878,
+      "step": 4675
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8663710355758667,
+      "learning_rate": 0.00011707860480439772,
+      "loss": 1.1014,
+      "step": 4676
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7111193537712097,
+      "learning_rate": 0.00011704898829660518,
+      "loss": 0.9375,
+      "step": 4677
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7736416459083557,
+      "learning_rate": 0.00011701937024853308,
+      "loss": 0.8987,
+      "step": 4678
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7856674194335938,
+      "learning_rate": 0.00011698975066285727,
+      "loss": 1.0611,
+      "step": 4679
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7942124009132385,
+      "learning_rate": 0.00011696012954225364,
+      "loss": 1.0369,
+      "step": 4680
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7673232555389404,
+      "learning_rate": 0.00011693050688939833,
+      "loss": 0.8537,
+      "step": 4681
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8419005870819092,
+      "learning_rate": 0.00011690088270696761,
+      "loss": 0.9412,
+      "step": 4682
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8815070390701294,
+      "learning_rate": 0.00011687125699763781,
+      "loss": 1.0334,
+      "step": 4683
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7238216400146484,
+      "learning_rate": 0.0001168416297640855,
+      "loss": 0.9714,
+      "step": 4684
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8091062307357788,
+      "learning_rate": 0.00011681200100898727,
+      "loss": 0.9395,
+      "step": 4685
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8089106678962708,
+      "learning_rate": 0.00011678237073501992,
+      "loss": 1.0512,
+      "step": 4686
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7475075721740723,
+      "learning_rate": 0.00011675273894486046,
+      "loss": 0.8836,
+      "step": 4687
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.825634241104126,
+      "learning_rate": 0.00011672310564118585,
+      "loss": 0.9447,
+      "step": 4688
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8098149299621582,
+      "learning_rate": 0.00011669347082667332,
+      "loss": 0.9268,
+      "step": 4689
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7696269154548645,
+      "learning_rate": 0.00011666383450400022,
+      "loss": 0.9186,
+      "step": 4690
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7242420315742493,
+      "learning_rate": 0.00011663419667584404,
+      "loss": 0.992,
+      "step": 4691
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.816675066947937,
+      "learning_rate": 0.00011660455734488236,
+      "loss": 1.1345,
+      "step": 4692
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7482806444168091,
+      "learning_rate": 0.00011657491651379296,
+      "loss": 0.9011,
+      "step": 4693
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.709613561630249,
+      "learning_rate": 0.00011654527418525367,
+      "loss": 0.9484,
+      "step": 4694
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7983031272888184,
+      "learning_rate": 0.00011651563036194253,
+      "loss": 1.0922,
+      "step": 4695
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7807701826095581,
+      "learning_rate": 0.00011648598504653774,
+      "loss": 0.9652,
+      "step": 4696
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.721126914024353,
+      "learning_rate": 0.00011645633824171751,
+      "loss": 0.9591,
+      "step": 4697
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.7826765179634094,
+      "learning_rate": 0.00011642668995016033,
+      "loss": 0.9512,
+      "step": 4698
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.7089790105819702,
+      "learning_rate": 0.0001163970401745447,
+      "loss": 0.7106,
+      "step": 4699
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.77752685546875,
+      "learning_rate": 0.00011636738891754931,
+      "loss": 0.9311,
+      "step": 4700
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.7967430949211121,
+      "learning_rate": 0.00011633773618185302,
+      "loss": 1.1004,
+      "step": 4701
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.7402793765068054,
+      "learning_rate": 0.00011630808197013477,
+      "loss": 0.8836,
+      "step": 4702
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6968073844909668,
+      "learning_rate": 0.00011627842628507366,
+      "loss": 1.0618,
+      "step": 4703
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.743408739566803,
+      "learning_rate": 0.00011624876912934892,
+      "loss": 0.9196,
+      "step": 4704
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.9421427249908447,
+      "learning_rate": 0.00011621911050563983,
+      "loss": 1.0486,
+      "step": 4705
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.8492775559425354,
+      "learning_rate": 0.00011618945041662599,
+      "loss": 1.31,
+      "step": 4706
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.8934208750724792,
+      "learning_rate": 0.00011615978886498695,
+      "loss": 0.9059,
+      "step": 4707
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.8813003301620483,
+      "learning_rate": 0.00011613012585340249,
+      "loss": 0.9019,
+      "step": 4708
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.924180269241333,
+      "learning_rate": 0.00011610046138455249,
+      "loss": 0.9498,
+      "step": 4709
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.7572709321975708,
+      "learning_rate": 0.00011607079546111692,
+      "loss": 0.8413,
+      "step": 4710
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.7872298359870911,
+      "learning_rate": 0.00011604112808577603,
+      "loss": 0.81,
+      "step": 4711
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.8559049963951111,
+      "learning_rate": 0.00011601145926121005,
+      "loss": 0.745,
+      "step": 4712
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.7605762481689453,
+      "learning_rate": 0.00011598178899009933,
+      "loss": 0.8235,
+      "step": 4713
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.768850564956665,
+      "learning_rate": 0.0001159521172751245,
+      "loss": 0.8782,
+      "step": 4714
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.9820607304573059,
+      "learning_rate": 0.00011592244411896618,
+      "loss": 1.1446,
+      "step": 4715
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.7328512668609619,
+      "learning_rate": 0.00011589276952430517,
+      "loss": 0.9038,
+      "step": 4716
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.8597776889801025,
+      "learning_rate": 0.00011586309349382245,
+      "loss": 0.9516,
+      "step": 4717
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.8076006770133972,
+      "learning_rate": 0.00011583341603019903,
+      "loss": 0.85,
+      "step": 4718
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.9194772243499756,
+      "learning_rate": 0.00011580373713611615,
+      "loss": 1.0622,
+      "step": 4719
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.7424063682556152,
+      "learning_rate": 0.00011577405681425506,
+      "loss": 0.8937,
+      "step": 4720
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.737322986125946,
+      "learning_rate": 0.00011574437506729726,
+      "loss": 1.0043,
+      "step": 4721
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.9296445250511169,
+      "learning_rate": 0.00011571469189792433,
+      "loss": 0.8489,
+      "step": 4722
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.7422108054161072,
+      "learning_rate": 0.00011568500730881795,
+      "loss": 0.9638,
+      "step": 4723
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.7797747850418091,
+      "learning_rate": 0.00011565532130265999,
+      "loss": 0.9662,
+      "step": 4724
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.8456657528877258,
+      "learning_rate": 0.00011562563388213236,
+      "loss": 0.8727,
+      "step": 4725
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.7890819907188416,
+      "learning_rate": 0.0001155959450499172,
+      "loss": 0.9409,
+      "step": 4726
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.9545003175735474,
+      "learning_rate": 0.0001155662548086967,
+      "loss": 0.9497,
+      "step": 4727
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.9436349868774414,
+      "learning_rate": 0.00011553656316115325,
+      "loss": 1.0286,
+      "step": 4728
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.7221186757087708,
+      "learning_rate": 0.00011550687010996924,
+      "loss": 0.9814,
+      "step": 4729
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.7962702512741089,
+      "learning_rate": 0.00011547717565782732,
+      "loss": 0.9702,
+      "step": 4730
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.8563812971115112,
+      "learning_rate": 0.00011544747980741021,
+      "loss": 0.9316,
+      "step": 4731
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6784048080444336,
+      "learning_rate": 0.0001154177825614008,
+      "loss": 0.8695,
+      "step": 4732
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6901435256004333,
+      "learning_rate": 0.00011538808392248199,
+      "loss": 0.8863,
+      "step": 4733
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.7063350677490234,
+      "learning_rate": 0.00011535838389333694,
+      "loss": 0.9459,
+      "step": 4734
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.8247732520103455,
+      "learning_rate": 0.00011532868247664885,
+      "loss": 0.9109,
+      "step": 4735
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.8078949451446533,
+      "learning_rate": 0.00011529897967510113,
+      "loss": 1.0646,
+      "step": 4736
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.8562504053115845,
+      "learning_rate": 0.00011526927549137716,
+      "loss": 1.0633,
+      "step": 4737
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.7846590280532837,
+      "learning_rate": 0.00011523956992816065,
+      "loss": 1.0485,
+      "step": 4738
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.8158285617828369,
+      "learning_rate": 0.00011520986298813527,
+      "loss": 0.9678,
+      "step": 4739
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6645245552062988,
+      "learning_rate": 0.00011518015467398489,
+      "loss": 0.9141,
+      "step": 4740
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.7792890667915344,
+      "learning_rate": 0.00011515044498839347,
+      "loss": 0.8894,
+      "step": 4741
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.7953566312789917,
+      "learning_rate": 0.00011512073393404514,
+      "loss": 1.0201,
+      "step": 4742
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.759485125541687,
+      "learning_rate": 0.00011509102151362412,
+      "loss": 1.0014,
+      "step": 4743
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.7176987528800964,
+      "learning_rate": 0.00011506130772981476,
+      "loss": 0.9569,
+      "step": 4744
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.7796375751495361,
+      "learning_rate": 0.00011503159258530147,
+      "loss": 0.9428,
+      "step": 4745
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.8124610185623169,
+      "learning_rate": 0.00011500187608276893,
+      "loss": 0.9826,
+      "step": 4746
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.7761412858963013,
+      "learning_rate": 0.00011497215822490186,
+      "loss": 0.7457,
+      "step": 4747
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6975809931755066,
+      "learning_rate": 0.00011494243901438501,
+      "loss": 0.8538,
+      "step": 4748
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.8496014475822449,
+      "learning_rate": 0.00011491271845390345,
+      "loss": 1.0841,
+      "step": 4749
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7893025875091553,
+      "learning_rate": 0.00011488299654614219,
+      "loss": 1.0616,
+      "step": 4750
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6957244277000427,
+      "learning_rate": 0.00011485327329378644,
+      "loss": 0.8699,
+      "step": 4751
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7174831032752991,
+      "learning_rate": 0.00011482354869952156,
+      "loss": 0.851,
+      "step": 4752
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7984290719032288,
+      "learning_rate": 0.000114793822766033,
+      "loss": 1.0114,
+      "step": 4753
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7365847826004028,
+      "learning_rate": 0.00011476409549600631,
+      "loss": 1.0527,
+      "step": 4754
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7391736507415771,
+      "learning_rate": 0.00011473436689212718,
+      "loss": 1.0829,
+      "step": 4755
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7169732451438904,
+      "learning_rate": 0.00011470463695708142,
+      "loss": 1.0403,
+      "step": 4756
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.8890628218650818,
+      "learning_rate": 0.00011467490569355499,
+      "loss": 0.8923,
+      "step": 4757
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7755351066589355,
+      "learning_rate": 0.00011464517310423391,
+      "loss": 0.9588,
+      "step": 4758
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7561038136482239,
+      "learning_rate": 0.00011461543919180436,
+      "loss": 0.8168,
+      "step": 4759
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.858525276184082,
+      "learning_rate": 0.0001145857039589526,
+      "loss": 1.0978,
+      "step": 4760
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7823034524917603,
+      "learning_rate": 0.00011455596740836512,
+      "loss": 0.9898,
+      "step": 4761
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7411243915557861,
+      "learning_rate": 0.00011452622954272838,
+      "loss": 0.8592,
+      "step": 4762
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.8100503087043762,
+      "learning_rate": 0.00011449649036472907,
+      "loss": 0.9123,
+      "step": 4763
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.8389270305633545,
+      "learning_rate": 0.0001144667498770539,
+      "loss": 0.9103,
+      "step": 4764
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.8715300559997559,
+      "learning_rate": 0.0001144370080823898,
+      "loss": 1.0515,
+      "step": 4765
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7054761052131653,
+      "learning_rate": 0.00011440726498342376,
+      "loss": 0.9034,
+      "step": 4766
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7008001804351807,
+      "learning_rate": 0.0001143775205828429,
+      "loss": 0.9998,
+      "step": 4767
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7793424725532532,
+      "learning_rate": 0.00011434777488333448,
+      "loss": 1.0721,
+      "step": 4768
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7706813216209412,
+      "learning_rate": 0.00011431802788758581,
+      "loss": 0.9,
+      "step": 4769
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7926739454269409,
+      "learning_rate": 0.00011428827959828441,
+      "loss": 1.1088,
+      "step": 4770
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7428482174873352,
+      "learning_rate": 0.00011425853001811784,
+      "loss": 0.999,
+      "step": 4771
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.8474048972129822,
+      "learning_rate": 0.00011422877914977381,
+      "loss": 0.9764,
+      "step": 4772
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.9430420994758606,
+      "learning_rate": 0.00011419902699594016,
+      "loss": 1.2097,
+      "step": 4773
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.902398943901062,
+      "learning_rate": 0.0001141692735593048,
+      "loss": 0.9944,
+      "step": 4774
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7506820559501648,
+      "learning_rate": 0.0001141395188425558,
+      "loss": 1.1252,
+      "step": 4775
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7798844575881958,
+      "learning_rate": 0.00011410976284838135,
+      "loss": 1.0435,
+      "step": 4776
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7502029538154602,
+      "learning_rate": 0.0001140800055794697,
+      "loss": 1.0445,
+      "step": 4777
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7140774726867676,
+      "learning_rate": 0.00011405024703850929,
+      "loss": 1.0022,
+      "step": 4778
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6743187308311462,
+      "learning_rate": 0.0001140204872281886,
+      "loss": 0.9194,
+      "step": 4779
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.8572053909301758,
+      "learning_rate": 0.00011399072615119623,
+      "loss": 0.9496,
+      "step": 4780
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7410863041877747,
+      "learning_rate": 0.00011396096381022103,
+      "loss": 0.9508,
+      "step": 4781
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7025617957115173,
+      "learning_rate": 0.00011393120020795178,
+      "loss": 0.7898,
+      "step": 4782
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7673794627189636,
+      "learning_rate": 0.00011390143534707744,
+      "loss": 0.9508,
+      "step": 4783
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.8125484585762024,
+      "learning_rate": 0.00011387166923028716,
+      "loss": 1.0228,
+      "step": 4784
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7624559998512268,
+      "learning_rate": 0.00011384190186027007,
+      "loss": 0.8023,
+      "step": 4785
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7645179629325867,
+      "learning_rate": 0.00011381213323971555,
+      "loss": 0.9577,
+      "step": 4786
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.8080947995185852,
+      "learning_rate": 0.00011378236337131301,
+      "loss": 0.9995,
+      "step": 4787
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.883863627910614,
+      "learning_rate": 0.00011375259225775192,
+      "loss": 0.9384,
+      "step": 4788
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7815911769866943,
+      "learning_rate": 0.00011372281990172204,
+      "loss": 0.9968,
+      "step": 4789
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7550731897354126,
+      "learning_rate": 0.00011369304630591304,
+      "loss": 0.8006,
+      "step": 4790
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.2814509868621826,
+      "learning_rate": 0.00011366327147301486,
+      "loss": 1.0837,
+      "step": 4791
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7658206224441528,
+      "learning_rate": 0.00011363349540571747,
+      "loss": 0.9169,
+      "step": 4792
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7140726447105408,
+      "learning_rate": 0.00011360371810671094,
+      "loss": 0.8128,
+      "step": 4793
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.785155177116394,
+      "learning_rate": 0.00011357393957868554,
+      "loss": 1.0296,
+      "step": 4794
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.8359742164611816,
+      "learning_rate": 0.00011354415982433151,
+      "loss": 0.9683,
+      "step": 4795
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7292312383651733,
+      "learning_rate": 0.00011351437884633933,
+      "loss": 0.8378,
+      "step": 4796
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7945870757102966,
+      "learning_rate": 0.00011348459664739956,
+      "loss": 1.2087,
+      "step": 4797
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.8345226645469666,
+      "learning_rate": 0.00011345481323020287,
+      "loss": 0.9874,
+      "step": 4798
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7681534886360168,
+      "learning_rate": 0.0001134250285974399,
+      "loss": 0.9763,
+      "step": 4799
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.8912960886955261,
+      "learning_rate": 0.00011339524275180167,
+      "loss": 1.0071,
+      "step": 4800
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.8802440762519836,
+      "learning_rate": 0.00011336545569597908,
+      "loss": 1.0812,
+      "step": 4801
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6936193108558655,
+      "learning_rate": 0.00011333566743266324,
+      "loss": 0.8666,
+      "step": 4802
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7325948476791382,
+      "learning_rate": 0.00011330587796454537,
+      "loss": 0.9684,
+      "step": 4803
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7145864963531494,
+      "learning_rate": 0.00011327608729431673,
+      "loss": 0.948,
+      "step": 4804
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7360267043113708,
+      "learning_rate": 0.0001132462954246688,
+      "loss": 0.8956,
+      "step": 4805
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7906669974327087,
+      "learning_rate": 0.00011321650235829309,
+      "loss": 0.9409,
+      "step": 4806
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7279953956604004,
+      "learning_rate": 0.00011318670809788119,
+      "loss": 0.9309,
+      "step": 4807
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7352005839347839,
+      "learning_rate": 0.00011315691264612492,
+      "loss": 0.9411,
+      "step": 4808
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7934585809707642,
+      "learning_rate": 0.00011312711600571604,
+      "loss": 0.9128,
+      "step": 4809
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.8576295375823975,
+      "learning_rate": 0.00011309731817934656,
+      "loss": 0.9998,
+      "step": 4810
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7300700545310974,
+      "learning_rate": 0.00011306751916970857,
+      "loss": 0.9852,
+      "step": 4811
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.8238819241523743,
+      "learning_rate": 0.0001130377189794942,
+      "loss": 0.9775,
+      "step": 4812
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7973847985267639,
+      "learning_rate": 0.00011300791761139575,
+      "loss": 0.9101,
+      "step": 4813
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.8104704022407532,
+      "learning_rate": 0.00011297811506810556,
+      "loss": 1.1568,
+      "step": 4814
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7424970269203186,
+      "learning_rate": 0.00011294831135231617,
+      "loss": 0.9855,
+      "step": 4815
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7911181449890137,
+      "learning_rate": 0.00011291850646672017,
+      "loss": 0.9468,
+      "step": 4816
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7143746018409729,
+      "learning_rate": 0.00011288870041401025,
+      "loss": 0.9719,
+      "step": 4817
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.8238941431045532,
+      "learning_rate": 0.00011285889319687923,
+      "loss": 0.9534,
+      "step": 4818
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7150064706802368,
+      "learning_rate": 0.00011282908481802004,
+      "loss": 1.0732,
+      "step": 4819
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6949624419212341,
+      "learning_rate": 0.00011279927528012561,
+      "loss": 0.8743,
+      "step": 4820
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.858025074005127,
+      "learning_rate": 0.00011276946458588917,
+      "loss": 0.9061,
+      "step": 4821
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.8336944580078125,
+      "learning_rate": 0.00011273965273800391,
+      "loss": 0.8504,
+      "step": 4822
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.8705890774726868,
+      "learning_rate": 0.00011270983973916314,
+      "loss": 1.0703,
+      "step": 4823
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.9273001551628113,
+      "learning_rate": 0.00011268002559206034,
+      "loss": 1.037,
+      "step": 4824
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.8158978223800659,
+      "learning_rate": 0.00011265021029938897,
+      "loss": 0.9633,
+      "step": 4825
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7523478269577026,
+      "learning_rate": 0.00011262039386384277,
+      "loss": 0.9112,
+      "step": 4826
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7214547395706177,
+      "learning_rate": 0.00011259057628811542,
+      "loss": 1.0415,
+      "step": 4827
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7658652663230896,
+      "learning_rate": 0.00011256075757490081,
+      "loss": 0.8409,
+      "step": 4828
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7982742190361023,
+      "learning_rate": 0.00011253093772689288,
+      "loss": 1.036,
+      "step": 4829
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7945194244384766,
+      "learning_rate": 0.00011250111674678563,
+      "loss": 0.8592,
+      "step": 4830
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7518361806869507,
+      "learning_rate": 0.00011247129463727328,
+      "loss": 0.9347,
+      "step": 4831
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.8912938833236694,
+      "learning_rate": 0.00011244147140105007,
+      "loss": 0.8304,
+      "step": 4832
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.731968104839325,
+      "learning_rate": 0.00011241164704081038,
+      "loss": 1.009,
+      "step": 4833
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.8283110857009888,
+      "learning_rate": 0.00011238182155924866,
+      "loss": 0.9884,
+      "step": 4834
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7054871320724487,
+      "learning_rate": 0.00011235199495905943,
+      "loss": 1.0458,
+      "step": 4835
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6782692670822144,
+      "learning_rate": 0.00011232216724293742,
+      "loss": 0.9012,
+      "step": 4836
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7592952251434326,
+      "learning_rate": 0.00011229233841357738,
+      "loss": 0.9308,
+      "step": 4837
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7900505065917969,
+      "learning_rate": 0.00011226250847367417,
+      "loss": 0.9745,
+      "step": 4838
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.729471743106842,
+      "learning_rate": 0.00011223267742592272,
+      "loss": 0.906,
+      "step": 4839
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7428840398788452,
+      "learning_rate": 0.00011220284527301812,
+      "loss": 1.0044,
+      "step": 4840
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7849833369255066,
+      "learning_rate": 0.00011217301201765556,
+      "loss": 1.0778,
+      "step": 4841
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7510427832603455,
+      "learning_rate": 0.00011214317766253026,
+      "loss": 1.0539,
+      "step": 4842
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7398820519447327,
+      "learning_rate": 0.00011211334221033767,
+      "loss": 0.9486,
+      "step": 4843
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7769248485565186,
+      "learning_rate": 0.00011208350566377315,
+      "loss": 0.9762,
+      "step": 4844
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.8509570360183716,
+      "learning_rate": 0.0001120536680255323,
+      "loss": 1.0259,
+      "step": 4845
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7902756929397583,
+      "learning_rate": 0.00011202382929831083,
+      "loss": 0.9836,
+      "step": 4846
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7448802590370178,
+      "learning_rate": 0.00011199398948480445,
+      "loss": 0.9311,
+      "step": 4847
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7944951057434082,
+      "learning_rate": 0.00011196414858770903,
+      "loss": 0.9934,
+      "step": 4848
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7836403846740723,
+      "learning_rate": 0.00011193430660972051,
+      "loss": 0.9239,
+      "step": 4849
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.8021852970123291,
+      "learning_rate": 0.00011190446355353497,
+      "loss": 0.9101,
+      "step": 4850
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.8140667080879211,
+      "learning_rate": 0.00011187461942184856,
+      "loss": 1.0416,
+      "step": 4851
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7513343691825867,
+      "learning_rate": 0.00011184477421735748,
+      "loss": 0.9801,
+      "step": 4852
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7877258658409119,
+      "learning_rate": 0.00011181492794275817,
+      "loss": 1.0065,
+      "step": 4853
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7535899877548218,
+      "learning_rate": 0.000111785080600747,
+      "loss": 0.8739,
+      "step": 4854
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7377232909202576,
+      "learning_rate": 0.00011175523219402048,
+      "loss": 0.7785,
+      "step": 4855
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7618876099586487,
+      "learning_rate": 0.00011172538272527535,
+      "loss": 0.9044,
+      "step": 4856
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.8464749455451965,
+      "learning_rate": 0.00011169553219720828,
+      "loss": 0.9384,
+      "step": 4857
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.8161110877990723,
+      "learning_rate": 0.00011166568061251605,
+      "loss": 1.0834,
+      "step": 4858
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.772560715675354,
+      "learning_rate": 0.00011163582797389567,
+      "loss": 0.894,
+      "step": 4859
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.8467875123023987,
+      "learning_rate": 0.00011160597428404407,
+      "loss": 0.9931,
+      "step": 4860
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7472082376480103,
+      "learning_rate": 0.00011157611954565844,
+      "loss": 0.8825,
+      "step": 4861
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.790768563747406,
+      "learning_rate": 0.00011154626376143593,
+      "loss": 0.933,
+      "step": 4862
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7249228358268738,
+      "learning_rate": 0.00011151640693407387,
+      "loss": 0.9692,
+      "step": 4863
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7793048620223999,
+      "learning_rate": 0.00011148654906626967,
+      "loss": 0.9523,
+      "step": 4864
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.8146613240242004,
+      "learning_rate": 0.00011145669016072078,
+      "loss": 0.9364,
+      "step": 4865
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6911619305610657,
+      "learning_rate": 0.00011142683022012478,
+      "loss": 0.932,
+      "step": 4866
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.9315606355667114,
+      "learning_rate": 0.00011139696924717941,
+      "loss": 0.8802,
+      "step": 4867
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.786227822303772,
+      "learning_rate": 0.00011136710724458238,
+      "loss": 1.0699,
+      "step": 4868
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7897815704345703,
+      "learning_rate": 0.00011133724421503157,
+      "loss": 0.9314,
+      "step": 4869
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.8168982267379761,
+      "learning_rate": 0.00011130738016122495,
+      "loss": 0.8716,
+      "step": 4870
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6992068886756897,
+      "learning_rate": 0.00011127751508586053,
+      "loss": 0.8776,
+      "step": 4871
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7909550666809082,
+      "learning_rate": 0.00011124764899163649,
+      "loss": 1.0275,
+      "step": 4872
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6491546630859375,
+      "learning_rate": 0.00011121778188125107,
+      "loss": 0.8812,
+      "step": 4873
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.8003860116004944,
+      "learning_rate": 0.00011118791375740254,
+      "loss": 1.0509,
+      "step": 4874
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7199599146842957,
+      "learning_rate": 0.00011115804462278933,
+      "loss": 0.9646,
+      "step": 4875
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.824131965637207,
+      "learning_rate": 0.00011112817448010999,
+      "loss": 0.8921,
+      "step": 4876
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7687273025512695,
+      "learning_rate": 0.00011109830333206311,
+      "loss": 0.883,
+      "step": 4877
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7793501019477844,
+      "learning_rate": 0.00011106843118134736,
+      "loss": 0.964,
+      "step": 4878
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7844024896621704,
+      "learning_rate": 0.0001110385580306615,
+      "loss": 1.0223,
+      "step": 4879
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7543127536773682,
+      "learning_rate": 0.00011100868388270442,
+      "loss": 1.0382,
+      "step": 4880
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.8293694853782654,
+      "learning_rate": 0.0001109788087401751,
+      "loss": 1.0587,
+      "step": 4881
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6826415061950684,
+      "learning_rate": 0.00011094893260577256,
+      "loss": 0.8231,
+      "step": 4882
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.746153712272644,
+      "learning_rate": 0.00011091905548219597,
+      "loss": 0.9327,
+      "step": 4883
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.8294541239738464,
+      "learning_rate": 0.00011088917737214452,
+      "loss": 0.9046,
+      "step": 4884
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.7316751480102539,
+      "learning_rate": 0.00011085929827831754,
+      "loss": 0.9777,
+      "step": 4885
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.8139753937721252,
+      "learning_rate": 0.00011082941820341448,
+      "loss": 0.9985,
+      "step": 4886
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.7943364381790161,
+      "learning_rate": 0.00011079953715013478,
+      "loss": 1.0446,
+      "step": 4887
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.7968236207962036,
+      "learning_rate": 0.00011076965512117805,
+      "loss": 1.1082,
+      "step": 4888
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.8494092226028442,
+      "learning_rate": 0.00011073977211924397,
+      "loss": 0.8863,
+      "step": 4889
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.90047287940979,
+      "learning_rate": 0.00011070988814703225,
+      "loss": 1.104,
+      "step": 4890
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.7975635528564453,
+      "learning_rate": 0.00011068000320724281,
+      "loss": 1.0129,
+      "step": 4891
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.7787724733352661,
+      "learning_rate": 0.00011065011730257552,
+      "loss": 0.9088,
+      "step": 4892
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6692328453063965,
+      "learning_rate": 0.00011062023043573047,
+      "loss": 0.9661,
+      "step": 4893
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.82075434923172,
+      "learning_rate": 0.00011059034260940773,
+      "loss": 0.941,
+      "step": 4894
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.7842402458190918,
+      "learning_rate": 0.00011056045382630744,
+      "loss": 1.0,
+      "step": 4895
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.7709221839904785,
+      "learning_rate": 0.00011053056408913001,
+      "loss": 0.876,
+      "step": 4896
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6689760684967041,
+      "learning_rate": 0.0001105006734005757,
+      "loss": 0.8053,
+      "step": 4897
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.7427790760993958,
+      "learning_rate": 0.00011047078176334502,
+      "loss": 0.9647,
+      "step": 4898
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.8376590013504028,
+      "learning_rate": 0.00011044088918013849,
+      "loss": 0.9913,
+      "step": 4899
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.7888361215591431,
+      "learning_rate": 0.00011041099565365674,
+      "loss": 0.8347,
+      "step": 4900
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.8089134097099304,
+      "learning_rate": 0.00011038110118660045,
+      "loss": 1.1087,
+      "step": 4901
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.9311762452125549,
+      "learning_rate": 0.00011035120578167049,
+      "loss": 1.0656,
+      "step": 4902
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.7809112071990967,
+      "learning_rate": 0.00011032130944156767,
+      "loss": 1.0239,
+      "step": 4903
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6649876832962036,
+      "learning_rate": 0.00011029141216899299,
+      "loss": 0.8859,
+      "step": 4904
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.776587724685669,
+      "learning_rate": 0.00011026151396664747,
+      "loss": 0.9574,
+      "step": 4905
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.9212097525596619,
+      "learning_rate": 0.00011023161483723228,
+      "loss": 0.9127,
+      "step": 4906
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.9740622043609619,
+      "learning_rate": 0.00011020171478344863,
+      "loss": 1.1391,
+      "step": 4907
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.8400608897209167,
+      "learning_rate": 0.0001101718138079978,
+      "loss": 0.9239,
+      "step": 4908
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7605606317520142,
+      "learning_rate": 0.00011014191191358119,
+      "loss": 1.0455,
+      "step": 4909
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7030631899833679,
+      "learning_rate": 0.00011011200910290024,
+      "loss": 1.0018,
+      "step": 4910
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.8395674228668213,
+      "learning_rate": 0.00011008210537865653,
+      "loss": 0.9589,
+      "step": 4911
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.8159804940223694,
+      "learning_rate": 0.00011005220074355171,
+      "loss": 0.9439,
+      "step": 4912
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7161035537719727,
+      "learning_rate": 0.00011002229520028746,
+      "loss": 0.7207,
+      "step": 4913
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7870673537254333,
+      "learning_rate": 0.00010999238875156559,
+      "loss": 1.1633,
+      "step": 4914
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6282675266265869,
+      "learning_rate": 0.00010996248140008794,
+      "loss": 0.8594,
+      "step": 4915
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.8430740833282471,
+      "learning_rate": 0.00010993257314855653,
+      "loss": 0.9314,
+      "step": 4916
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6887407302856445,
+      "learning_rate": 0.00010990266399967337,
+      "loss": 0.8555,
+      "step": 4917
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7878185510635376,
+      "learning_rate": 0.00010987275395614058,
+      "loss": 1.0859,
+      "step": 4918
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7717889547348022,
+      "learning_rate": 0.00010984284302066037,
+      "loss": 1.0218,
+      "step": 4919
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.771571934223175,
+      "learning_rate": 0.00010981293119593503,
+      "loss": 1.0549,
+      "step": 4920
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7822331786155701,
+      "learning_rate": 0.00010978301848466692,
+      "loss": 1.0124,
+      "step": 4921
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.9212383031845093,
+      "learning_rate": 0.00010975310488955846,
+      "loss": 1.082,
+      "step": 4922
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7388486862182617,
+      "learning_rate": 0.00010972319041331223,
+      "loss": 0.9977,
+      "step": 4923
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7861277461051941,
+      "learning_rate": 0.00010969327505863076,
+      "loss": 0.9302,
+      "step": 4924
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.8004422187805176,
+      "learning_rate": 0.00010966335882821678,
+      "loss": 0.9582,
+      "step": 4925
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7476869821548462,
+      "learning_rate": 0.00010963344172477308,
+      "loss": 1.1044,
+      "step": 4926
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.813045084476471,
+      "learning_rate": 0.00010960352375100242,
+      "loss": 1.0339,
+      "step": 4927
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.9535208344459534,
+      "learning_rate": 0.0001095736049096078,
+      "loss": 0.9714,
+      "step": 4928
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.8231042623519897,
+      "learning_rate": 0.00010954368520329217,
+      "loss": 0.9928,
+      "step": 4929
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7152572870254517,
+      "learning_rate": 0.00010951376463475858,
+      "loss": 0.8452,
+      "step": 4930
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.711341142654419,
+      "learning_rate": 0.00010948384320671027,
+      "loss": 0.9219,
+      "step": 4931
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7866460680961609,
+      "learning_rate": 0.00010945392092185043,
+      "loss": 1.0716,
+      "step": 4932
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.7420953512191772,
+      "learning_rate": 0.00010942399778288233,
+      "loss": 0.8871,
+      "step": 4933
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.8441842794418335,
+      "learning_rate": 0.0001093940737925094,
+      "loss": 1.0337,
+      "step": 4934
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.7497798800468445,
+      "learning_rate": 0.0001093641489534351,
+      "loss": 0.9642,
+      "step": 4935
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.9009296894073486,
+      "learning_rate": 0.00010933422326836294,
+      "loss": 0.8992,
+      "step": 4936
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.8999432325363159,
+      "learning_rate": 0.00010930429673999658,
+      "loss": 1.135,
+      "step": 4937
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.9044147729873657,
+      "learning_rate": 0.00010927436937103968,
+      "loss": 0.9029,
+      "step": 4938
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.7883457541465759,
+      "learning_rate": 0.00010924444116419604,
+      "loss": 0.9975,
+      "step": 4939
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.7764599919319153,
+      "learning_rate": 0.00010921451212216945,
+      "loss": 0.8132,
+      "step": 4940
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.7747498750686646,
+      "learning_rate": 0.00010918458224766386,
+      "loss": 0.859,
+      "step": 4941
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.7943763136863708,
+      "learning_rate": 0.0001091546515433833,
+      "loss": 1.011,
+      "step": 4942
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.7181071043014526,
+      "learning_rate": 0.00010912472001203177,
+      "loss": 0.8989,
+      "step": 4943
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.7549866437911987,
+      "learning_rate": 0.00010909478765631349,
+      "loss": 1.0726,
+      "step": 4944
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.798565149307251,
+      "learning_rate": 0.0001090648544789326,
+      "loss": 1.0303,
+      "step": 4945
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.7307342290878296,
+      "learning_rate": 0.00010903492048259345,
+      "loss": 0.9789,
+      "step": 4946
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.7824006080627441,
+      "learning_rate": 0.00010900498567000039,
+      "loss": 0.9985,
+      "step": 4947
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.8141332268714905,
+      "learning_rate": 0.00010897505004385788,
+      "loss": 0.9185,
+      "step": 4948
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.8538034558296204,
+      "learning_rate": 0.00010894511360687039,
+      "loss": 0.9765,
+      "step": 4949
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.7719243764877319,
+      "learning_rate": 0.00010891517636174253,
+      "loss": 1.0315,
+      "step": 4950
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.8093717098236084,
+      "learning_rate": 0.00010888523831117899,
+      "loss": 1.0684,
+      "step": 4951
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.8469390869140625,
+      "learning_rate": 0.00010885529945788446,
+      "loss": 1.1159,
+      "step": 4952
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.8024957776069641,
+      "learning_rate": 0.00010882535980456376,
+      "loss": 1.1199,
+      "step": 4953
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6433467268943787,
+      "learning_rate": 0.00010879541935392176,
+      "loss": 0.8979,
+      "step": 4954
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.7785211205482483,
+      "learning_rate": 0.00010876547810866344,
+      "loss": 1.0434,
+      "step": 4955
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.8724049925804138,
+      "learning_rate": 0.0001087355360714938,
+      "loss": 0.8796,
+      "step": 4956
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.7402239441871643,
+      "learning_rate": 0.00010870559324511793,
+      "loss": 0.9601,
+      "step": 4957
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.8949133157730103,
+      "learning_rate": 0.00010867564963224101,
+      "loss": 1.026,
+      "step": 4958
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6975107192993164,
+      "learning_rate": 0.00010864570523556827,
+      "loss": 0.9029,
+      "step": 4959
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.9479396939277649,
+      "learning_rate": 0.000108615760057805,
+      "loss": 1.0211,
+      "step": 4960
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.8513799905776978,
+      "learning_rate": 0.0001085858141016566,
+      "loss": 1.045,
+      "step": 4961
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.8274711966514587,
+      "learning_rate": 0.0001085558673698285,
+      "loss": 0.9805,
+      "step": 4962
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.720975399017334,
+      "learning_rate": 0.00010852591986502625,
+      "loss": 0.9955,
+      "step": 4963
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.8408166766166687,
+      "learning_rate": 0.00010849597158995542,
+      "loss": 0.8946,
+      "step": 4964
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.7662861943244934,
+      "learning_rate": 0.00010846602254732158,
+      "loss": 1.1358,
+      "step": 4965
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.8497337102890015,
+      "learning_rate": 0.00010843607273983063,
+      "loss": 0.9136,
+      "step": 4966
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.0238080024719238,
+      "learning_rate": 0.00010840612217018822,
+      "loss": 0.963,
+      "step": 4967
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6822003722190857,
+      "learning_rate": 0.0001083761708411003,
+      "loss": 0.6186,
+      "step": 4968
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.7723561525344849,
+      "learning_rate": 0.00010834621875527278,
+      "loss": 1.0276,
+      "step": 4969
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.792113721370697,
+      "learning_rate": 0.00010831626591541157,
+      "loss": 1.037,
+      "step": 4970
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.7487592101097107,
+      "learning_rate": 0.00010828631232422291,
+      "loss": 0.9075,
+      "step": 4971
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.8601886630058289,
+      "learning_rate": 0.00010825635798441283,
+      "loss": 0.8613,
+      "step": 4972
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.8530653119087219,
+      "learning_rate": 0.00010822640289868753,
+      "loss": 1.0797,
+      "step": 4973
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.7469228506088257,
+      "learning_rate": 0.00010819644706975332,
+      "loss": 0.8785,
+      "step": 4974
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.7503280639648438,
+      "learning_rate": 0.0001081664905003165,
+      "loss": 1.0749,
+      "step": 4975
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.7468873262405396,
+      "learning_rate": 0.00010813653319308353,
+      "loss": 1.1124,
+      "step": 4976
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.7453671097755432,
+      "learning_rate": 0.00010810657515076086,
+      "loss": 0.916,
+      "step": 4977
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.7897058725357056,
+      "learning_rate": 0.000108076616376055,
+      "loss": 0.947,
+      "step": 4978
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.804443895816803,
+      "learning_rate": 0.00010804665687167262,
+      "loss": 1.016,
+      "step": 4979
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6845722794532776,
+      "learning_rate": 0.0001080166966403203,
+      "loss": 0.8888,
+      "step": 4980
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.7716487646102905,
+      "learning_rate": 0.00010798673568470485,
+      "loss": 1.0317,
+      "step": 4981
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.7068995833396912,
+      "learning_rate": 0.00010795677400753306,
+      "loss": 0.9932,
+      "step": 4982
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.8245829343795776,
+      "learning_rate": 0.00010792681161151177,
+      "loss": 0.9657,
+      "step": 4983
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.9497877359390259,
+      "learning_rate": 0.00010789684849934796,
+      "loss": 1.0754,
+      "step": 4984
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7938148975372314,
+      "learning_rate": 0.00010786688467374859,
+      "loss": 1.02,
+      "step": 4985
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7475448846817017,
+      "learning_rate": 0.0001078369201374207,
+      "loss": 1.0186,
+      "step": 4986
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7335013747215271,
+      "learning_rate": 0.0001078069548930715,
+      "loss": 0.9782,
+      "step": 4987
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.8388778567314148,
+      "learning_rate": 0.00010777698894340813,
+      "loss": 1.0908,
+      "step": 4988
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.8602745532989502,
+      "learning_rate": 0.0001077470222911378,
+      "loss": 0.8835,
+      "step": 4989
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7140038013458252,
+      "learning_rate": 0.00010771705493896788,
+      "loss": 0.8857,
+      "step": 4990
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.8438007831573486,
+      "learning_rate": 0.00010768708688960575,
+      "loss": 1.0315,
+      "step": 4991
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7744883298873901,
+      "learning_rate": 0.00010765711814575883,
+      "loss": 1.0601,
+      "step": 4992
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.8040426969528198,
+      "learning_rate": 0.00010762714871013467,
+      "loss": 0.9017,
+      "step": 4993
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.8766114711761475,
+      "learning_rate": 0.00010759717858544078,
+      "loss": 0.8371,
+      "step": 4994
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7012424468994141,
+      "learning_rate": 0.00010756720777438482,
+      "loss": 0.9728,
+      "step": 4995
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7083522081375122,
+      "learning_rate": 0.00010753723627967451,
+      "loss": 0.966,
+      "step": 4996
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.8759257197380066,
+      "learning_rate": 0.00010750726410401754,
+      "loss": 1.1248,
+      "step": 4997
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.9319406747817993,
+      "learning_rate": 0.0001074772912501218,
+      "loss": 1.1054,
+      "step": 4998
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.8455122709274292,
+      "learning_rate": 0.00010744731772069512,
+      "loss": 1.0103,
+      "step": 4999
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7683991193771362,
+      "learning_rate": 0.00010741734351844544,
+      "loss": 0.9358,
+      "step": 5000
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7314528226852417,
+      "learning_rate": 0.00010738736864608079,
+      "loss": 0.8393,
+      "step": 5001
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7942683696746826,
+      "learning_rate": 0.00010735739310630916,
+      "loss": 0.9442,
+      "step": 5002
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.0038470029830933,
+      "learning_rate": 0.00010732741690183879,
+      "loss": 1.0515,
+      "step": 5003
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7769905924797058,
+      "learning_rate": 0.00010729744003537776,
+      "loss": 0.973,
+      "step": 5004
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7850512862205505,
+      "learning_rate": 0.00010726746250963431,
+      "loss": 0.9856,
+      "step": 5005
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.8909277319908142,
+      "learning_rate": 0.00010723748432731682,
+      "loss": 0.9844,
+      "step": 5006
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7409754991531372,
+      "learning_rate": 0.00010720750549113359,
+      "loss": 1.0171,
+      "step": 5007
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7498346567153931,
+      "learning_rate": 0.00010717752600379303,
+      "loss": 1.1193,
+      "step": 5008
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7579041123390198,
+      "learning_rate": 0.00010714754586800364,
+      "loss": 0.9046,
+      "step": 5009
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.8222535252571106,
+      "learning_rate": 0.00010711756508647395,
+      "loss": 0.9735,
+      "step": 5010
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6744101643562317,
+      "learning_rate": 0.00010708758366191254,
+      "loss": 0.9995,
+      "step": 5011
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7639139294624329,
+      "learning_rate": 0.00010705760159702811,
+      "loss": 0.9265,
+      "step": 5012
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7082836627960205,
+      "learning_rate": 0.0001070276188945293,
+      "loss": 0.831,
+      "step": 5013
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7807806730270386,
+      "learning_rate": 0.00010699763555712495,
+      "loss": 0.8165,
+      "step": 5014
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.8013209104537964,
+      "learning_rate": 0.00010696765158752383,
+      "loss": 0.9743,
+      "step": 5015
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7248284816741943,
+      "learning_rate": 0.00010693766698843482,
+      "loss": 0.9452,
+      "step": 5016
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7307727336883545,
+      "learning_rate": 0.00010690768176256689,
+      "loss": 0.9154,
+      "step": 5017
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.8232489824295044,
+      "learning_rate": 0.00010687769591262904,
+      "loss": 1.0497,
+      "step": 5018
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.945981502532959,
+      "learning_rate": 0.00010684770944133031,
+      "loss": 0.92,
+      "step": 5019
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.828311562538147,
+      "learning_rate": 0.00010681772235137977,
+      "loss": 0.9855,
+      "step": 5020
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7558428645133972,
+      "learning_rate": 0.00010678773464548663,
+      "loss": 0.8314,
+      "step": 5021
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.9145514369010925,
+      "learning_rate": 0.00010675774632636011,
+      "loss": 1.0534,
+      "step": 5022
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.711399257183075,
+      "learning_rate": 0.00010672775739670949,
+      "loss": 0.9327,
+      "step": 5023
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.89848792552948,
+      "learning_rate": 0.00010669776785924404,
+      "loss": 1.046,
+      "step": 5024
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.8986443877220154,
+      "learning_rate": 0.00010666777771667321,
+      "loss": 1.1313,
+      "step": 5025
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7316709756851196,
+      "learning_rate": 0.0001066377869717064,
+      "loss": 0.9042,
+      "step": 5026
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.8240397572517395,
+      "learning_rate": 0.00010660779562705312,
+      "loss": 1.1324,
+      "step": 5027
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.8348107933998108,
+      "learning_rate": 0.00010657780368542293,
+      "loss": 1.0284,
+      "step": 5028
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7479124665260315,
+      "learning_rate": 0.00010654781114952543,
+      "loss": 1.0645,
+      "step": 5029
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7343984246253967,
+      "learning_rate": 0.00010651781802207023,
+      "loss": 0.8719,
+      "step": 5030
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.2171791791915894,
+      "learning_rate": 0.00010648782430576713,
+      "loss": 0.9959,
+      "step": 5031
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.8316954374313354,
+      "learning_rate": 0.00010645783000332578,
+      "loss": 1.0267,
+      "step": 5032
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.8453816771507263,
+      "learning_rate": 0.00010642783511745611,
+      "loss": 1.0098,
+      "step": 5033
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.722510814666748,
+      "learning_rate": 0.0001063978396508679,
+      "loss": 0.7125,
+      "step": 5034
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.8132488131523132,
+      "learning_rate": 0.00010636784360627108,
+      "loss": 1.1609,
+      "step": 5035
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7693787217140198,
+      "learning_rate": 0.0001063378469863757,
+      "loss": 1.0599,
+      "step": 5036
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6765629053115845,
+      "learning_rate": 0.00010630784979389168,
+      "loss": 0.9047,
+      "step": 5037
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.8674545288085938,
+      "learning_rate": 0.00010627785203152918,
+      "loss": 1.0258,
+      "step": 5038
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.982782244682312,
+      "learning_rate": 0.0001062478537019983,
+      "loss": 0.962,
+      "step": 5039
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7085098624229431,
+      "learning_rate": 0.00010621785480800914,
+      "loss": 0.9935,
+      "step": 5040
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.9388718605041504,
+      "learning_rate": 0.00010618785535227209,
+      "loss": 1.1226,
+      "step": 5041
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.8830339312553406,
+      "learning_rate": 0.00010615785533749729,
+      "loss": 1.2911,
+      "step": 5042
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7700386643409729,
+      "learning_rate": 0.00010612785476639518,
+      "loss": 0.8633,
+      "step": 5043
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7826847434043884,
+      "learning_rate": 0.00010609785364167606,
+      "loss": 0.9842,
+      "step": 5044
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.9161370396614075,
+      "learning_rate": 0.00010606785196605033,
+      "loss": 1.0458,
+      "step": 5045
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.8588224053382874,
+      "learning_rate": 0.00010603784974222861,
+      "loss": 0.9512,
+      "step": 5046
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7238483428955078,
+      "learning_rate": 0.00010600784697292134,
+      "loss": 0.838,
+      "step": 5047
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7885895371437073,
+      "learning_rate": 0.0001059778436608391,
+      "loss": 1.0034,
+      "step": 5048
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7344606518745422,
+      "learning_rate": 0.00010594783980869254,
+      "loss": 0.9331,
+      "step": 5049
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7655723690986633,
+      "learning_rate": 0.00010591783541919232,
+      "loss": 0.8249,
+      "step": 5050
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7848523259162903,
+      "learning_rate": 0.00010588783049504915,
+      "loss": 0.9683,
+      "step": 5051
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7566481828689575,
+      "learning_rate": 0.00010585782503897388,
+      "loss": 0.9803,
+      "step": 5052
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.9054071307182312,
+      "learning_rate": 0.00010582781905367725,
+      "loss": 0.9562,
+      "step": 5053
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7615681290626526,
+      "learning_rate": 0.0001057978125418702,
+      "loss": 0.9088,
+      "step": 5054
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7379032373428345,
+      "learning_rate": 0.00010576780550626357,
+      "loss": 0.9506,
+      "step": 5055
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.8043782711029053,
+      "learning_rate": 0.0001057377979495684,
+      "loss": 0.9858,
+      "step": 5056
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.8147229552268982,
+      "learning_rate": 0.00010570778987449568,
+      "loss": 1.0159,
+      "step": 5057
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.8456796407699585,
+      "learning_rate": 0.00010567778128375642,
+      "loss": 1.0781,
+      "step": 5058
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7417508363723755,
+      "learning_rate": 0.00010564777218006181,
+      "loss": 1.003,
+      "step": 5059
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7917277812957764,
+      "learning_rate": 0.00010561776256612295,
+      "loss": 0.9502,
+      "step": 5060
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.8203314542770386,
+      "learning_rate": 0.00010558775244465102,
+      "loss": 0.8545,
+      "step": 5061
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.869928777217865,
+      "learning_rate": 0.00010555774181835733,
+      "loss": 1.0155,
+      "step": 5062
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.8022637367248535,
+      "learning_rate": 0.00010552773068995314,
+      "loss": 1.0879,
+      "step": 5063
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.672001838684082,
+      "learning_rate": 0.00010549771906214979,
+      "loss": 0.9135,
+      "step": 5064
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7225290536880493,
+      "learning_rate": 0.0001054677069376586,
+      "loss": 0.7734,
+      "step": 5065
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.8035221099853516,
+      "learning_rate": 0.00010543769431919108,
+      "loss": 0.9138,
+      "step": 5066
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.8685542941093445,
+      "learning_rate": 0.00010540768120945865,
+      "loss": 1.0208,
+      "step": 5067
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7406560778617859,
+      "learning_rate": 0.00010537766761117286,
+      "loss": 0.9031,
+      "step": 5068
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7587171196937561,
+      "learning_rate": 0.00010534765352704523,
+      "loss": 0.8852,
+      "step": 5069
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6719124913215637,
+      "learning_rate": 0.00010531763895978739,
+      "loss": 0.8829,
+      "step": 5070
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.8159770965576172,
+      "learning_rate": 0.000105287623912111,
+      "loss": 0.9413,
+      "step": 5071
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.8835862874984741,
+      "learning_rate": 0.00010525760838672769,
+      "loss": 1.1545,
+      "step": 5072
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.8177557587623596,
+      "learning_rate": 0.00010522759238634927,
+      "loss": 0.8752,
+      "step": 5073
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7685697674751282,
+      "learning_rate": 0.00010519757591368744,
+      "loss": 0.9306,
+      "step": 5074
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.8134411573410034,
+      "learning_rate": 0.00010516755897145408,
+      "loss": 0.7861,
+      "step": 5075
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.8483935594558716,
+      "learning_rate": 0.00010513754156236102,
+      "loss": 0.9184,
+      "step": 5076
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.9010506868362427,
+      "learning_rate": 0.00010510752368912014,
+      "loss": 0.9618,
+      "step": 5077
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7965332269668579,
+      "learning_rate": 0.00010507750535444346,
+      "loss": 1.0317,
+      "step": 5078
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6694085597991943,
+      "learning_rate": 0.0001050474865610429,
+      "loss": 0.8317,
+      "step": 5079
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7621757388114929,
+      "learning_rate": 0.00010501746731163049,
+      "loss": 0.9535,
+      "step": 5080
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7820752859115601,
+      "learning_rate": 0.00010498744760891834,
+      "loss": 0.9037,
+      "step": 5081
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7151625156402588,
+      "learning_rate": 0.00010495742745561854,
+      "loss": 0.8744,
+      "step": 5082
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.742006778717041,
+      "learning_rate": 0.00010492740685444322,
+      "loss": 1.0673,
+      "step": 5083
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.8351051807403564,
+      "learning_rate": 0.00010489738580810462,
+      "loss": 1.0883,
+      "step": 5084
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6983124613761902,
+      "learning_rate": 0.0001048673643193149,
+      "loss": 0.9804,
+      "step": 5085
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.8488489985466003,
+      "learning_rate": 0.0001048373423907864,
+      "loss": 1.0387,
+      "step": 5086
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.8636870980262756,
+      "learning_rate": 0.00010480732002523139,
+      "loss": 1.0967,
+      "step": 5087
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7339608073234558,
+      "learning_rate": 0.00010477729722536225,
+      "loss": 0.9628,
+      "step": 5088
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7165808081626892,
+      "learning_rate": 0.00010474727399389134,
+      "loss": 0.8708,
+      "step": 5089
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6733713746070862,
+      "learning_rate": 0.0001047172503335311,
+      "loss": 0.9271,
+      "step": 5090
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6857487559318542,
+      "learning_rate": 0.00010468722624699401,
+      "loss": 0.9018,
+      "step": 5091
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7812450528144836,
+      "learning_rate": 0.00010465720173699259,
+      "loss": 0.9974,
+      "step": 5092
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.8059720396995544,
+      "learning_rate": 0.00010462717680623933,
+      "loss": 0.8986,
+      "step": 5093
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7781034708023071,
+      "learning_rate": 0.00010459715145744686,
+      "loss": 1.031,
+      "step": 5094
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7936933040618896,
+      "learning_rate": 0.00010456712569332777,
+      "loss": 0.9529,
+      "step": 5095
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7723500728607178,
+      "learning_rate": 0.00010453709951659475,
+      "loss": 1.0107,
+      "step": 5096
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6995753049850464,
+      "learning_rate": 0.00010450707292996047,
+      "loss": 0.9326,
+      "step": 5097
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.8494675159454346,
+      "learning_rate": 0.00010447704593613768,
+      "loss": 1.0187,
+      "step": 5098
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.855739176273346,
+      "learning_rate": 0.00010444701853783911,
+      "loss": 1.005,
+      "step": 5099
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7245144248008728,
+      "learning_rate": 0.00010441699073777762,
+      "loss": 0.8173,
+      "step": 5100
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.8450302481651306,
+      "learning_rate": 0.00010438696253866603,
+      "loss": 0.8375,
+      "step": 5101
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.767345666885376,
+      "learning_rate": 0.00010435693394321717,
+      "loss": 0.9019,
+      "step": 5102
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.8089317083358765,
+      "learning_rate": 0.00010432690495414405,
+      "loss": 0.9923,
+      "step": 5103
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7338062524795532,
+      "learning_rate": 0.00010429687557415954,
+      "loss": 0.8316,
+      "step": 5104
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6969124674797058,
+      "learning_rate": 0.00010426684580597666,
+      "loss": 0.8215,
+      "step": 5105
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.8065308332443237,
+      "learning_rate": 0.0001042368156523084,
+      "loss": 1.0661,
+      "step": 5106
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7004128098487854,
+      "learning_rate": 0.00010420678511586785,
+      "loss": 0.7736,
+      "step": 5107
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7430830597877502,
+      "learning_rate": 0.00010417675419936809,
+      "loss": 0.9779,
+      "step": 5108
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.9451195597648621,
+      "learning_rate": 0.00010414672290552223,
+      "loss": 0.8763,
+      "step": 5109
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.8106129169464111,
+      "learning_rate": 0.00010411669123704344,
+      "loss": 1.0782,
+      "step": 5110
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6853575110435486,
+      "learning_rate": 0.00010408665919664491,
+      "loss": 0.8011,
+      "step": 5111
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.8419955968856812,
+      "learning_rate": 0.00010405662678703984,
+      "loss": 1.0292,
+      "step": 5112
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7453970313072205,
+      "learning_rate": 0.00010402659401094152,
+      "loss": 1.0889,
+      "step": 5113
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.819788932800293,
+      "learning_rate": 0.00010399656087106324,
+      "loss": 1.0317,
+      "step": 5114
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7040184140205383,
+      "learning_rate": 0.00010396652737011828,
+      "loss": 1.0628,
+      "step": 5115
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.740817666053772,
+      "learning_rate": 0.00010393649351082006,
+      "loss": 0.9928,
+      "step": 5116
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.8002640604972839,
+      "learning_rate": 0.00010390645929588196,
+      "loss": 1.0403,
+      "step": 5117
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7260842323303223,
+      "learning_rate": 0.00010387642472801733,
+      "loss": 0.9452,
+      "step": 5118
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7765595316886902,
+      "learning_rate": 0.00010384638980993972,
+      "loss": 1.0988,
+      "step": 5119
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.8337041139602661,
+      "learning_rate": 0.00010381635454436252,
+      "loss": 0.9593,
+      "step": 5120
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7491742968559265,
+      "learning_rate": 0.00010378631893399933,
+      "loss": 0.9047,
+      "step": 5121
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7119678258895874,
+      "learning_rate": 0.00010375628298156365,
+      "loss": 0.9697,
+      "step": 5122
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.8019363880157471,
+      "learning_rate": 0.00010372624668976909,
+      "loss": 0.8877,
+      "step": 5123
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6626631021499634,
+      "learning_rate": 0.00010369621006132923,
+      "loss": 0.651,
+      "step": 5124
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.8198673129081726,
+      "learning_rate": 0.0001036661730989577,
+      "loss": 0.9331,
+      "step": 5125
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7067912817001343,
+      "learning_rate": 0.0001036361358053682,
+      "loss": 1.0908,
+      "step": 5126
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7765764594078064,
+      "learning_rate": 0.00010360609818327443,
+      "loss": 0.8615,
+      "step": 5127
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.8399764895439148,
+      "learning_rate": 0.00010357606023539008,
+      "loss": 1.1321,
+      "step": 5128
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.8030591011047363,
+      "learning_rate": 0.00010354602196442897,
+      "loss": 0.7552,
+      "step": 5129
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7561306953430176,
+      "learning_rate": 0.00010351598337310482,
+      "loss": 0.9402,
+      "step": 5130
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7164328098297119,
+      "learning_rate": 0.00010348594446413148,
+      "loss": 0.9956,
+      "step": 5131
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7632815837860107,
+      "learning_rate": 0.00010345590524022283,
+      "loss": 0.9489,
+      "step": 5132
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7721334099769592,
+      "learning_rate": 0.00010342586570409267,
+      "loss": 0.7268,
+      "step": 5133
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.8274480700492859,
+      "learning_rate": 0.00010339582585845498,
+      "loss": 0.9895,
+      "step": 5134
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7295353412628174,
+      "learning_rate": 0.00010336578570602362,
+      "loss": 0.9588,
+      "step": 5135
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.811366617679596,
+      "learning_rate": 0.00010333574524951257,
+      "loss": 0.9832,
+      "step": 5136
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7662849426269531,
+      "learning_rate": 0.00010330570449163586,
+      "loss": 1.0832,
+      "step": 5137
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7018322944641113,
+      "learning_rate": 0.00010327566343510746,
+      "loss": 0.9579,
+      "step": 5138
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.8456294536590576,
+      "learning_rate": 0.00010324562208264139,
+      "loss": 0.9006,
+      "step": 5139
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7415403127670288,
+      "learning_rate": 0.00010321558043695174,
+      "loss": 0.8829,
+      "step": 5140
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.8214705586433411,
+      "learning_rate": 0.00010318553850075265,
+      "loss": 0.9886,
+      "step": 5141
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7233492732048035,
+      "learning_rate": 0.00010315549627675816,
+      "loss": 0.919,
+      "step": 5142
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.9265508055686951,
+      "learning_rate": 0.00010312545376768246,
+      "loss": 0.8932,
+      "step": 5143
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.8299009799957275,
+      "learning_rate": 0.00010309541097623967,
+      "loss": 0.8363,
+      "step": 5144
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6598271727561951,
+      "learning_rate": 0.00010306536790514406,
+      "loss": 0.6518,
+      "step": 5145
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.8086226582527161,
+      "learning_rate": 0.0001030353245571098,
+      "loss": 0.9596,
+      "step": 5146
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7078489661216736,
+      "learning_rate": 0.00010300528093485115,
+      "loss": 0.8122,
+      "step": 5147
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7355614900588989,
+      "learning_rate": 0.00010297523704108242,
+      "loss": 1.0471,
+      "step": 5148
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7539297938346863,
+      "learning_rate": 0.00010294519287851783,
+      "loss": 0.9966,
+      "step": 5149
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7892729043960571,
+      "learning_rate": 0.00010291514844987174,
+      "loss": 0.966,
+      "step": 5150
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.8298482298851013,
+      "learning_rate": 0.00010288510375785851,
+      "loss": 0.9127,
+      "step": 5151
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7779788374900818,
+      "learning_rate": 0.00010285505880519249,
+      "loss": 1.0337,
+      "step": 5152
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.8932836651802063,
+      "learning_rate": 0.00010282501359458807,
+      "loss": 1.1316,
+      "step": 5153
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7916525602340698,
+      "learning_rate": 0.00010279496812875969,
+      "loss": 0.9774,
+      "step": 5154
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7860032916069031,
+      "learning_rate": 0.0001027649224104217,
+      "loss": 0.9521,
+      "step": 5155
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7322259545326233,
+      "learning_rate": 0.00010273487644228868,
+      "loss": 0.9672,
+      "step": 5156
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.733378529548645,
+      "learning_rate": 0.00010270483022707506,
+      "loss": 0.9651,
+      "step": 5157
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7881492972373962,
+      "learning_rate": 0.00010267478376749531,
+      "loss": 1.0659,
+      "step": 5158
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.8630042672157288,
+      "learning_rate": 0.00010264473706626404,
+      "loss": 0.9151,
+      "step": 5159
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7404335737228394,
+      "learning_rate": 0.0001026146901260957,
+      "loss": 0.8849,
+      "step": 5160
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7069669365882874,
+      "learning_rate": 0.00010258464294970492,
+      "loss": 0.8823,
+      "step": 5161
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.755982518196106,
+      "learning_rate": 0.0001025545955398063,
+      "loss": 1.0988,
+      "step": 5162
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7744724750518799,
+      "learning_rate": 0.00010252454789911443,
+      "loss": 0.9114,
+      "step": 5163
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6935697793960571,
+      "learning_rate": 0.00010249450003034395,
+      "loss": 1.1287,
+      "step": 5164
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7566258907318115,
+      "learning_rate": 0.00010246445193620953,
+      "loss": 1.0492,
+      "step": 5165
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7999916672706604,
+      "learning_rate": 0.00010243440361942581,
+      "loss": 1.0886,
+      "step": 5166
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7703303098678589,
+      "learning_rate": 0.00010240435508270752,
+      "loss": 0.9508,
+      "step": 5167
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.8494589328765869,
+      "learning_rate": 0.00010237430632876934,
+      "loss": 0.9582,
+      "step": 5168
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.777059018611908,
+      "learning_rate": 0.00010234425736032607,
+      "loss": 1.1731,
+      "step": 5169
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6870027780532837,
+      "learning_rate": 0.00010231420818009238,
+      "loss": 0.956,
+      "step": 5170
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.8410665988922119,
+      "learning_rate": 0.00010228415879078308,
+      "loss": 1.1189,
+      "step": 5171
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.8162976503372192,
+      "learning_rate": 0.00010225410919511301,
+      "loss": 1.0337,
+      "step": 5172
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7400126457214355,
+      "learning_rate": 0.00010222405939579692,
+      "loss": 1.0547,
+      "step": 5173
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7200222015380859,
+      "learning_rate": 0.00010219400939554964,
+      "loss": 0.9803,
+      "step": 5174
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6938434839248657,
+      "learning_rate": 0.00010216395919708603,
+      "loss": 0.8514,
+      "step": 5175
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7400026321411133,
+      "learning_rate": 0.000102133908803121,
+      "loss": 1.0465,
+      "step": 5176
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.8585032224655151,
+      "learning_rate": 0.00010210385821636934,
+      "loss": 0.8917,
+      "step": 5177
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7792884111404419,
+      "learning_rate": 0.00010207380743954602,
+      "loss": 1.0266,
+      "step": 5178
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.8200801610946655,
+      "learning_rate": 0.00010204375647536595,
+      "loss": 0.9524,
+      "step": 5179
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.8211904764175415,
+      "learning_rate": 0.00010201370532654404,
+      "loss": 1.0111,
+      "step": 5180
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.8270883560180664,
+      "learning_rate": 0.00010198365399579528,
+      "loss": 1.0467,
+      "step": 5181
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.825343132019043,
+      "learning_rate": 0.00010195360248583458,
+      "loss": 0.9845,
+      "step": 5182
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.9144495725631714,
+      "learning_rate": 0.000101923550799377,
+      "loss": 0.9904,
+      "step": 5183
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6933146119117737,
+      "learning_rate": 0.00010189349893913745,
+      "loss": 0.9956,
+      "step": 5184
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7854787707328796,
+      "learning_rate": 0.00010186344690783101,
+      "loss": 0.9385,
+      "step": 5185
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.8274226188659668,
+      "learning_rate": 0.00010183339470817271,
+      "loss": 0.9576,
+      "step": 5186
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7524675726890564,
+      "learning_rate": 0.00010180334234287754,
+      "loss": 0.9829,
+      "step": 5187
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7909670472145081,
+      "learning_rate": 0.00010177328981466064,
+      "loss": 1.0907,
+      "step": 5188
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7706515192985535,
+      "learning_rate": 0.00010174323712623704,
+      "loss": 0.8728,
+      "step": 5189
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.837437629699707,
+      "learning_rate": 0.00010171318428032179,
+      "loss": 1.1172,
+      "step": 5190
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.70237797498703,
+      "learning_rate": 0.00010168313127963008,
+      "loss": 0.8799,
+      "step": 5191
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7743717432022095,
+      "learning_rate": 0.000101653078126877,
+      "loss": 0.9472,
+      "step": 5192
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.8090443015098572,
+      "learning_rate": 0.00010162302482477764,
+      "loss": 1.1221,
+      "step": 5193
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.8558246493339539,
+      "learning_rate": 0.0001015929713760472,
+      "loss": 0.9798,
+      "step": 5194
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6947144269943237,
+      "learning_rate": 0.00010156291778340079,
+      "loss": 0.7357,
+      "step": 5195
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.9039539694786072,
+      "learning_rate": 0.00010153286404955362,
+      "loss": 1.0181,
+      "step": 5196
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7144058346748352,
+      "learning_rate": 0.00010150281017722088,
+      "loss": 0.7697,
+      "step": 5197
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.8410497307777405,
+      "learning_rate": 0.00010147275616911774,
+      "loss": 0.9568,
+      "step": 5198
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.8819612860679626,
+      "learning_rate": 0.00010144270202795943,
+      "loss": 1.121,
+      "step": 5199
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.9397442936897278,
+      "learning_rate": 0.00010141264775646114,
+      "loss": 0.9298,
+      "step": 5200
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.8475532531738281,
+      "learning_rate": 0.0001013825933573381,
+      "loss": 0.9965,
+      "step": 5201
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.9101167917251587,
+      "learning_rate": 0.00010135253883330563,
+      "loss": 0.8776,
+      "step": 5202
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7638474702835083,
+      "learning_rate": 0.00010132248418707888,
+      "loss": 0.9963,
+      "step": 5203
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.9219881892204285,
+      "learning_rate": 0.00010129242942137322,
+      "loss": 0.9861,
+      "step": 5204
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7822911739349365,
+      "learning_rate": 0.00010126237453890386,
+      "loss": 1.101,
+      "step": 5205
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.8252258896827698,
+      "learning_rate": 0.0001012323195423861,
+      "loss": 0.9579,
+      "step": 5206
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.9163365960121155,
+      "learning_rate": 0.00010120226443453526,
+      "loss": 1.0257,
+      "step": 5207
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7574987411499023,
+      "learning_rate": 0.00010117220921806664,
+      "loss": 0.87,
+      "step": 5208
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.8726235628128052,
+      "learning_rate": 0.00010114215389569555,
+      "loss": 1.0485,
+      "step": 5209
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.8032419681549072,
+      "learning_rate": 0.00010111209847013732,
+      "loss": 1.1435,
+      "step": 5210
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.8374848961830139,
+      "learning_rate": 0.00010108204294410726,
+      "loss": 1.0055,
+      "step": 5211
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6941425800323486,
+      "learning_rate": 0.0001010519873203208,
+      "loss": 0.8382,
+      "step": 5212
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.87954181432724,
+      "learning_rate": 0.00010102193160149322,
+      "loss": 0.9534,
+      "step": 5213
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7242146730422974,
+      "learning_rate": 0.0001009918757903399,
+      "loss": 0.9834,
+      "step": 5214
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.8842020630836487,
+      "learning_rate": 0.0001009618198895762,
+      "loss": 1.0604,
+      "step": 5215
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.72428959608078,
+      "learning_rate": 0.00010093176390191757,
+      "loss": 0.8531,
+      "step": 5216
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.8273959755897522,
+      "learning_rate": 0.0001009017078300793,
+      "loss": 1.0667,
+      "step": 5217
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7096107602119446,
+      "learning_rate": 0.00010087165167677686,
+      "loss": 0.9236,
+      "step": 5218
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.8173868656158447,
+      "learning_rate": 0.00010084159544472562,
+      "loss": 0.9115,
+      "step": 5219
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.8997098207473755,
+      "learning_rate": 0.00010081153913664099,
+      "loss": 0.9685,
+      "step": 5220
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7394409775733948,
+      "learning_rate": 0.0001007814827552384,
+      "loss": 0.8171,
+      "step": 5221
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.8413199782371521,
+      "learning_rate": 0.00010075142630323324,
+      "loss": 0.8588,
+      "step": 5222
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7438793182373047,
+      "learning_rate": 0.00010072136978334102,
+      "loss": 0.8727,
+      "step": 5223
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.8331936001777649,
+      "learning_rate": 0.00010069131319827709,
+      "loss": 1.1788,
+      "step": 5224
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7141227722167969,
+      "learning_rate": 0.0001006612565507569,
+      "loss": 0.6969,
+      "step": 5225
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7840338349342346,
+      "learning_rate": 0.00010063119984349598,
+      "loss": 0.9761,
+      "step": 5226
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.9646100997924805,
+      "learning_rate": 0.00010060114307920972,
+      "loss": 1.2251,
+      "step": 5227
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.8005926012992859,
+      "learning_rate": 0.00010057108626061358,
+      "loss": 0.9434,
+      "step": 5228
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.8847595453262329,
+      "learning_rate": 0.00010054102939042302,
+      "loss": 1.1027,
+      "step": 5229
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.792830228805542,
+      "learning_rate": 0.00010051097247135351,
+      "loss": 0.9165,
+      "step": 5230
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.663872241973877,
+      "learning_rate": 0.00010048091550612055,
+      "loss": 0.8339,
+      "step": 5231
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7960761785507202,
+      "learning_rate": 0.0001004508584974396,
+      "loss": 0.9844,
+      "step": 5232
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7719110250473022,
+      "learning_rate": 0.0001004208014480261,
+      "loss": 1.0782,
+      "step": 5233
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.839648425579071,
+      "learning_rate": 0.00010039074436059561,
+      "loss": 1.0546,
+      "step": 5234
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.8012295365333557,
+      "learning_rate": 0.00010036068723786356,
+      "loss": 0.9247,
+      "step": 5235
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7242897152900696,
+      "learning_rate": 0.00010033063008254546,
+      "loss": 1.0221,
+      "step": 5236
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 10472,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 2618,
+  "total_flos": 4.065545021544202e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}